OBJECTIVE 2: EXPERIENCES, LESSONS AND POLICY...

282
OBJECTIVE 2: EXPERIENCES, LESSONS AND POLICY IMPLICATIONS FINAL REPORT July 1999 John Bachtler and Sandra Taylor European Policies Research Centre University of Strathclyde 40 George Street Glasgow G1 1QE United Kingdom Tel: +44 141 548 3955 Fax: +44 141 548 4898 E-mail: [email protected] [email protected]

Transcript of OBJECTIVE 2: EXPERIENCES, LESSONS AND POLICY...

OBJECTIVE 2: EXPERIENCES, LESSONS AND

POLICY IMPLICATIONS

FINAL REPORT July 1999

John Bachtler and Sandra Taylor

European Policies Research Centre University of Strathclyde

40 George Street Glasgow G1 1QE United Kingdom

Tel: +44 141 548 3955 Fax: +44 141 548 4898

E-mail: [email protected] [email protected]

ii

1

TABLE OF CONTENTS

1. OBJECTIVE 2: EXPERIENCES AND LESSONS............................................. 3 1.1 Introduction ................................................................................................... 3 1.2 The Evolution of Objective 2 ......................................................................... 6

2. IMPLEMENTING OBJECTIVE 2 PROGRAMMES: OVERVIEW................. 17 2.1 Introduction ................................................................................................. 17 2.2 Regional Development Context.................................................................... 17 2.3 Regional Policy Context............................................................................... 19 2.4 Institutional Arrangements for Structural Fund Management and Delivery... 22 2.5 Management Trends..................................................................................... 25 2.6 Programme Delivery .................................................................................... 30 2.7 Conclusions and Recommendations ............................................................. 34

3. THE IMPLEMENTATION OF OBJECTIVE 2 PROGRAMMES: COUNTRY-BY-COUNTRY REVIEW....................................................................................... 39

3.1 Austria ......................................................................................................... 39 3.2 Belgium ....................................................................................................... 41 3.3 Denmark ...................................................................................................... 43 3.4 Finland......................................................................................................... 45 3.5 France .......................................................................................................... 47 3.6 Germany ...................................................................................................... 49 3.7 Italy ............................................................................................................. 51 3.8 Luxembourg................................................................................................. 52 3.9 Netherlands.................................................................................................. 53 3.10 Spain........................................................................................................ 55 3.11 Sweden .................................................................................................... 57 3.12 United Kingdom....................................................................................... 58

4. REGIONAL DEVELOPMENT PLANS AND STRATEGIES ......................... 61 4.1 Regional Development Plans and Programmes: 1989-93.............................. 61 4.2 Reprogramming of CSFs: 1992-93............................................................... 68 4.3 Regional Development Plans and Programmes: 1994-96.............................. 69 4.4 Reprogramming: 1997-99 ............................................................................ 82 4.5 Conclusions and Recommendations ............................................................. 87

5. JOB CREATION.............................................................................................. 93 5.1 Introduction ................................................................................................. 93 5.2 Context ........................................................................................................ 93 5.3 Objective 2 Programmes and Employment................................................... 97 5.4 Programme Implementation and Job Creation ............................................ 103 5.5 Conclusions and Recommendations ........................................................... 119

6. RESEARCH, TECHNOLOGICAL DEVELOPMENT AND INNOVATION. 125 6.1 RTDI Policy in Regional Development ...................................................... 125 6.2 RTDI Policy and the EC............................................................................. 126 6.3 RTDI and Objective 2 ................................................................................ 128 6.4 The 1989-1993 Programming Period.......................................................... 129 6.5 The 1994-96 Programming Period.............................................................. 131

2

6.6 The 1997-99 Programming Period.............................................................. 137 6.7 Objective 2 RTDI Policies: Approaches to Implementation........................ 142 6.8 RTDI in Objective 2: successful projects and approaches ........................... 146 6.9 Conclusions and Recommendations ........................................................... 149

7. ENVIRONMENTAL INTEGRATION IN OBJECTIVE 2 PROGRAMMES . 157 7.1 Introduction ............................................................................................... 157 7.2 The EU and Sustainable Development........................................................ 158 7.3 EU Regional Policy and the Environment .................................................. 160 7.4 Evolution of Programming and the Environment........................................ 162 7.5 The current Programming Period, 1997-99................................................. 168 7.6 Conclusions and Recommendations ........................................................... 178

8. GENDER MAINSTREAMING IN EU REGIONAL POLICY ....................... 189 8.1 Introduction ............................................................................................... 189 8.2 Context – the Evolution of Equal Opportunities Policy............................... 189 8.3 Gender Equality and Objective 2................................................................ 194 8.4 Integrating Equal Opportunities: the 1997-99 Programmes......................... 197 8.5 Equal Opportunities as a Dimension of Programme Development .............. 200 8.6 Programme Implementation ....................................................................... 212 8.7 Monitoring and Evaluation......................................................................... 215 8.8 Conclusions and Recommendations ........................................................... 217

9. MONITORING AND EVALUATION........................................................... 227 9.1 Introduction ............................................................................................... 227 9.2 The Evolution of Monitoring and Evaluation.............................................. 227 9.3 A Systematic Approach to Monitoring ....................................................... 233 9.4 Evaluation.................................................................................................. 239 9.5 Conclusions and Recommendations ........................................................... 249

10. SUMMARY AND CONCLUSIONS ......................................................... 259 10.1 Introduction............................................................................................ 259 10.2 Effective Programme Management and Delivery.................................... 260 10.3 Preparing Plans and Strategies................................................................ 263 10.4 Making Partnerships work better ............................................................ 266 10.5 Creating Jobs.......................................................................................... 268 10.6 Encouraging Innovation ......................................................................... 270 10.7 Promoting Sustainable Development ...................................................... 273 10.8 Encouraging Gender Mainstreaming ...................................................... 275 10.9 Monitoring and Evaluation ..................................................................... 277 10.10 And finally… ......................................................................................... 279

3

1. OBJECTIVE 2: EXPERIENCES AND LESSONS

1.1 Introduction This paper comprises the Final Report of the study Objective 2: Experiences, Lessons and Policy Implications prepared on behalf of DG XVI of the European Commission.

The origins of the study lie in the need to understand better the long-term evolution of programme interventions addressing industrial reconversion. Across a decade of experience of implementing Objective 2 programmes, the approach to programming has progressed significantly as actors in the process (Commission services, national officials, programme managers, secretariats, partners and beneficiaries) have gained familiarity with the principles and operational requirements of strategy development, programme management, programme implementation, partnership, monitoring and evaluation. Comparison of the design and implementation of regional development programmes over time across different EU Member States suggests that it is possible to identify a developmental process of ‘learning’ in undertaking different elements of programming.

While an enormous body of information and research on the implementation of the Funds has been generated through programme documentation, evaluation studies, Commission publications and academic research (especially in the geographical, regional economics and political science literature), the coverage, detail and accessibility of the material is variable.

The distinctive contribution of this study is to provide a synthesis and assessment of the achievements, experiences and lessons of Objective 2 interventions over the past decade. In particular, it takes both a horizontal perspective, examining commonalities and contrasts across Member States and regions, as well as a longitudinal view, identifying progressive changes in the approach to programming over time and successive programming periods – 1989-93, 1994-96 and 1997-99.

The study is characterised by three other features. First, it disaggregates the process of programming into discrete components for the purposes of analysis: plan preparation, strategy development, programme management, programme delivery, partnership, monitoring and evaluation. Second, it ensures that each aspect of programming is set within the context of wider trends and policy issues in the economic development field such as institutional change or evaluation. Third, at the heart of the analysis of Objective 2 strategies is a detailed assessment of progress being made in the so-called ‘horizontal’ priorities – job creation, RTDI, environment and equal opportunities – each of which is discussed in a separate chapter.

The outcome of the study is intended to assist understanding of past trends and to stimulate future thinking and debate about the shape of next generation of Objective 2 programmes.

This first chapter outlines the objectives and method of the study, the structure of the report and the regulatory context for the implementation of Objective 2 interventions.

4

1.1.1 Aims and Objectives The aim of the study is to provide a definitive assessment of the achievements, experiences and lessons of Objective 2 interventions under the Structural Funds. It is intended to provide an authoritative review of all activity carried out under Objective 2 to date as an input for the design and implementation of future programmes under the new Objective 2, as well as for the preparation of the Second Report on Economic and Social Cohesion.

In examining experience of Objective 2 over the decade 1989-99, the project has five specific objectives:

��to review Objective 2 reconversion strategies to support industrial restructuring;

��to analyse the main strategic options deployed in the Objective 2 programmes to support specific horizontal themes;

��to assess issues relating to programme management;

��to examine the socio-economic impact of Objective 2 interventions; and

��to derive policy implications for future programmes.

1.1.2 Method The study comprises a review of the entire experience of Objective 2 programming to date, in order to provide new insights into the achievements and experiences of Objective 2 intervention across the EU. It has used both quantitative and qualitative information sources and provides illustrative examples of successful and distinctive practice. The methodology has encompassed five tasks: a review of reconversion strategies; the analysis of strategic options to address horizontal themes; review of the delivery of Structural Funds; assessment of socio-economic impacts; and identification of policy implications.

The study has been carried out mainly through desk research, supplemented by a limited programme of interviews involving key Commission officials and selected Member State officials, senior programme managers and experts in Objective 2 regions. The desk research consisted of a synthesis and review of all available documentary and data sources available to date relating to Objective 2, especially the following.

(a) Evaluation reports. Objective 2 ex ante, interim and ex post evaluation studies carried out for the 1989-93, 1994-1996 and 1997-99 programming periods, notably: ex ante appraisal reports for 1994-96 and 1997-99; ex post evaluation of 1989-93 CSFs/OPs (Ernst &Young) and regional group evaluations; intermediate evaluation reports for 1994-96; the synthesis and extended synthesis of regional development strategy evaluations for the 1994-6 and 1995-99 SPDs (EPRC/ACT); and thematic/horizontal evaluations on partnership, RTD etc.

(b) Commission documentation. Annual Reports on the Structural Funds, plus Commission documentation eg. Communications COM (95)111, COM(96)316, and COM(97)524 and the First Cohesion Report.

5

(c) IQ-Net. Output from the EPRC-managed IQ-Net consortium Improving the Quality of Objective 2 Programmes through Exchange of Experience which has yielded a series of thematic papers on issues including programme management, monitoring and interim evaluation, Fund integration, RTDI policy approaches, equal opportunities, environmental sustainability and project quality.

(d) Member State policy information. Official and unofficial policy documentation and results from interview research with Member State policy officials carried out by EPRC.

(e) Academic literature. National and international literature on Objective 2 regions and strategies, including the ‘grey’ literature of working papers and conference papers.

The study has been managed by John Bachtler and Sandra Taylor with the assistance of Mary Louise Rooney, Ruth Downes, Rona Michie, Ross Brown, Rona Fitzgerald, Keith Clement, Christelle McKeegan and Sue Laverge. The contributions of Patricia Noble, Philip Raines, Jane Brooks-Burnett, Charlotte Damborg, Henrik Halkier and Rosemarie Rey are also gratefully acknowledged.

1.1.3 Structure of the Report The Final Report has ten chapters. Turn to the last section of the report, Chapter 10, for the summary and conclusions of the study. For each of the issues discussed in the report, the main findings are presented, accompanied by recommendations for future programming. These recommendations are set out in more detail in the relevant chapters.

Chapter 1, following this summary of the aims, method and structure, briefly reviews the key regulatory developments underpinning Structural Fund interventions under Objective 2 and the volume of support currently being provided for each Member State.

The second and third chapters begin the comparative assessment with a review of the implementation arrangements for administering Objective 2 programmes. Chapter 2 briefly reviews recent trends in the institutional infrastructure for economic development and the regional policies of Member States to provide a context for discussing how the Structural Funds (and Objective 2 programmes in particular) have been implemented in different countries and regions. The chapter also considers developments in programme management and project delivery over the past ten years. This overview is complemented by short, country-by-country surveys of implementation arrangements in Chapter 3. For each of the 12 Member States with Objective 2 areas, the national approach to regional policy is briefly summarised, particularly the role of subnational levels of government, as well as the key administrative responsibilities for implementing the Funds.

Chapter 4 is concerned with the process of preparing Objective 2 regional development plans and the shape of regional development strategies. It traces the evolution of plan preparation and strategy development through the successive programming periods, examining how the plans and strategies were

6

created, the form and content of strategies (in terms of expenditure allocation) and the strengths and weaknesses of the process at each stage.

Chapters 5 to 8 focus on four key horizontal elements of Objective 2 strategies – job creation, RTDI, environment and equal opportunities. In each case, the chapters review the background to these strategic priorities, especially the reasons for the importance accorded to them. The response of Objective 2 programmes is then examined, charting the progressive changes in approach across programming periods from 1989 to the present and drawing out conclusions, not just for the strategic content but also the delivery of these priorities.

Chapter 9 examines the complex regulatory issues of monitoring and evaluation. The chapter reviews the evolution of monitoring and evaluation over the past ten years and then discusses the process of monitoring with a systematic assessment of good practice. The evaluation of Objective 2 interventions is also considered with a comparative perspective on three major evaluation exercises conducted over the past three programming periods.

1.2 The Evolution of Objective 2 Before beginning the review of Objective 2 programming, it is useful to consider the regulatory context for the implementation of EU interventions. The following sections provide a précis of the major regulatory decisions and expenditure allocations concerning Objective 2 areas over the past decade, together with a summary of the support being provided for each Member State.

1.2.1 Pre 1988 Prior to the 1988 reform of the Structural Funds, EU support for regions undergoing industrial change was available under the 1975 European Regional Development Fund (ERDF) Regulation. The budget was allocated on the basis of predetermined, negotiated quotas, with Member States’ own regional aid maps being used. Although eligible projects for ERDF support included investment in industrial, handicraft and service projects, assistance was largely confined to grants for infrastructure investments and investments in industrial and service sector schemes. The focus was on project-by-project aid, as opposed to the current ‘programming’ approach. Financial resources were limited.

The ERDF underwent two major reforms in 1979 and 1984, and the budget slowly increased over the period. In 1979, a small ‘non-quota’ section of the ERDF was introduced to support ‘programmes’ in special eligible areas. Early non-quota programmes typically concentrated on areas affected by the decline of the textile, steel and shipbuilding industries. The ERDF also began to offer newer types of flexible assistance designed to stimulate indigenous development in depressed areas and disadvantaged inner city areas through the Integrated Development Operations which were first launched in Belfast and Naples.

The next revision of Community regional policy culminated in new guidelines in 1984, redefining the objectives of the ERDF to emphasise the development and structural adjustment of lagging regions and the conversion of declining

7

industrial regions. In a drive to make Community regional policy more effective, the Commission proposed that the quota-free part of the ERDF be extended considerably, investment be tied more closely to regional development plans and a common procedure be developed to designate areas eligible for Community regional policy assistance. This met resistance from Member States and the compromise was a prolongation of the quota system, but within flexible minimum and maximum limits.1

Alongside the ERDF, the European Social Fund (ESF), which had been in existence since 1957, now had a regional dimension with the addition in 1971 of the remit to address long-term structural unemployment and underemployment, particularly in underdeveloped regions and regions affected by decline in a principal industry.2

1.2.2 The 1988 Reform of the Structural Funds After the accession of Spain, Greece and Portugal to the Community, the increased regional imbalances were perceived as a threat to the realisation of the Single Market. The rationale behind the 1988 reforms was to improve the effectiveness of the Structural Funds, and therefore economic and social cohesion in the Community. Common criteria for designating problem regions, the scope and forms of assistance, guidelines and reporting requirements were developed. The tasks of the Structural Funds were redefined and a number of principles for the implementation of EU regional policy were introduced as follows:

��the submission of plans by the Member States under priority objectives;

��the implementation of partnership between competent authorities at national, regional and local levels;

��the additionality of Community measures;

��the compatibility of structural policy with other Community policies;

��the concentration of resources, with particular emphasis on the least prosperous regions; and

��the co-ordination and combination of different Community instruments.

Of key importance, there was agreement progressively to double the Funds over the period 1989 to 1992. The commitments for the Structural Funds as a whole in 1987 amounted to 7,200 MECU; by 1993 this reached 14,500 MECU (1989 prices). Also central was the notion of co-ordinating the Funds around a set of six policy objectives to which the Funds were individually or jointly assigned:

��Objective 1 - promoting the development and structural adjustment of the regions whose development is lagging behind;

��Objective 2 - converting regions seriously affected by industrial decline;

1 Michie R and Fitzgerald R (1997) The Evolution of the Structural Funds in J Bachtler and I Turok (Eds.) (1997) The Coherence of EU Regional Policy: Contrasting Perspectives on the Structural Funds, Jessica Kingsley, London. 2 Yuill D, Allen K, Bachtler J, Wishlade F and Clement K (Eds.) (1990) European Regional Incentives, 1990-91, 10th Edition, Bowker Saur, London.

8

��Objective 3 - combating long-term unemployment;

��Objective 4 - facilitating the occupational integration of young people;

��Objective 5a - speeding up the adjustment of agricultural structures and improvement of conditions for fish processing and marketing; and

��Objective 5b - promoting the development of rural areas.

Objectives 1, 2 and 5b were specifically regional in nature, involving measures restricted to certain eligible regions or parts of regions. This was a major innovation in European regional policy. For the first time, an area designation typology was applied across the Community and the European Commission played the central role in area designation for the Funds.

Unlike Objective 1, the Objective 2 regions were not set out in the new Structural Funds regulations, which only indicated the designation criteria to be applied. Regions were to be designated at NUTS level III and had to have: an average level of unemployment higher than that of the Community during the previous three years; industrial employment, as a percentage of employment, in excess of the Community average in any of the previous 15 years; and an observable fall in industrial employment relative to the reference year.

In addition to this essentially quantitative basis for area designation, the Regulations indicated that Objective 2 regions could include: areas adjacent to those satisfying the main criteria; urban communities with an unemployment rate at least 50 percent above the Community average which had recorded a substantial fall in industrial unemployment; and other areas which had recorded substantial job losses over the previous three years or were experiencing or threatened with such losses in industrial sectors vital to their economic development, with a consequent serious worsening of their unemployment.

The designation of Objective 2 regions was officially announced in March 1989. On the basis of the areas designated, a three-stage process was implemented - Member States submitted regional development plans; on the basis of these plans, the Commission, in consultation with Member States, drew up the Community Support Frameworks (CSFs) which set out the priorities for assistance, the forms of assistance and the sources and levels of funding; and Operational Programmes (OPs) were formulated on the basis of the Community Support Frameworks. The OPs outlined operational aid measures submitted by the Member States in the form of applications for assistance adopted by the Commission. Agreement on the CSFs was not reached until the end of 1989; CSFs for 53 of the 60 Objective 2 regions were adopted on 20 December 1989. The principal Operational Programmes within the CSFs were finally approved in July 1990.

For Objectives 1 and 5b, the programming period was five years (1989-93), but, in view of the perceived potential for quicker change in relation to industrial conversion, Objective 2 programmes were intended to operate over two periods, 1989-1991 and 1992-1993. However, the protracted negotiation process and resulting delays in agreeing the CSFs and OPs meant that regions only initially benefited from their status for two years, rather than the full three-year period. Reflecting these difficulties, in April 1991 the Commission

9

approved a proposal to maintain the original list of Objective 2 areas for a further two years until the end of 1993.3

Table 1.1: Objective 2 allocations, 1989-93

Sources: * CEC (1996) First Report on Economic and Social Cohesion, Luxembourg. ** Ernst & Young (1997) Ex Post Evaluation of the 1989-93 Objective Programmes: Synthesis Report, DGXVI Series: Evaluation and Documents, No.4, Commission of the European Communities, Brussels.

In total, Objective 2 CSF allocations for the 1989-93 period amounted to about 6.1 billion ECU, representing 9.7 percent of the total Structural Fund allocation during that programming period. The UK received some 40 percent and Spain and France were also allocated significant shares.

In August 1990, the list of Objective 2 areas was extended to include areas qualifying for support under the RECHAR Community Initiative for coal-mining areas, which previously fell outside the scope of the Structural Funds.

1.2.3 The 1993 Reform The negotiation of the Maastricht Treaty in 1991 reaffirmed the Community’s commitment to economic and social cohesion, and the Commission subsequently produced proposals based on the policies agreed at Maastricht. This was followed in March 1992 by the Commission’s mid-term review of the EU structural policies. These documents recommended specific changes to the Structural Funds which eventually led to the 1993 reform of the Structural Funds. The changes introduced under the 1993 reform did not represent the kind of radical overhaul that took place in 1988; there was an explicit

3 Yuill D, Allen K, Bachtler J, Wishlade F and Clement K (Eds.) (1992) European Regional Incentives, 1992-93, 12th Edition, Bowker Saur, London.

Member State

Structural Fund

allocation*

ERDF ** ESF ** % of national population

covered

Number of programmes

Belgium 214.0 154.3 59.3 22.1 5 Denmark 25.0 18.6 6.2 4.9 2 France 1225.0 993.3 232.0 18.3 17 Germany 581.0 420.5 160.2 12.4 7 Greece - - - 0.0 0 Ireland - - - 0.0 0 Italy 387.0 272.1 114.8 6.6 9 Luxembourg 12.0 11.0 0.5 38.0 1 Netherlands 165.0 106.0 58.7 9.9 3 Portugal - - - 0.0 0 Spain 1506.0 1134.4 205.3 22.2 1 UK 2015.0 1455.4 475.0 35.5 9 Total 6130.0 4565.6 1312.0 16.8 54

10

commitment to retain the main principles of the earlier reform.4 On the other hand, the revised Regulations contained important new developments, most notably:

��an increase in Structural Fund resources to 27.4 billion ECU by 1999 (1992 prices), 70 percent of which was to be concentrated on Objective 1 regions;

��an extension of the areas eligible for Objective 1 and changes to the procedures for designating Objectives 2 and 5b;

��a simplification of decision-making procedures involving a reduction in the number of planning stages from three to two, with a Single Programming Document (SPD) replacing the CSF and OP except in most of the Objective 1 regions; and

��the introduction of systematic evaluation.

With respect to area designation processes, important changes were made to the procedures for selecting Objective 2 areas. Under the revised Regulations, Member States took the lead role in proposing areas to be designated under Objective 2. Of particular importance, the Member States could take account of national policy priorities, and use could be made of national statistical data in devising area designation proposals. This represented a significant change over the position in the 1989-93 period when the Commission had been much more influential in the process.

For Objective 2, the primary designation criteria (at NUTS III level) continued to be above (EU) average rates of unemployment and industrial employment and decreases in industrial employment. Under secondary criteria, eligibility could also be extended to certain other areas seriously affected by unemployment or adjacent to areas which met the main criteria. The eligibility criteria were to be reviewed after three years as of 1 January 1997.

The scope of the Structural Funds was broadened somewhat during the 1993 reform. In Objective 2 areas (as in other regions) the coverage of ERDF and ESF measures was broadened to include research and development and education investments.

4 Bachtler J, Michie R and Wishlade F (1993) European Community Regional Policy, in Yuill D, Allen K, Bachtler J and Wishlade F (Eds.) European Regional Incentives, 1993-94, 13th Edition, Bowker Saur, London.

11

The list of areas was approved in early 1994 on the basis of proposals submitted by Member States. The Member States’ proposals accounted for population coverage proportions well in excess of the Commission’s indicative ceiling specified in the regulations so a certain amount of ‘adjustment’ was necessary.5 The final outcome was coverage of 16.8 percent of the Community population for Objective 2. Compared to the previous funding period (1989-93) the coverage of Objective 2 decreased in Belgium, Germany, Spain, Luxembourg and the UK, although the total coverage remained stable. When the three new Member States acceded to the EU in 1995, a further 2.3 million people were added to the population covered by Objective 2, bringing the total to 6.46 million or 16.4 percent of the total population.

As can be seen from the table, the UK was once again one of the major beneficiaries of Objective 2 funding, closely followed by France.

Following Member State criticisms regarding the complexity of planning and decision-making procedures during the 1988-93 programming period, the three phases of programming were reduced to two, and Member States were able to present simultaneously their development plans and draft programmes as part of a single document. The next step after the selection of eligible areas

5 Bachtler J, Michie R and Wishlade F (1994) ‘EU Regional Policy: Implementation of the Revised Structural Funds' in Yuill D et al (Eds.) European Regional Incentives 15th Edition, Bowker Saur, London.

Table 1.2: Objective 2 allocations, 1994-96 Member State Total

Structural Fund

allocation

Of which, ERDF

1994-96

Of which, ESF

1994-96

% of national population

covered

Number of programmes

Austria* 7.5 4 Belgium 160.0 130.0 30.0 14.2 4 Denmark 56.0 44.2 11.8 8.5 2 Finland** 69.2 55.2 14.0 25.1 1 France 1763.3 1452.6 310.7 25.1 19 Germany 733.0 513.7 219.3 8.8 9 Greece - - - 0.0 0 Ireland - - - 0.0 0 Italy 684.0 542.3 141.7 11.0 11 Luxembourg 7.0 6.0 1.0 34.6 1 Netherlands 300.0 206.0 94.0 17.4 5 Portugal - - - 0.0 0 Spain 1130.0 870.1 259.9 20.4 8 Sweden* 11.5 5 UK 2142.0 1606.9 535.1 30.9 13 Total 7044.5 5427.0 1617.5 16.4 82 * 1995-99 – see Table 1.3 for financial details, ** 1995-96 Source: CEC (1995) The New Regional Programmes under Objectives 1 and 2 Commission of the European Communities COM(95)111 final, 29.03.95, Brussels.

12

was the submission of Regional Development Plans by Member States for ex ante appraisal and negotiation to Single Programming Document (or CSF) stage. The Objective 2 plans were submitted in March/April 1994, with SPD approval taking place in the latter part of 1994.

Table 1.3: Objective 2 allocations, 1997-99 Member State Total

Structural Fund

allocation

Of which, ERDF

1997-99

Of which, ESF

1997-99

% of national population

covered

Number of programmes

Austria* 108.2 73.9 34.3 7.5 4 Belgium 216.2 173.1 43.1 14.2 4 Denmark 68.2 52.7 15.5 8.5 2 Finland 135.3 105.5 29.8 25.1 1 France 2246.3 1852.1 394.2 25.1 19 Germany 901.1 639.9 261.2 8.8 9 Greece - - - 0.0 0 Ireland - - - 0.0 0 Italy 967.8 782.7 185.1 11.0 11 Luxembourg 9.8 8.0 1.8 34.6 1 Netherlands 442.2 292.3 149.9 17.4 5 Portugal - - - 0.0 0 Spain 1485.0 1132.4 352.6 20.4 8 Sweden* 160.0 121.7 38.3 11.5 5 UK 2675.8 2014.2 661.6 30.9 13 Total 9381.6 7248.5 2133.1 16.4 82 * 1995-99 Source: CEC (1997) The New Regional Programmes 1997-99 under Objective 2, Commission of the European Communities COM 97 (524) final, 14.11.97, Brussels.

The 1993 reforms specified a new six-year programming period (1994-1999), again except for Objective 2, for which the period was split into two three-year phases. Again, it was thought that the speed of transition in industrial regions might require (or allow for) changes to the areas of the programmes after three years. In fact, the changes were very limited, involving marginal adjustments to partially eligible urban areas of Spain (Zaragoza, Barcelona, Palma de Mallorca, Madrid), Italy (Pisa, Milan, Varese, Reggio Emilia, Modena, Rome, Trieste, Florence and Venice) and the Netherlands (Overig Groningen). The re-programming process took place during 1996, with most programmes approved between March and August 1997. Structural Fund allocations under Objective 2 during the 1997-99 programming period once again heavily favoured the UK, with France and Spain being the other major beneficiaries.

1.2.4 Objective 2 in the Member States In the 1994-99 period, Objective 2 allocations ranged from 15 MECU in Luxembourg and 99 MECU in Austria to 4.6 billion ECU in the United Kingdom, for Objective 2 areas ranging from eight percent of the national

13

population in Austria to 31 percent in the UK and 35 percent in Luxembourg. Individual Member State allocations (in 1994 prices) and areas are as follows.

��Austria - 1.57 billion ECU from the Structural Funds over the 1995-99 period, of which 1.43 billion ECU is being distributed through mainstream programmes and 143 MECU through Community Initiatives. Regional expenditure is managed through one Objective 1 programme (Burgenland), four Objective 2 programmes and seven Objective 5b programmes. The Objective 2 allocation for 1995-99 for the regions of Niederösterreich, Oberösterreich, Steiermark and Vorarlberg is 99 MECU. The Structural Fund areas cover 39.9 percent of the national population, 7.5 percent under Objective 2.

��Belgium - 2.1 billion ECU from the Structural Funds over the 1994-99 period, of which 1.8 billion ECU is being distributed through mainstream programmes and the remaining 287 MECU through Community Initiatives. Regional expenditure is managed through one Objective 1 programme (Hainaut), four Objective 2 programmes and three Objective 5b programmes. The Objective 2 allocation for 1994-99 for the regions of Liège, Limburg, Turnhout and Aubange is 342 MECU. The Structural Fund areas cover 31.5 percent of the national population, 14.2 percent under Objective 2.

��Denmark - 843 MECU from the Structural Funds over the 1994-99 period, of which 741 MECU is being distributed under mainstream programmes and the remaining 102 MECU through Community Initiatives. Regional expenditure is managed through two Objective 2 programmes and one Objective 5b programme. The Objective 2 allocation for 1994-99 for the regions of Lolland and Nordjylland is 119 MECU. The Structural Fund areas cover 15.3 percent of the national population, 8.5 percent under Objective 2.

��Finland - 1.65 billion ECU from the Structural Funds over the 1995-99 period, of which 1.1 billion ECU is being distributed through mainstream programmes and the remaining 150 MECU through Community Initiatives. Regional expenditure is managed through one Objective 6 programme, one Objective 2 programme and two Objective 5b programmes (one for mainland Finland and one for the Åland Islands). The Objective 2 allocation for 1995-99 is 179 MECU and covers six areas in a single SPD: West Coast (Varsinais Suomi and Satakunta), South Karelia, Central Ostrobothnia (Kokkola), Päijät-Häme, Eastern Gulf (Ita Uusimaa and Kymenlaakso) and Central Finland. The Structural Fund areas cover 54 percent of the national population, 15.7 percent under Objective 2.

��France - 14.94 billion ECU from the Structural Funds over the 1994-99 period, of which 13.34 billion ECU is being distributed through mainstream programmes and 1.6 billion ECU through Community Initiatives. Regional expenditure is managed through six Objective 1 programmes (for Nord Pas-de-Calais in mainland France, plus Corsica and overseas territories), 19 Objective 2 programmes and 20 Objective 5b programmes. The Objective 2 allocation for 1994-99 is 3.77 billion ECU for the regions of Alsace, Aquitaine, Auvergne, Basse-Normandie,

14

Bourgogne, Bretagne, Centre, Champagne-Ardenne, Franche-Comté, Haute-Normandie, Languedoc-Roussillon, Lorraine, Midi-Pyrénées, Nord-Pas-de-Calais, Pays de la Loire, Picardie, Poitou-Charentes, Provence-Alpes-Côtes d’Azur and Rhône-Alpes. The Structural Fund areas cover 46.2 percent of the national population, 25.1 percent under Objective 2.

��Germany - 21.72 billion ECU from the Structural Funds over the 1994-99 period, of which 19.52 billion ECU is being distributed through mainstream programmes and 1.6 billion ECU through Community Initiatives. Regional expenditure is managed through six Objective 1 programmes (for the eastern German Länder), nine Objective 2 programmes, and eight Objective 5b programmes. The Objective 2 allocation for the 1994-99 period is 1.57 billion ECU for the eligible areas in Bayern, Bremen, Hessen, Niedersachsen, Nordrhein-Wesfalen, Rheinland-Pfalz, Saarland, Schleswig-Holstein and West Berlin. The Structural Fund areas cover 39.1 percent of the national population, 8.8 percent under Objective 2.

��Italy - 21.65 billion ECU from the Structural Funds over the 1994-99 period, of which 19.75 billion ECU is being distributed through mainstream programmes and 1.6 billion ECU through Community Initiatives. Regional expenditure is managed through eleven Objective 2 programmes. The Objective 2 allocation for the 1994-99 period is 1.46 billion ECU for the regions of Emilia-Romagna, Friuli-Venezia-Giulia, Lazio, Liguria, Lombardia, Marche, Piemonte, Toscana, Umbria, Valle d’Aosta and Veneto. The Structural Fund areas cover 56 percent of the national population, 11 percent under Objective 2.

��Luxembourg - 104 MECU from the Structural Funds over the 1994-99 period, of which 84 MECU is being distributed through mainstream programmes and 20 MECU through Community Initiatives. Regional expenditure is managed through a single Objective 2 programme and a single Objective 5b programme. The Objective 2 allocation for the 1994-99 period is 15 MECU for Esch-zur-Alzette and Capellen. The Structural Fund areas cover 42.5 percent of the national population, 34.6 percent under Objective 2.

��The Netherlands - 2.62 billion ECU from the Structural Funds over the 1994-99 period, of which 2.19 billion ECU is being distributed through mainstream programmes and 421 MECU through Community Initiatives. Regional expenditure is managed through one Objective 1 programme (Flevoland), five Objective 2 programmes and six Objective 5b programmes. The Objective 2 allocation for the 1994-99 period is 650 MECU for the regions of Arnhem-Nijmegen, Groningen-Drenthe, Zuidoost-Brabant, Twente-Overijssel and Zuid-Limburg. The Structural Fund areas cover 24.3 percent of the national population, 17.4 percent under Objective 2.

��Spain - 34.44 billion ECU from the Structural Funds over the 1994-99 period, of which 32.44 billion ECU is being distributed through mainstream programmes and 2.87 billion ECU through Community Initiatives. Regional expenditure is managed through Objective 1 interventions (including several multi-regional programmes and 13

15

regional programmes), eight Objective 2 programmes and seven Objective 5b programmes. The Objective 2 allocation for the 1994-99 period is 2.45 billion ECU for the regions of la Rioja, Aragón, Baleares, Cataluña, Navarra, Madrid and País Vasco, each with an operational programme (combining ERDF and ESF), and an additional ESF programme covering all of the regions. The Structural Funds cover 84.5 percent of the national population, 20.4 percent under Objective 2.

��Sweden - 1.38 billion ECU from the Structural Funds over the 1995-99 period, of which 1.25 billion ECU is being distributed through mainstream programmes and 125 MECU through Community Initiatives. Regional expenditure is managed through an Objective 6 programme, five Objective 2 programmes and five Objective 5b programmes. The Objective 2 allocation for the 1994-99 period is 157 MECU for the regions of Ångermanlandskusten, Blekinge, Fyrstad, Norra Norrlandskusten and Bergslagen. The Structural Funds cover 26 percent of the national population, 11.5 percent under Objective 2.

��The United Kingdom - 13.16 billion ECU from the Structural Funds over the 1994-99 period, of which 11.59 billion ECU is being distributed through mainstream programmes and 1.57 billion ECU through Community Initiatives. Regional expenditure is managed through three Objective 1 programmes, 13 Objective 2 programmes and 11 Objective 5b programmes. The Objective 2 allocation for the 1994-99 period is 4.58 billion ECU for the regions of East Midlands, West Cumbria and Furness, Plymouth, Thanet, West Midlands, Gibraltar, Greater Manchester, Lancashire and Cheshire, East London and the Lee Valley, North-East England, Yorkshire and Humberside, Western Scotland, Eastern Scotland and Industrial South Wales. The Structural Funds cover 41.7 percent of the national population, 30.9 percent under Objective 2.

17

2. IMPLEMENTING OBJECTIVE 2 PROGRAMMES: OVERVIEW

2.1 Introduction The management of Structural Fund intervention is strongly influenced by national institutional and policy contexts. These contexts are dynamic: most EU Member States have witnessed profound changes in the scope, substance and institutional management of regional development over the past two decades. National governments continue to administer their own regional policies, many dating back to the 1950s and 1960s, but regional development is no longer the sole prerogative of central government, and the institutional environment has become increasingly unstable with the involvement of sub-national authorities and groups. The implementation of the Structural Funds has been shaped by these trends, through successive reforms of EU regional policy from 1988 to the present. In turn, the Funds have themselves also been a contributor to the institutional instability, influencing the changing relationships between and within levels of government and the administration of regional development.

The following two chapters review the institutional and programming arrangements for the management and administration of Objective 2 programmes. This chapter begins with a discussion of the dynamic institutional environment for economic development and then reviews the institutional structures used for implementing the regional policies of the Member States. The institutional arrangements for implementing Structural Funds, and specifically for Objective 2 programmes, are then discussed, highlighting recent trends and patterns across the Member States. The subsequent chapter then provides a systematic account of Objective 2 implementation in each of the 12 Member States with eligible Objective 2 areas.

2.2 Regional Development Context In recent years, institutional structures have been one of the most dynamic areas of regional development. The primacy of central government institutions in the field of regional development has been complemented or challenged by the involvement of local and regional organisations engaged in economic development activities. A trend towards regionalisation is evident, through the deconcentration or devolution of national government responsibilities, and there is a growing role for private sector organisations as partners or delivery agents for government authorities. At every level, the structure, organisation, responsibilities and functions of government organisations have been subject to review, and new forms of interrelationship are being sought within and between levels of government.

The growing role for sub-national levels in economic development emanates partly from regions and localities themselves. With competition among localities to provide a favourable environment for investment, regions are under pressure to improve competitiveness or face losing out in terms of processes such as new firm formation and innovation. Greater social, political

18

and cultural awareness of regional identity is associated with regions taking more direct responsibility for their own socio-economic development.

A further contributory factor is the changing role of national government. Many northern European countries have seen a decline in the priority and expenditure allocated to regional policy over the past 10-15 years.6 Policy objectives have shifted away from promoting the inter-regional distribution of economic growth towards encouraging regional structural change. In some cases, central authorities have attempted explicitly to encourage greater initiative and flexibility at regional and local levels, anticipating that locally managed development programmes would be more efficient and realistic.7 In others, the progressive withdrawal of central government from many areas of direct economic intervention, including a lessening commitment to traditional regional policy, has stimulated new forms of economic policy which are rooted in the specific requirements of local communities.8 It has been argued that the regionalisation of economic development is more effective, providing closer proximity to factor markets, a better knowledge of local entrepreneurial capacity, greater ease in the identification of policy targets through local networks, more opportunity for innovative approaches, the engendering of regional identity, and quicker responsiveness to social and economic problems.9

A second trend in the institutional context for regional development is the growing commitment to co-operation among regional development organisations. This applies to vertical co-operation at both national and regional level, and also to the new forms of horizontal inter-relationship between different levels of government. In attempting to take a strategic perspective towards regional development, notably improvements in regional competitiveness, policy-makers are having to address issues that go beyond support for productive investment and employment to encompass research and technological development, information and advice, environmental quality and amenity, and human resources. Such issues require co-ordinated or integrated policy approaches, forcing different policy organisations to work together.

A further reason is the need for flexibility, requiring a co-operative approach to the formulation and implementation of regional development, and involving partnerships of locally-based coalitions of interest which represent the full range of relevant organisations and levels of government. More emphasis is

6 Bachtler J (1997) New Dimensions of Regional Policy in Western Europe, in M Keating and J Loughlin (Eds.) The Political Economy of Regionalism, Frank Cass, London, pp77-89. 7 Rousseau M O and Zariski R (1987) Regionalism and Regional Devolution in Comparative Perspective, Praeger Publishers, New York. 8 Roberts P (1989) Local Economic Development - Alternative Forms of Local and Regional Policy for the 1990s, in L Albrechts, F Moulaert, P Roberts, and E Swyngedouw (Eds.) (1989) Regional Policy at the Crossroads, Regional Studies Association, Jessica Kingsley Publishers, London. 9 Begg I (1994) The Case for Decentralised Industrial Policy, in G Hajdú and G Horváth (Eds.) (1994) European Challenges and Hungarian Responses in Regional Policy, Centre for Regional Studies, Pecs. Armstrong H and Taylor J (1993) Regional Economics and Policy, Harvester Wheatsheaf, London. Pickvance C and Preteceille E (1991) State Restructuring and Local Power - A Comparative Perspective, Pinter Publishers, London.

19

placed on interactions, modes of co-operation and network activities than on the functional hierarchies between levels of governments.10

Although regional and local levels may be increasingly regarded as the most appropriate level for implementing regional development policies, central government is far from becoming obsolete in terms of its involvement in regional economic development. Central government maintains an active role in seeking solutions to regional problems - whether for national economic efficiency or social equity. Central involvement is necessary to ensure that regional policy is adequately funded in the regions of greatest need, to ensure the effective co-ordination of regional policy and to identify and promote regional concerns among other ‘non-spatial’ national economic policies. In several key policy areas of relevance for regional competitiveness, pressures towards regionalisation are not practicable, for example the regulation of corporate governance, national systems of innovation, the labour market and the industrial relations system. Further, some Member States have resisted any substantial rise in the autonomy and powers of regional and local authorities, unless mediated by regional offices of the State.11

The reshaping of territorial organisations, and the reallocation of regional development responsibilities with shifts in autonomy, is controversial.12 Notwithstanding the rise of ‘meso-government’, the future governance of regional policy in larger EU states is far from clear.

2.3 Regional Policy Context Against this institutional backcloth, the regional policy context for implementing the Structural Funds within individual EU countries is highly differentiated. Virtually every Member State has a national regional policy, designed and administered to address its individual regional problems and disparities, but the definition and coverage of policy vary greatly.13 Member States prioritise the policy objectives of efficiency and equity differently: in Germany and Italy there is a constitutional commitment to equalisation of income or living conditions, and, as in the Nordic countries (where there is also a strong emphasis on equity) regional policy has a high profile. Elsewhere efficiency objectives are more important, for example in the regional policies of the UK, France and the Netherlands, where spending on regional policy is relatively small.

In some countries, regional policy cuts across a wide range of policy areas. The Nordic regional policy tradition involves nearly all government

10 Friedmann, J (1991) The industrial transition: A comprehensive approach to regional development, in E M Bergman, G Maier, and F Tödtling (1991) Regions Reconsidered - Economic Networks, Innovation and Local Development in Industrialised Countries, Mansell Publishing Ltd, London. Hansen N, Higgins B, and Savoie B J (1990) Regional Policy in a Changing World, Plenum Press, New York. Balme R, Garraud P, Hoffmann-Martinot V, le May S and Ritaine E (1994) Analysing territorial policies in Western Europe, European Journal of Political Research, Vol. 25:4, June. 11 Wiehler F and Stumm T (1995) The powers of regional and local authorities and their role in the European Union, European Planning Studies, 2, p.227-251. 12 Balme R (1997) Regional Policy and European Governance, in M Keating and J Loughlin (Eds.) op. cit. 13 Yuill D, Bachtler J and Wishlade F (1999) European Regional Incentives, 1998-99, Bowker-Saur, London. Bachtler J (1999) Regional Policy in Europe, in P Coffey (Ed.) Europe – Toward the Next Enlargement, Kluwer.

20

departments having to take account of the regional dimension, and several ministries pursue ‘regional policies’. In France, regional policy is considered to be aménagement du territoire, literally the management of space, to which several policy areas contribute. By contrast, regional policy in the UK is defined as ‘regional industrial policy’, operated almost exclusively through financial incentives to industry.

Regional policy has traditionally been exclusively (or significantly) a central government responsibility. This still applies in unitary states, such as the UK or Finland, where both the design and administration of regional policy are undertaken by central government ministries. A variant of the centralised approach to regional policy management is delegated administrative authority, embodying a separation of political and administrative responsibilities for policy. This approach, which characterises the systems operating in Sweden and France, involves an administrative unit with an ‘arms length’ relationship with a parent ministry. In both systems, a feature over the past decade has been a progressive decentralisation of administrative responsibilities for regional policy to regional offices of the State. More advanced examples of this trend, with devolution rather than decentralisation, are the approaches to regional policy in the Netherlands, Spain and Denmark, where policy responsibility has been transferred to regional authorities. At the end of the institutional spectrum is the regionalised approach to regional policy implementation, where the primary responsibility for regional development is devolved to regions. This is reflected in the federal countries - Austria, Germany and Belgium - which generally have a constitutional allocation of responsibilities to regional levels of government.

The relationship between these national regional policies and the Structural Fund Objective 2 programmes differs greatly. As the following chapter makes clear, some countries (eg. France) consider EU regional policy to be complementary to their own national policies, with considerable integration in the administration of the two policy areas. In both Austria and Germany, the Structural Funds have been largely integrated within the existing support system for administering economic development, channelling EU regional aid to beneficiaries largely through existing Federal and Land regional and sectoral assistance programmes. Elsewhere, the Structural Funds and national regional policy have been kept completely separate (as in the UK and Sweden): while the same administrative systems are used (Government Offices in the UK, CABs in Sweden), the focus of national regional policy is on regional subsidies to strategic enterprises (Regional Selective Assistance in the UK, Regional Development Grants in Sweden), and the Objective 2 programmes prioritise support for the business environment and small business.

In many cases, there has been a tension between the national and regional policy priorities of Member States and the emphases and coverage of the Structural Funds. “The extraordinary breadth of the ERDF concept and the strong emphasis on measures to foster ‘endogenous development’ have led to

21

conflicts with member states having a more narrow perception of what is conducive to ‘regional development’.”14

Interestingly, over time there is some evidence of convergence between EU and Member State regional policies.15 During the 1990s, legislation in Italy and Germany modified the coverage of national regional policies to improve their coherence with the Structural Funds: in Italy through an extension of regional policy (under legislation passed in 1992) from ‘special intervention’ in the Mezzogiorno to a broader regional development policy for problem areas throughout the country; and in the German case (in 1995) by adapting the eligibility of the ‘Joint Task’ (Gemeinschaftsaufgabe) Framework to incorporate ‘non-investment projects’ such as consulting, training, human capital and R&D within the eligibility conditions for regional assistance. In both cases, a primary motive was to utilise Structural Fund receipts on a broader basis. In Germany, national regional policy is also assisting the preparation of ‘regional concepts’ to enhance the capacity of sub-regional strategic thinking and planning.

Both Finland and Sweden have undertaken some decentralisation of the management of regional development as a result of introducing the Structural Funds, as well as adapting their policies.16 In Finland, the priorities of the Structural Funds were largely taken over into the priorities for national regional policy as part of the preparations for EU accession, widening the scope of regional policy from its traditional focus on enterprise-based support. Most recently, legislation in Sweden introduced ‘regional growth agreements’ on the Structural Fund model into the national regional policy system, encouraging local collaboration in the development of regional strategies and improving the coordination of state resources at the local level.

These trends should not be overstated. In some cases, the adaptations to the Funds are relatively marginal or small-scale, in others the changes can only be partly attributed to the Structural Funds. Indeed, some have argued17 that there has been no recognisable direct effect of the Funds’ delivery system on national administrative structures and systems. Rather, the influence of the Funds has been restricted to indirect, procedural innovations in the planning, management and evaluation of programmes and strengthening the participation of regional actors in the implementation of economic development strategies.

14 Conzelmann T (1998) ‘Europeanisation’ of Regional Development Policies? Linking the Multi-Level Governance Approach with Realities of Learning and Policy Change, European Integration Online Papers, 2(4) http: //eiop.or.at/eiop 15Yuill D, Bachtler J and Wishlade F (1999) op.cit. 16 CEC (1996) Communication of the Commission on the implementation of EU regional policies in Austria, Finland and Sweden, Commission of the European Communities, COM(96) 316 final, 3.7.1996, Brussels. 17 Lang J, Naschold F and Reissert B (1998) Reforming the implementation of European Structural Funds: a next development step, FS II 98-202, Wissenschaftszentrum Berlin für Sozialforschung, Berlin.

22

2.4 Institutional Arrangements for Structural Fund Management and Delivery Implementing the Structural Funds within national policy contexts has been a complex process. In each Member State, the European Commission, national government and sub-national actors have different degrees of control and influence, reflecting factors such as the existence of regional institutions, the allocation of competencies between national, regional and local levels, political interests and linkages, the scale of EU funding, the number and scope of programmes, and administrative experience of managing economic development. Complexity is also attributable to the fluidity and dynamism of administrative arrangements: implementation structures and delivery systems encompass both formal government institutions and temporary coalitions of public and private actors. In many cases, national institutional systems have progressively been challenged over successive programming periods. Tensions between European, state and regional/local actors are reorganising political and power relationships between levels of government (for example through internal administrative decentralisation) or creating opportunities for new actors to become involved in resource allocation and policy formulation.

The scale of EU regional policy intervention also varies. Whereas Denmark received 843 MECU over the 1994-99 period, implemented through three programmes (two of them under Objective 2), the United Kingdom was allocated 13.16 billion ECU, distributed through 27 programmes (13 under Objective 2). In these circumstances, it is hardly surprising that each Member State is implementing the Structural Funds in a distinctive manner.

The proliferation of participants and their multiple interactions produce a variety of implementation arrangements between (and sometimes within) countries which inhibits generalisation. On the one hand, it has been argued that national governments have a pre-eminent role as ‘gatekeepers’ for the Structural Funds,18 centralising decision-making and restricting the involvement of sub-national actors. Others observe greatly enhanced regional involvement in policymaking and prominence of regional interests, although in diverse ways.19 For others the local impact is more significant: rather than promoting a Europe of regions, the Structural Funds “are instead encouraging new configurations in the heart of every region”.20

The problem with analysing the Structural Funds is that all these statements have some validity, even within the context of one Member State. In the UK, for instance, there was little scope for policy initiative for sub-national authorities during the early 1990s; most aspects of Structural Fund management were centrally controlled by the Department of Trade and Industry, with the original UK CSFs being “drawn up within strict Whitehall

18 Anderson J (1991) Skeptical reflections on a Europe of regions: Britain, Germany and the European Regional Development Fund, Journal of European Public Policy, 10(4). Martin S (1997) The Effects of EU Regional Policy in Local Institutional Structures and Policies, in J Bachtler and I Turok (Eds.) The Coherence of EU Regional Policy: Contrasting Perspectives on the Structural Funds, Jessica Kingsley Publishers, London, pp51-66. 19 Keating M and Hooghe L (1998) By-passing the nation state? Regions and the EU policy process, in J Richardson (ed.) European Union: Power and Policy-Making, Routledge, London, pp216-229. 20 Smith A (1998) The sub-regional level: Key battleground for the Structural Funds? in P Le Galès and C Lequesne C (Eds.) Regions in Europe, Routledge, London, pp50-66.

23

and Treasury guidelines, leaving little scope for local experimentation”.21 In the course of the 1990s, commentators identified a (re)discovery of regional thinking and planning22 and the emergence of new, innovative patterns of institutional arrangements at local levels.23 The unevenness of central-local relations in the UK, variously described as creative, chaotic and anarchic, may be extreme, but it applies to many other Member States which have variable and dynamic institutional arrangements for managing the Structural Funds.24

Against this background, it is difficult to present a typology of institutional arrangements for implementing the Funds. At a general level, however, it is possible to distinguish between two broad approaches among Member States.

The first group of countries are those exercising strong central government control on implementation. It is a feature of virtually all EU Member States that central government plays an important role in Structural Fund implementation, certainly in negotiating with Commission services and designating assisted areas, but also in supervising strategy formulation, expenditure, monitoring and evaluation. In some countries this central government role is much more marked than in others, as in France, Sweden, the UK, Luxembourg, Denmark and Finland. Either central government departments (the National Agency for Industry & Trade in Denmark, the Ministry of Interior in Finland, the Ministry for Industry & Trade in Sweden) or the regional representatives of central government (eg. the Regional Prefects in France or Government Offices in England) chair the Monitoring Committees and take responsibility for the final funding decisions, although some aspects of programming management may be devolved or delegated to special executives or committees.

In a second group of countries, implementation is devolved wholly or partly to units of regional self-government. Programming (including budgetary resources and some Commission liaison) is decentralised to the regions, and the role of central government is restricted to general co-ordination of the Funds, higher level Commission negotiation, inter-ministerial co-ordination and evaluation of ‘good practice’. This applies mainly to Germany where the Länder are legally responsible for regional development; the Federal Government chairs the Monitoring Committees but the Länder are responsible for implementation. Substantial devolution also applies to the Netherlands, Austria and Belgium where there is a high degree of regional autonomy (complete regional autonomy in the case of Belgium) over the implementation process, and programmes are largely managed and controlled by regional authorities (provinces). In Italy and Spain, too, the management of Objective 2 programmes (including the chairing of Monitoring Committees and some

21 Keating M and Jones B (1995) The UK Experience, in B Jones and M Keating (Eds.) The European Union and the Regions, Clarendon Press, Oxford, pp89-113. 22 Roberts P and Hart T (1997) The Design and Implementation of European Programmes for Regional Development in the UK: A Comparative Review, in J Bachtler and I Turok (Eds.) op. cit., pp193-216. 23 Dabinett G (1997) The EU Community Initiatives and the Management of Industrial Change in the UK, in J Bachtler and I Turok (Eds.) op.cit. pp109-122. 24 Roberts P et al (1997) One Union, Fifteen Systems: European Union regional programmes - lessons from practice and a review of future options, Paper to the European Urban & Regional Research Network ‘Regional Frontiers’ Conference, Frankfurt/Oder, September.

24

Commission liaison) is largely devolved to regional authorities - the Autonomous Communities in Spain, the regions in Italy.

With respect to the ministerial responsibilities for managing the Structural Funds among Objective 2 Member States, it is universally the case that the national ministry of finance or treasury is the government department which receives all monies from the European Commission and which has overall accounting responsibility for the disbursement of the Funds. Thereafter, the finance ministry typically distributes resources to the national government ministries responsible for the individual Funds. Typically, Fund management is exercised by ministries of economics/industry for ERDF and ministries of employment/labour/social affairs for ESF (and ministries of agriculture for EAGGF); these government departments are responsible for co-ordinating the planning and implementation of Fund-specific actions within programmes, for the allocation and payment of funding in line with programme management and project allocation systems, and for ensuring adequate monitoring and evaluation. Under some federal systems (notably Germany and Belgium but not Austria), the transfers from the national ministry of finance go directly to the finance departments/treasuries at regional level, from which the regional economics or employment ministries responsible for the funds then draw down funding as required.

There is considerable regional variation within countries. Italy is a good example: the extent of partnership and the balance of responsibility between central and regional authorities varies significantly between regional programmes, and this regional divergence appears to be growing. In France, regionally targeted Structural Fund programmes are always managed by regional offices of the State, but the degree of influence of elected regional and local authorities varies greatly, as does the amount of responsibility devolved to the sub-regional offices of the State in the départements. This is also true of the UK (in terms of the different approaches to implementation in England, Scotland and Wales) and Belgium (Flanders and Wallonia). These regional differences reflect the disparities between regions in terms of regional political leadership and autonomy, institutional capability, network relations between regional/local actors, the character of the policies being pursued and the significance of Structural Fund programming relative to other policy activities.

In summary, there is little commonality in the institutional arrangements for managing the Funds, with a variety of different approaches to the regional management of Objective 2 programmes across Member States. In some regions, programme management is undertaken by the local arms of State organisations (France, Sweden, England), while in others it is partly or wholly the responsibility of regional government organisations of varying constitutional status (Belgium, Germany, Austria, Spain, Italy and Finland). In a limited number of regions in selected Member States, the day-to-day management and administration of programmes has been delegated to independent executives, with varying degrees of autonomy from the official ‘programming authority’ (Scotland, Wales, Netherlands, Sweden, Flanders).

25

2.5 Management Trends Apart from the commonalities and contrasts between countries, an important question is how institutional arrangements for managing the Structural Funds have developed over time, in particular how national institutions and regional policies have adapted to the challenge of managing Objective 2 programmes.

In the early stages of programming, the initial programmes were disjointed in management and administration. Most countries initially sought to limit the impact of EU regional policy on their systems of governance, while maximising their control over the Funds. In almost all the Objective 2 Member States national regional policy officials were suspicious and critical of the policy objectives and implementation requirements of the Structural Funds.25 In part this reflected the administrative challenge of implementing EU regional policy and the difficulty of adapting to the distinctive management culture of the European Commission26 The institutional management (for the first time) of multi-annual programming, integrated operations - especially between ERDF and ESF - and partnership was highly demanding. Also, there was a lack of acceptance of the Structural Funds as an independent area of policy and a resistance to changing established administrative structures and systems.

Initially, programme management was essentially department or agency based, each administering department operating programme measures through established structures and with little cross/inter-departmental liaison or communication. Early approaches to programme delivery were opportunistic, reflecting the approach to programme management. The perception among partners of the programme as a vehicle for drawing down funding was shared among beneficiaries who saw the programme in terms of ‘grant opportunities’; indeed, partners were often the same as beneficiaries. The need to commit resources rapidly, following a delayed start to the CSFs/OPs led to a widespread tendency to take projects ‘off the shelf’ or select projects on a ‘first come first served’ basis.27 Evaluation research found that project planning could be weak, sometimes derived from simple abstract ideas or borrowed - largely unadapted - from other regions or agencies; prior research and piloting was unusual and consultation and co-operation between agencies limited.28

25See, for example, Bachtler J and Turok (1997) An Agenda for Reform, in J Bachtler and I Turok (Eds.) op. cit. or the policymaker contributions in the same volume: Wells R (1997) UK Policy Perspectives; Lagrange R (1997) French Policy Perspectives; Drerup D (1997) German Policy Perspectives; Kleyn W H and Bekker M (1997) Integrated Regional Development under EU Regional Policy: The Dutch Experience. Similar sentiments are expressed in Huber W (1996) Regionalpolitik in Ősterreich und Kohäsionspolitik der EU, Discussion Paper, Federal Chancellery, Vienna; and Sapper A (1997) Regionalpolitik in Ősterreich vor dem Hintergrund der Strukturfonds und Regionalpolitik der Europäischen Union, in I Kretschmer (Ed.) Geographischer Jahresbericht aus Ősterreich. EU Sonderband, Institut fűr Geographie der Universität Wien. 26Eskelinen H et al (1997) Appraisal of the Finnish Objective 2 Programme: Reflections on the EU Approach to Regional Policy, Regional Studies, 31(2), 167-172. Huber W (1996) Regionalpolitik in Österreich und Köhasionspolitik der EU, Diskussionspapier, Federal Chancellery, Austria, mimeo. 27 CEC (1997a) The Structural Funds in 1996: 8th Annual Report, Commission of the European Communities, OOPEC, Luxembourg. 28 Turok I (1997) Evaluating European support for business development: evidence from the Structural Funds in Scotland, Entrepreneurship & Regional Development, 9, 335-352.

26

Over time, the centralism in Structural Fund management has tended to lessen. Particularly in Member States with longer experience of EU regional development programmes, institutional structures have been progressively adapted over the past 2-3 programming periods, in most cases evolving more integrated and inclusive arrangements and decentralising responsibility to regional levels. In a range of countries, the Structural Funds have promoted (or at least influenced) either the devolution of power to elected regional authorities (Denmark, the Netherlands, Italy, Spain) and/or its deconcentration to offices of the State and its ministries in the regions (France, Finland, Sweden, UK).

The boundaries between government ministries are also being broken down in the interests of effective implementation. In the UK, for instance, integrated government offices were established in the English regions in the mid-1990s, in part to facilitate easier inter-departmental coordination for managing the Structural Funds. A similar process of unifying the district offices of national government ministries within regional ‘employment and economic development centres’ has been carried out in Finland (T&E Centres). In Germany and Austria, where economics and labour market ministries continue to operate independently, even at Land level, the locus for institutional co-operation is sub-regional - ‘regional conference’ arrangements in Germany, regional management offices in Austria.

PROMOTING AWARENESS OF STRUCTURAL FUND PROGRAMMES REGIONAL MANAGEMENT OFFICES IN AUSTRIA

A major challenge for all Structural Fund programme managers is how to promote awareness of programmes. In Austria, a distinctive approach to programme implementation has been established through the formation of Regional Management Offices, which promote local involvement in Structural Fund programming and assist in project development. While regional managers or consultants had operated in conjunction with national regional policy in some areas prior to EU accession, the comprehensive coverage introduced in line with the Structural Funds was a new development.

Regional Managers do not have a formal role in the operational implementation of schemes or in political or interest group activity. The support they bring to Structural Fund programmes varies, but typical functions include:

• = transfer of information into the region, including innovative project ideas, information on financial support and specific consultancies;

• = lobbying and information dissemination about the region,

• = assisting cooperation and coordination within the region, including tasks such as improving contacts and information flows between regional actors, stimulating joint participation in projects, consensus building and bringing cooperation partners together; and

• = providing initial consultancy support, ie. acting as a first point of contact for regional firms and helping them identify possible partners, funding sources etc.

The regional managers are not designed to act as competitors for other regional development organisations but instead help to stimulate cooperation between them.

27

INVOLVING LOCAL ACTORS IN STRUCTURAL FUND PROGRAMMING REGIONAL CONFERENCES IN NORDRHEIN-WESTFALEN

As NRW is the largest region in Europe, its government authorities and the social partners at the state level are too distant from the specific problems of the sub-regions. To counter this, NRW has developed a specific management approach to facilitate the targeted involvement of local authorities, chambers of commerce, chambers of craft industries, the unions, the universities, specialised authorities such as the labour offices and other important players at the sub-regional level.

The basis of this ‘regionalised structural policy’ are ‘regional conferences (or committees)’ which meet in each of the 15 sub-regions. These conferences discuss sub-regional strengths and weaknesses and then use this to determine regional development priorities. The principles governing the conferences are co-operation, co-ordination and consensus. Special parallel regional labour market conferences operate for labour market policy.

The development plans and project proposals of the conferences form the basis upon which structural policy at the state level is built. In addition, those projects put forward by the regional conferences receive the highest priority for funding. As a result, the Objective 2 programme primarily funds projects developed by the social partners and local or sub-regional authorities. The role of the Land government in the conferences is to ensure high standards and conformity with state and European policy guidelines.

Implementation systems have matured significantly. During the 1994-96 programming periods, programme administration moved from a ‘list of projects’ approach to a more systematic and co-ordinated method of allocating resources. The programming concept became more firmly embedded: the strategic management of multi-annual programmes began to be accepted as “more conducive to efficient and effective planning” than responding ad hoc to annual ‘wish lists’.29 Monitoring Committees were playing a more significant role in many regions, and mechanisms such as secretariats, advisory groups and local delivery mechanisms were being introduced in an attempt to promote synergies, raise quality and create integrated management structures.

29 House of Lords (1997) Reducing Disparities within the European Union: the Effectiveness of the Structural and Cohesion Funds, Select Committee on the European Communities, HL Paper 64, The Stationery Office, London.

28

INDEPENDENT PROGRAMME SECRETARIATS - UK AND SWEDEN

In Objective 2 regions in Scotland (UK), programme implementation and management are undertaken by independent management bodies supporting the Member State authority and the wider partnership. Termed the ‘oil in the machine’, Scottish ‘programme executives’ such as Strathclyde European Partnership undertake virtually all aspects of programme administration apart from payments: servicing partnership committees and implementing committee decisions; managing programme publicity and project application processes for ERDF and ESF, including the provision of application advice; monitoring financial and physical progress; and general liaison with implementing authorities and the European Commission. In some cases, these executives have moved or are moving towards private sector status.

Among the perceived advantages of the secretariats is their independence from narrow institutional dependencies, which means that they can represent the views of and respond to a collective partnership. This independence is reinforced by the manner in which the executives are funded: a small charge is levied on successful project applicants, the amount varying between programmes, and being proportionally lower for the largest programmes, where economies of scale can be achieved. The fact that the secretariats are funded directly by their partnership also helps to ensure lean and efficient programme delivery.

Most Swedish Objective 2 areas administer the Structural Funds through the existing County Administration Boards (CABs), but Fyrstad, following the Scottish example, established an independent Secretariat to administer its programme, in this case because the eligible area spanned two CAB areas which had little track-record of joint working. The autonomy which the Secretariat in Fyrstad has from the local state apparatus allows it to work more freely when dealing with other regional actors. It may, therefore, make the Funds more acceptable to other local political bodies in the area compared with Sweden’s other Objective 2 regions.

INCORPORATING THE POLITICAL DIMENSION: THE REGIONAL STEERING COMMITTEE IN NORDJYLLAND

In Nordjylland, the ‘classic’ Objective 2 programme delivery structures (the Monitoring Committee which steers the programme, the Executive Committee which takes decisions, and the secretariat, which is housed in the Industrial Policy Division of Nordjylland County Administration and services the programme) are supplemented by a further structure, the Regional Steering Committee (RSC). This Committee, which comprises mayors and other senior political figures is chaired by the leader of the county council and provides a permanent forum for dialogue between regional actors.

The central role of the RSC is to help to ensure political backing at the regional level for the overall priorities of the Objective 2 programme, for the new initiatives and programmes it contains and for major individual projects. In addition, the Committee serves to improve co-ordination between Objective 2 activities and the region’s own development policies. At an operational level, the RSC brings a democratic dimension to shaping the general principles by which project decisions will be made, while the ‘technical’ committees apply these frameworks in making project decisions.

29

However, efficient management and delivery of programmes was still often hampered by the existence of too many small programme elements, limiting effectiveness and causing enormous administrative demands. The measure or sub-programme remained a rather exclusive unit with barriers to information and consultation (often related to marginality of the Funds in relation to national/regional/local policy expenditure). In the context of a government review in the UK, it was noted that the ERDF process was “characterised by short lead-in times for applications to be submitted, long delays in processing and minimal time allowed for the execution of the proposed projects”.30 The review noted that only one-quarter of the 4,500 applications submitted and appraised under the 1994-96 Objective 2 programmes were actually approved, associated with long decision-making times, slow payment procedures, inadequate ERDF/ESF integration and lack of a strategic approach.31

As noted in more detail in later chapters, the methodology for programme preparation in 1996-97 seems to have been more ‘bottom up’ than ever before in many regions, more integrated and better informed. This in turn led to greater coherence within programming documents. For example, while ERDF/ESF coordination is still patchy - much of ESF spending being project-based rather than programme driven - greater efforts were made to ‘package’ ERDF and ESF resources together, and to seek a closer articulation between demand and supply in labour market interventions. In addition, many aspects of implementation arrangements were reviewed in the course of reprogramming, leading to numerous improvements. New administrative structures were introduced in several cases with a better division of responsibilities (including the introduction of a new Executive in Wales). Mechanisms to encourage the greater involvement of local partners in programming became more widespread, including specialist advisory groups (for RTD, labour market, environment, equal opportunities issues etc), more effective and imaginative means of communication and dissemination of information, and the wider involvement of elected members. In some cases, new systems were introduced to afford specific local and/or thematic partnerships enhanced responsibility in aspects of programme delivery, the primary example being UK proposals for ‘Action Plans.32

30 Quote from a joint paper by local authority associations, reported in Jacobs J (1997) Streamlining the Structural Funds – Government Reviews Administration in England, in European Information Service, Issue 177 – February. 31 HMSO (1996) ‘Lifting the Burden - Efficiency Scrutiny Report on Information Flows between Central and Local Government, 14 November 1996. 32 DETR (1997) EC Structural Fund Programmes in England: Action Plans for ERDF and ESF, Guidance Note No 1, Department of the Environment Transport and the Regions, London.

30

Whereas programming developments during the 1994-96 period were primarily associated with embedding the regional-level structures for organising plan preparation, strategy development, programme management and partner co-ordination, the 1997-99 period was characterised by more widespread investment in management and monitoring systems. Because of the need to commit resources quickly at the outset of the earlier programme periods, it was not until the mid-1990s that programming authorities began to introduce rigorous and strategic formal project appraisal and selection procedures. It was also recognised that programme monitoring and evaluation practices were very weak. Across numerous Objective 2 programmes - in Belgium, Netherlands, Luxembourg, Austria, Sweden, the UK and Denmark – interim evaluations conducted in 1996-97 emphasised the priority being given to the development of more systematic, robust and comprehensive monitoring indicators, to track both physical activity and impacts. In Finland and the UK, investment in management information systems and procedures was highlighted to accelerate decision-making, to speed up the flow of funds from award bodies to beneficiaries, and to establish systematic routines for information processing and storage.

2.6 Programme Delivery Structural Fund programmes are ultimately delivered through the generation, appraisal, selection and then implementation of project proposals. Progressively, programme delivery has become more pro-active, attempting to take a more integrated, developmental approach to the generation and selection of projects. The first stage in laying the groundwork for successful delivery is the programme development phase, during which a diagnostic process identifies and makes provision for interventions which the full partnership believes are best able to positively impact on the regional

GREATER RESPONSIBILITY TO PARTNERS FOR PROGRAMME DESIGN AND DELIVERY - UK ACTION PLANS

In England, to enhance 1997-99 programme implementation, Monitoring Committees (MC) were encouraged to invite sub-groups of partners, defined geographically and thematically, to put forward integrated ‘Action Plans’ (AP) for the implementation of parts of their SPD. Once an AP and its associated budget was approved by the SPD decision-making committee and the MC, the meso-level partnerships involved would take responsibility for implementing it, appraising and approving projects submitted under it, and co-financing them using their designated envelope of Funds. The main anticipated benefits of the APs were enhanced co-ordination between the ERDF and ESF and between domestic and European economic development policy, stronger partnership at the local level, streamlined programme delivery and a more strategic approach to implementation.

A variant on the AP has been the ‘Package’ approach, in which meso-level partnerships – especially in community economic development areas – submit an integrated plan setting out what they wish to achieve under the SPD across the programming period, in the form of a ‘Package’ of outline projects and a rationale justifying them. Once approved, this framework is used as a guide for project decision-making by pre-existing programme level decision-making structures. The mechanism helps to provide a more strategic and integrated approach to SPD implementation at the level of the locality, yet involves none of the inefficiencies involved with delegating aspects of project decision-making, financial management and programme monitoring to the sub-partnerships.

31

economy. In addition to setting out policy scope, there is increasing awareness of the role programming documents play in shaping the detail of project applications, including by explicitly emphasising and clarifying expectations about key programming objectives such as job creation.33 Subsequent to programme approval, pre-application advice is being provided by increasing numbers of programmes to improve the relevance and technical quality of project applications.

Two types of decision are taken on all projects submitted to Structural Fund programmes for financial support – the first to approve the domestic portion of co-financing and the second for the Structural Fund contribution. Project decision-making as it relates to the indigenous portion of project co-financing, whether it be of national, regional, local or private origin, remains the responsibility of the relevant domestic co-financing authority. Diverse systems have emerged for decision-making on the Structural Fund element of project finance. In some Member States, including Germany and Austria, project decision-making has largely been subsumed within pre-existing institutional structures. As noted above, in Germany, the implementation of the Structural Funds has been integrated, as far as possible, within the existing system and structure of regional policy administration, programme management and project decisions being undertaken by the relevant Land authorities - ministries of economics for ERDF and ministries of labour or social affairs for ESF.

In other Member States, while decision-making for some project areas is subsumed into what are effectively the pre-existing circuits (eg. through the Business Aid Working Groups in Finland), distinct mechanisms have been established for decision-making on many Structural Fund projects. For example, in Finland, Sweden, France and the UK, it is the task of special interdisciplinary Programme Management Committees (PMCs) in the regions (usually composed of key co-financing partners) to make most decisions on which projects should receive Structural Fund support. These committees are assisted by prior project appraisals, generally undertaken by programme administrators and/or an expert agency/expert agencies in relevant fields. There is again variety in how the PMCs are configured, even within Member States. In France, some programmes operate a single region-wide committee, and others, several committees for specified geographic areas.

Although in some less-developed regions the project selection stage may still be limited to a formal analysis of eligibility criteria for expenditure, in the best cases it is moving towards an holistic, strategic approach along the lines of the “virtuous circle of programme delivery” (see figure below). Faced with increasing numbers of applications and, in some cases, decreasing resources, the processing of applications has become more professional and systematic. In particular, formal and systematic appraisal and selection procedures have been introduced by many programmes, with some developing sophisticated methods for scoring, weighting and ranking projects (especially in the UK and

33 Chapter 4 addresses the programme development phase in more detail, while Chapters 5, 6, 7 and 8 illustrate how this phase can be managed to improve responses to the horizontal programming priorities. Chapter 9 shows how it can be used to help lay the foundations for effective programme monitoring and evaluation.

32

the Netherlands), or involving independent experts and/or specialist panels in the selection process. External experts are also used to address problems experienced by regions with ensuring take-up under specific programme measures eg. increasing SME participation in Finnish programmes.

Figure 2.1: The virtuous circle of programme delivery

Programme marketing

Programme development/

refinement

Evaluation

Monitoring

Project appraisal& selection

Project implementation

In the UK, most programmes use scoring systems to compare the merit of project applications. These systems emerged due to situations of overbid for European funds which necessitated a robust, transparent, consistent and replicable way of judging projects. Scoring systems, which build on the selection criteria set out in programming documents to guide applicants, have increased in maturity over successive programming rounds. In particular, they have become less open to manipulation by partners shaping proposals to meet scoring thresholds. In Scotland, for example, there is no points threshold at which a project is approved: projects with a low score can be approved if the expert opinion of Advisory Groups is that they are worthwhile, while projects which receive a high score yet are of limited merit can be rejected. Likewise, in England, selection systems also increasingly supplement scores with a qualitative review. Here, the initial weakness of mechanical scoring systems was political: they enabled partners to avoid criticising each others’ projects directly (and so harming the chances of their own projects being approved).

Once projects have been approved under a given programme, their implementation is monitored to ensure that they are undertaken as planned, and to capture initial information about their outputs and impacts. Effective computerised management information systems have been introduced which assist in tracking projects from initial proposal submission to completion (Scotland, England and Wales, the Netherlands, Bremen, Aquitaine), although such systems have on occasion caused problems and hindered programme

33

implementation (Finland). A significant contribution of these systems has been to enable more comprehensive monitoring data to be collated more frequently, providing up-to-date management information which can then be used to improve programme delivery, for example by identifying where absorption rates are insufficient.

Programme delivery has both improved and developed over the last decade. The main deficiencies continue to be those which are endemic to the system - unrealistic timescales for programme and project planning and implementation; excessive bureaucracy; problems in determining the eligibility of projects through imprecise guidelines leading to delays in approval decisions and fund payment; and the difficulties project implementers have experienced in responding to the project design and management requirements of Structural Fund programmes (including integrating the horizontal issues and ensuring robust and appropriate monitoring).

‘ADVISORY GROUPS’ SUPPORTING PROJECT APPRAISAL

Specialist Advisory Groups are used to varying degrees in the EU (Finland/ Netherlands/UK/Denmark) to provide support at various stages of the project generation and appraisal process. In some countries, these groups assist decision-making by providing technical advice, while in others their role is more substantive.

The three advisory groups in Päijät-Häme (Finland) are involved in tourism, technology and training projects at an early stage, running group meetings to actively assist applicants with the project development process. They undertake an initial and unofficial appraisal of possible project applications, and have an active role in improving them - co-ordinating the resources which could be drawn upon, bringing proposers of similar projects together, and improving the case for applications.

Fyrstad is alone among the Swedish Objective 2 areas in establishing Advisory Groups. Composed of local civil servants and officials from local municipalities, these groups aid project applications prior to them reaching the more politically charged atmosphere of the PMC. They can also act as a ‘filtering mechanism’, notifying programme managers of potential political issues arising from certain projects. Similarly, the use of more general reference groups (composed of banking interests, trades unions and cooperative organisations) is different across Objective 2 regions in Sweden.

In Scotland, Advisory Groups are not involved in project development, but instead support the project appraisal process. These Groups, piloted by the Western Scotland programme, bring specialist and impartial input to the appraisal of projects and so improve the quality of decisions which can be made by generalist Programme Management Committees. In the Western Scotland programme, there are Advisory Groups for each programme priority (eg. Technology and Innovation, Community Economic Development, Business Development) plus a Labour Market Advisory Group. To encourage critical appraisal, these groups consist of both programme partners with specific expertise and impartial experts, including academics. Usually, projects are appraised by Group members prior to their meeting, and then presented to the Group with a recommendation. This is discussed and, depending on the comments of the other panel members, a final recommendation is made. Where there are difficulties with a project, or more clarification is needed, then the applicant can be invited to give a brief presentation and answer questions. To ensure fairness, applicant organisations which are represented on an Advisory Group are excluded from discussions about their own proposals.

34

2.7 Conclusions and Recommendations Every Member State has a different approach to programme implementation, reflecting different structures of territorial administration, the division of powers and policy responsibilities between national and sub-national authorities, and the relationship between Structural Fund and national regional policies. Furthermore, there is considerable regional variation within countries, which appears to be growing.

It is a feature of virtually all EU Member States that established government authorities - national governments in unitary states, regional authorities in federal countries and some others - play a dominant role in Structural Fund programme management. Indeed, most Member States have sought to limit the impact of EU regional policy on their systems of governance. Whereas some countries consider EU regional policy to be complementary to their own national policies, others have sought to restrict the scope of the Funds. In certain cases, EU regional expenditure has been channelled rigidly through the existing funding circuits of regional policy, while elsewhere the Structural Funds and national regional policy have been kept completely separate. Over time, there has been some convergence between EU and Member State regional policies, although more in terms of procedural adaptations, such as changes in eligibility conditions for financial incentive schemes, than direct effects on national administrative structures and systems.

One of the main causes of problems for Member States arises from different actors (at national, regional and local levels) working at different stages of the programming process, often with conflicting goals and agendas, and with differing relationships with Commission services. In this context, the level of sub-national involvement in programme management is a hotly contested issue. Lower levels of government have often perceived the Structural Funds as an opportunity to gain more autonomy with regard to their economic development responsibilities and resources (especially in unitary states), and this has increased the awareness and involvement of regional and local authorities with respect to strategic planning and development.

Over time, the Structural Funds have contributed to a change in attitude and structures. In some regions, participation in Structural Fund programmes has led to deconcentrated services of the State, formerly accustomed to operating independently in accordance with their own discrete remits, becoming more conscious of how their actions fit into an overall strategic context. This applies to the Government Offices in the UK, Regional Prefects in France, CABs in Sweden, and to the District Offices in Finland. Rationalisation of central government administration with greater integration of (or co-operation between) the regional offices of sectoral ministries is also apparent, as in England and Finland. Elsewhere, implementation of Fund administration has strengthened the devolution of economic development, for instance to the Provinces in the Netherlands, the Regional Councils in Finland, and the Regions in Italy. A similar process is evident at sub-regional level in Germany with the creation of 12 regional committees or ‘conferences’ in Nordrhein-Westfalen. Novel institutional arrangements have been established for collective participation in strategy-making, implementation and monitoring. Of

35

particular note is the use of independent ‘programme management executives’ - pioneered in the Western Scotland Objective 2 areas but now being used in other Objective 2 regions in Scotland, Wales and Sweden (and currently under consideration in other Member States). In Denmark, political interests have been actively incorporated into Fund management in the Nordjylland Objective 2 regions by creating a Regional Steering Committee to ensure political backing for development issues and to co-ordinate Objective 2 with other economic development initiatives. In Austria, a network of Regional Management Offices supports Structural Fund implementation at the local level, encouraging information dissemination, project generation and co-operation, especially among municipalities.

It is equally clear that there is still a considerable gap between the advocacy of regionalisation and partnership, outlined in this chapter, and the reality of regional development. The management of programmes continues to be dominated by the compartmentalisation of priorities and measures; substantial shares of programme expenditure are still allocated to government departments or agencies over which programme managers have limited influence and no control. This is not exclusively a problem of central government institutions; considerable funds are channelled through existing schemes of the sectoral departments of the Objective 2 Länder in Germany and Austria and the Assessorate of the Italian Regions with questionable strategic orientation and policy additionality.

With respect to the process of administration, in the early stages of programming, implementation of the Funds was characterised by management deficits at every level. At regional level, it was often anticipated that programme management could be accommodated within government departments or agencies, many of which had no experience or frame of reference for designing regional strategies or managing multiannual programmes. At Member State level, many national regional policy officials were often highly suspicious and critical of the policy objectives and implementation requirements of the Structural Funds and, in some cases, resisted complying with the principles of the Funds. In retrospect, officials will freely admit that it took them 4-5 years to appreciate the terminology and procedural requirements of the Funds and introduce appropriate management and monitoring systems. Institutional inflexibility hindered integrated approaches to implementation. At Commission level, the management and resource capabilities of Commission services were inadequate to deal with the administrative complexity of the Funds, resulting in the familiar litany of complaints – endemic delays, complicated decision-making processes, bureaucracy, lack of co-operation between Directorates-General, and inconsistent advice.

36

2.7.1 Programme Management Recommendations

More effort needs to be made to promote attitudinal change within Member State authorities, in particular to encourage flexibility and new thinking in strategic approaches to industrial change.

The management of the Funds has improved greatly – at all levels - over the past decade, but problems still remain. There is still unwillingness in some Member States to accept that it requires a different approach from long-established national practice. This translates into resistance to adopt a positive, pro-active approach to implementing the Funds, and the attempted sidelining of concepts like partnership. Also, institutional inflexibility and compartmentalisation persists at Commission and national levels, most obviously evident in the difficulties of integrating ERDF and ESF in practice. Notwithstanding the existence of the regulations, changes in attitude will only come about if it is possible to demonstrate that a strategic approach to industrial change, through real partnerships of economic and social actors, really can make a difference to the efficiency and effectiveness of regional development. This means promoting the dissemination of good practice and providing pro-active support for programme managers. The European Commission has a critical role in disseminating good practice in economic development and the innovative management of regional development, and it should develop this more vigorously.

The key to successful programme management is to understand the inter-relationship between the different elements of the programming cycle and to manage them as an integrated ‘virtuous circle’.

Implementation systems have matured significantly since 1989. Progressively programme delivery has become more pro-active, attempting to take a more integrated, developmental approach in the generation and selection of projects. During the 1994-96 programming period, programme administration moved from a ‘list of projects’ approach to a more systematic and co-ordinated method of allocating resources. The programming concept became more firmly embedded: the strategic management of multi-annual programmes began to be accepted as more conducive to efficient and effective planning.

In the best cases, programme delivery is moving towards an holistic, strategic approach along the lines of the ‘virtuous circle of programme delivery’. This requires programme managers and partners to see the programme cycle as a continuum. It begins with the development of the programme, itself built on input from previous programme experience. Programme marketing involves the active generation of projects (interacting with potential applicants to bring forward projects meeting strategic objectives and priorities) and providing pre-application advice to improve the relevance and technical quality of project applications even before submission. Then, professional and systematic approaches to project selection require formal and systematic appraisal and

37

selection procedures, exploiting the increasingly sophisticated methods for scoring, weighting and ranking projects, or involving independent experts and/or specialist panels in the selection process. Thereafter it is important to see project award as the start of a process of close and continuous monitoring and ‘project aftercare’, supported by effective computerised management information systems able to track projects from initial proposal submission to completion. The experience generated in the stages of the programming cycle in turn contributes to refining the strategy and programming practice.

In order to improve programme management, information flows, especially about required programming practice, need to be improved at all levels and made more consistent.

Uneven and insufficient information flows at all levels about required programme management practices (eg. relating to project eligibility, financial management, monitoring and evaluation obligations, etc) translate through into inconsistent and sometimes insufficient practices on the ground. Such variation can even occur within a single programme, for example where a decentralised or regionalised system of delivery encourages inconsistent approaches to programming. A fundamental problem is how to ensure standard and sufficient management of the Funds by all Structural Fund actors in a Member State, including government departments at central, regional and local levels, in a manner that complies with the Structural Fund regulations and national administrative and accounting practices.

Research for this report shows that information flows are still inadequate in many cases, with considerable variation in information and awareness of policy and programming developments, and should be improved.

Successful Structural Fund programming is increasingly demanding in terms of resources and expertise. Drawing on existing practice, a more explicit debate is required at all levels about how programme development and delivery should be organised and resourced and how those involved should be trained to meet the specific challenges of this changing environment.

Many of the changes to Structural Fund management systems over the last decade have been driven by the growing complexity of programme delivery, which demands significant resources and expertise, not only from partnership members, but also in terms of secretariat services. Structural Fund programmes have to be conceived as strategic plans in their own right, developed and delivered through the active participation of wide partnerships, supported by sophisticated information management systems and with dedicated, inter-related responses to each stage of the programming cycle (programme development, project generation, project selection, monitoring and evaluation). They not only have to maximise economic development impacts, but must also take appropriate account of horizontal priorities such as

38

sustainable development and gender mainstreaming, which may not yet be fully integrated into the domestic economic development context. In addition, programme design and delivery mechanisms cannot be static, but instead have to adapt on a rolling basis through processes of ‘bureaucratic innovation’ to changing theories of economic development, evolving regulatory requirements and the lessons of experience at the programme level.

Too little is known about the full cost of delivering Structural Fund programmes and the extent and nature of responses to their financial and human resource implications. Evaluation evidence, however, suggests that programmes are frequently prevented from achieving their full potential by expertise and resource limitations which inhibit sufficient and creative responses to programming challenges. In addition, it is rare that any form of structured training is available to prepare those working in the Structural Fund field. The current reform of Structural Fund systems provides an opportunity to more explicitly review how programming is organised, prepared and resourced. Implicit in this is a debate about the optimal balance between the costs and benefits of Structural Fund programming.

Better use needs to be made of the partnership resource. This means not just involving more partners but using them more effectively at appropriate stages of the programming cycle.

The input of the partnership to programme design, delivery and implementation is central. Improved approaches are needed to partnership management, learning from the ‘good practice’ that exists, to involve more partners more actively at more stages in the programming process where their input could improve the quality of programming. A key issue in assessing partnerships is the extent of ‘strategic capacity’ - whether they make decisions or rubber-stamp them. Programme authorities and partnerships should work together to identify which partners, or groups of partners, can most effectively contribute which functions at what stage of the programming process, and ensure that they are properly prepared to undertake these functions. A complementary aspect of this is to ensure that excessive demands are not made on partners’ time.

A key question is how to establish and maintain coherent and effective partnerships, working strategically to shared aims, focusing on strategy rather than projects. To avoid a programme’s ‘committee community’ effectively becoming an ‘exclusive club’, which could alienate important groups such as the social partners, proceedings should be open and widely publicised. Lack of interest in participation among relevant actors and the wider community need measures to promote understanding of the programme and the Structural Funds and the opportunity to shape the economic development future of the region.

39

3. THE IMPLEMENTATION OF OBJECTIVE 2 PROGRAMMES: COUNTRY-BY-COUNTRY REVIEW

The previous chapter provided an overview of trends in the institutional management of regional policy among the EU Member States and the relationship between the administrative arrangements for implementing national regional policy and the Structural Funds. It also reviewed the evolution of programming arrangements and project delivery over time. The following chapter provides more in-depth information on the implementation of Objective 2 programmes in different national contexts. It examines the specific institutional arrangements for managing the Structural Funds in more detail, focusing on each of the 12 Member States with Objective 2 regions in turn.

3.1 Austria Despite being a federal country, political power in Austria is concentrated at central government level under a system which encourages the Federal Government to act as the legislative branch with the lower tiers (partly) consigned to a more administrative function. Competence for regional policy is not allocated constitutionally either to the Federal Government or to the States (Länder), although in practice there is of legislative and administrative responsibilities for regional development between Federal, state and local levels. The system is distinctive in its complexity and the importance of informal links and contact networks. At Federal level, a range of government ministries has regional development functions and instruments. The system is co-ordinated partly by the Federal Chancellery (Bundeskanzleramt) and partly by the Austrian Conference of Regional Planning (Österreichische Raumordnungskonferenz ÖROK). At regional level, each of the nine Austrian states elaborates its own regional development programme which, following approval by the Federal Government, is implemented autonomously by the districts. The Federal level often has a more formal role under so-called federal-state agreements for co-funded packages of regional development measures.

The Structural Funds in Austria have been largely integrated within the existing support system for administering economic development. Primarily to satisfy competing political and administrative interests, and an unwillingness to establish a new administrative structure for handling the Structural Funds, EU regional policy support has been distributed across numerous government departments at national and regional levels. This has resulted in a highly fragmented structure for implementing the Structural Funds with a plethora of EU-funded incentive schemes; according to Commission figures, 34 Federal and 134 state schemes have been co-financed under Objectives 1, 2 and 5b during the 1995-99 programming period.34

34 Downes R and Bachtler J (1998) Komplexitätsreduktion - regionale Wirtschaftsförderung in Österreich, Raum - Österreichische Zeitschrift für Raumplanung und Regionalpolitik, 30, pp.39-41. Downes R and Bachtler J (1998) Business aid and the Structural Funds in Austria, Regional and Industrial Policy Research Papers, EPRC, University of Strathclyde, Glasgow.

40

The Federal Chancellery has overall responsibility for the Structural Funds, including negotiations with the Commission. Austria-wide coordination among federal departments, and between federal, Land and local levels, is the task of ÖROK. With respect to financial management, all EU monies are initially transferred to the Federal Ministry of Finance from where they are drawn down by so-called fund-corresponding (fondskorrespondierende) Ministries. At national level, the overall responsibility for co-ordination of ERDF lies with the Federal Chancellery; monies for the co-financing of Federal or Land measures are distributed according to a predetermined quota to Federal ministries (economics, R&D, transport, environment etc), federal agencies and Land government departments and agencies for economic promotion. Overall responsibility for ESF lies with the Federal Ministry of Social Affairs and Health (BMAGS) with implementation undertaken principally by the Labour Market Service, an independent institution responsible for administering employment and training policies which operates through a network of regional and local offices. Each Land also has responsibilities for implementing ESF actions at regional level under Objective 2.

Programme management is the responsibility of the Länder, with a coordination role being exercised by the Federal Government. Formal implementation is managed by a single Monitoring Committee for all four Objective 2 regions (Steiermark, Niederösterreich, Oberösterreich, Vorarlberg), chaired by the Federal Chancellery. Individual project award decisions are taken by the funding agencies (federal ministries, Land government departments, regional development agencies, special banks) responsible for the schemes or programmes being co-financed by the Funds. Where more than one agency is involved in an award, the decision is made on the basis of informal contacts rather than formal committee structures. Indeed, there are no central project decision-making committees; the award criteria, award systems and award rates used are those already operating for each incentive scheme. This approach also applies to the delivery of projects: both project approval and implementation are co-ordinated through networks of informal relationships between managers of aid schemes at regional and Federal levels.35

Since Structural Fund expenditure is channelled through existing funding circuits, with predetermined allocations for different government departments and agencies, there is little scope for active programme management in terms of strategic steering of the programme. At Land level, the emphasis of programme management is more on programme co-ordination, a task undertaken by departments of economic development (Steiermark, Vorarlberg) or EU policy departments/offices (Niederösterreich, Oberösterreich).36 This administrative workload is considerable: in addition to overseeing the administration of the Objective 2 programmes, these offices are

35 CEC (1996) Communication of the Commission on the Implementation of EU Regional Policies in Austria, Finland and Sweden, COM(96) 316 final, 3.7.96, Brussels. 36 InTeReg, BAW and EPRC (1998) Zwischenbewertung der Interventionen der Ziel-2 Programme, des RESIDER II und des RECHAR II Programmes in der Programmperiode 1995-99 der Strukturfonds EFRE und ESF der Europäischen Kommission in Österreich, Report to ÖROK, Joanneum Research, Graz.

41

generally also responsible for the Objective 5b programmes, Community Initiatives and the administration of other (non-EU) regional development interventions. Repeating the point made earlier, with such a large number of aid schemes and administering agencies, many with small budgets, the system is administratively complex, bureaucratic and opaque - even to Austrian policymakers.37 In each Land, different Programme Groups and Working Groups have been established to assist both vertical and horizontal coordination at Land level.38 However, as noted in the Objective 2 interim evaluation, “with the high level of institutional fragmentation of programmes in Steiermark and Niederösterreich, it appears to be extremely difficult to maintain an overview of project implementation as a result of the multiplicity of organisations involved”.39

At sub-regional level, EU accession and the implementation of the Structural Funds has been associated with the emergence of a dense network of local and regional initiatives and advisory networks, accelerating the trend of the past decade towards regionalising/localising economic development.40 A distinctive feature of the Austrian implementation system is the use of Regional Management Offices established in many of the larger regions. The responsibilities of the regional managers vary greatly between regions, but their common aim is to support the local implementation of the Structural Funds through activities such as the dissemination of information and marketing of the programmes, the promotion of project applications and the co-ordination of multi-fund and cross-border projects. Regional ESF consultants also operate in many areas, but again within separate structures and with relatively little co-operation with the activities of their ERDF counterparts. Notwithstanding efforts to improve co-ordination, a common feature is the lack of co-operation between these local and regional initiatives and, indeed, sometimes competition and rivalry.

3.2 Belgium The economic development system in Belgium is the most regionalised in the EU. The Belgian government is responsible for providing the basic legislative framework and has a role in fiscal policy, employment aid and enforcing compliance with State aid regulations, but central authorities are not formally involved in formulating regional policy objectives and strategies. The Regions (Wallonia and Flanders) and Communities each have their own regional development policies, implemented under their own regulatory procedures, with separate and distinctive institutional structures.

37 Schwarz W (1997) EU-Regionalpolitik in Niederösterreich - Konzeption, Umsetzung und Bewertung, in I Kretshmer (Ed.) Geographischer Jahresbericht aus Österreich, EU Sonderband, Institut für Geographie der Universität Wien. Sapper A (1997) Regionalpolitik in Österreich vor dem Hintergrund der Strukturfonds und Regionalpolitik der Europäischen Union, in I Kretschmer (Ed.) Ibid. Huber W (1996) Regionalpolitik in Österreich und Kohäsionspolitik der EU, Discussion Paper, Federal Chancellery, Vienna. 38 Gruber M (1997) Perspectives on regional policy from a new Member State: Austria, in J Bachtler and I Turok (Eds.) op. cit. 39 InTeReg et al (1998) op.cit. p.136. 40 Schwarz (1997) op. cit. Lichenauer-Kranich T (1996) EU-Regionalprogramme fűr Oberösterreich, in I Kretschmer (Ed.) (1997) op. cit.

42

Belgian arrangements for the implementation of Structural Fund programmes reflect the exceptional extent of devolution, that has taken place over the past three decades, within an increasingly federal state as the country has tried to accommodate the co-existence of different language communities.41 Up to the 1970s, decentralisation of economic development was initially organised at the level of the (nine) provinces, a process strengthened with the creation of provincial-level Regional Development Agencies in 1972 to organise regional development (co-ordinating both public and private interests) within a national planning framework. These arrangements were superseded by a major reform in 1980 that assigned wide-ranging powers for economic development to new regional governments for Wallonia and Flanders and to the Communities with further responsibilities progressively devolved during subsequent years.42 The result is central government with very limited economic development powers, strong regional governments, a network of Regional Development Agencies and “informal input from local authorities that can plug into the regional system or into the Regional Development Agencies at the provincial level”.43

Negotiation on EU regional policy matters in the Council of Ministers is carried out by the Regions directly and by the Federal Secretary for European Affairs, concerning Federal matters and representing the views of the three Communities (which also participate directly on occasion). In terms of specific programming, while the Ministry of Economics and the Ministry of Employment and Labour are the official interlocutors of the Commission for ERDF and ESF respectively, in practice the Federal level is minimally involved in Structural Funds programming. The Regions and Communities, which are considered equal, rather than subordinate to, the national authorities, are the sole responsible governmental authorities for decisions on the content of programmes and their implementation.44 Representatives of the Federal Government are involved, but as observers, in the preparation of Objective 2 SPDs and in the Monitoring Committees.45 The fluidity of Belgian arrangements for sub-national governance, however, means that “the regulatory context for the administration of the Structural Funds at times lags behind the division of powers and operational competencies which exist in the state”.46

The two Regions have different approaches to the implementation of the Funds - to the extent that there is little effective co-ordination within Belgium in Structural Fund programming. The Walloon regional government has a more dominant role than its counterpart in Flanders. The overall co-ordination

41 De Rynck S (1996) Europe and Cohesion Policy-Making in the Flemish Region, in L Hooghe (Ed.) Cohesion Policy and European Integration: Building Multi-Level Governance, Clarendon Press, Oxford. 42 The Belgian federation comprises six units – three Regions (Flanders, Wallonia and the Brussels region) and three Communities (the French, Flemish and German-speaking communities) which have been assigned extensive economic development and cultural/social responsibilities respectively. (See: Hooghe L (1995) Belgian Federalism and the EC, in B Jones and M Keating (Eds.) The European Union and the Regions, Clarendon Press, Oxford.) 43 De Rynck (1996) op. cit., p.138. 44 De Rynck S (1994) The Europeanization of Regional Development Policies in the Flemish Region, EUI Working Paper RSC No 94/5, European University Institute, Florence, Italy. 45 Hooghe L (1995) op. cit. 46 Stern E (1997) The ‘partnership principle’ in European Structural Funds, Report to the European Commission (DG XVI), Tavistock Institute, London.

43

role in Wallonia is played by the Directorate-General for Employment and Economics (Direction Générale de l’Economie et de l’Emploi), which liaises with the Commission. A special regional bureau administers the Funds, including the two Walloon Objective 2 programmes (Liège-Meuse-Vesdre and Aubange). In Liège a single Monitoring Committee chaired by the Walloon Government Minister for International Relations deals with overall programming issues (co-ordination, monitoring, evaluation and programme resourcing). Two further ‘technical Monitoring Committees’, for the ERDF and ESF, examine the ensemble of projects emerging under each Fund. The ESF committee is chaired by the Walloon Government Minister for Employment.47

In Flanders, the role of the regional government is more passive and limited, giving local public and private organisations a much greater role in managing the programmes.48 This reflects the local nature of economic development problems in Flanders, and the difficulties experienced by the Flemish government in operating the programme in the early 1990s, subsequently leading to intermediaries (notably the Regional Development Agencies and intercommunal associations) playing a central role in the planning, management and delivery of programmes49. The two Flemish Objective 2 programmes, Turnhout and Limburg, operate the following structures: an SPD Monitoring Committee, in which the Commission and social partners participate; a Programme Committee which makes project approvals; and a full-time secretariat. The Limburg Objective 2 Monitoring Committee is in fact composed of three separate committees - for ERDF, ESF and the ECSC - with a different secretariat serving each. The ERDF committee is served by the Limburg Regional Development Agency, the ESF by a Social Investment Company (SIM) and the ECSC by the non-profit Guidance Centre for the Limburg Mining Area (BLM).50 A Programme Management Unit provides the overall secretariat, while a Permanent Working Group (PWL) is the central decision-making authority. Decision-making is advised by local co-ordination committees which were established at the request of the Commission and report to the PWL.

3.3 Denmark Regional policy in Denmark is managed by the National Agency for the Development of Industry and Trade (NAIT, Ervervsfremme Styrelsen) reporting to the Department of Trade and Industry. Since the abolition of central government regional incentive support in 1991, the regional development functions of the Agency are restricted to Structural Fund programmes and initiatives, as well as the supervision and co-ordination of sub-national development activities in the regions and municipalities. Regional

47 Rider (1994) Évaluation ex-ante du Plan de Développement Régional Objectif No 2 1994-96 du Bassin Industriel Meuse-Vesdre (Liège), Report prepared for DG XVI of the European Commission. 48 Hooghe (1995) op. cit. 49 Roberts P et al (1997) One Union, Fifteen Systems: European Union regional programmes – lessons from practice and a review of future options, Paper to the European Urban & Regional Research Network ‘Regional Frontiers’ Conference, Frankfurt/Oder, September. De Rynck (1996) op. cit. 50 CEC (1995) Intermediate Evaluation of Community Initiatives (Resider, Renaval, Rechar), Volume I Final Report, Prepared by Arup Economics and Planning for DG XVI of the European Commission, March 1995.

44

development is primarily conducted by regionally based actors, either deconcentrated agencies of central government or departments of the administrative regions, but there are restrictions on the types of development initiative that can be taken by sub-national authorities. For example, local governments are permitted to provide collective business services but not subsidies to individual firms.

Institutional arrangements for managing EU regional policy have seen growing regional and local involvement over the past decade. At national level, the Department of Trade and Industry has overall responsibility for co-ordinating the Structural Funds and leading negotiations with Commission services. Compared to many other Member States, the Danish Government has not attempted to use Structural Fund resources (under Objectives 2 and 5b) to co-finance substantial national programmes or schemes, particularly because since 1991 no national-level regional aid schemes have operated in Denmark. EU regional development programmes are essentially regionally determined, the role of national government being to establish strategic principles, supervise management and approve the co-financing of some measures.51

The development and implementation of the Objective 2 SPDs52 is led by the county (Amt) authorities, but with a differential level of responsibility between the Funds. The ERDF is co-ordinated by the NAIT, which also retains responsibility for making the final decision on applications, undertaking payments and monitoring on-going projects. The role of the counties is operational steering of the programmes and, together with the district municipalities (Kommune), to put forward projects for consideration. By contrast, ESF operates in a more devolved manner: the Ministry of Labour has overall responsibility for monitoring and controlling ESF, especially financial aspects and Commission liaison, and (via the Directorate General for Employment, Placement and Vocational Training) provides general guidance and supervision. However, operational administration of the ESF for Objective 2 is devolved to regional level; the county-based Executive Committees take all decisions pertaining to financing.

In terms of management structures, the two Objective 2 Monitoring Committees are chaired by central government (NAIT) and have general responsibility for the programmes. Operational decision-making on programme management is delegated to county-led Regional Executive Committees for Nordjylland and Lolland respectively. A secretariat provides administrative support to both committee levels. Whereas in Lolland, the Executive Committee comprises both elected politicians and administrative staff, in Nordjylland the Executive Committee only consists of officials, and a separate Regional Steering Committee has been established (comprising mayors, senior political figures and chaired by the leader of the county council) to provide political backing for major development projects and

51 Halkier H (1998) Danish Regions and the Europeanization of Regional Policy, European Studies Occasional Papers 27, ERU, Aalborg University, Denmark. 52 Halkier H (1997) The North Jutland Objective 2 Programme: Institutions and Implementation, IQ-Net Occasional Paper No.1, EPRC, Glasgow.

45

ensure co-ordination between Objective 2 interventions and other regional development initiatives.

In 1996-97, a series of laws and executive orders modified the above system of Structural Fund administration.53 The main aim was to remove central government bodies from the early stages of project development by withdrawing from the Executive Committees where they would be making recommendations on applications for which they would subsequently make the approval decisions. Under the revised structure, in Nordjylland applications are now first professionally scrutinised by new Advisory Committees (for industry and services, education, and tourism) which then submit advice to a new Co-ordinating Committee (replacing the former Executive Committee), made up of representatives of the county and district levels as well as social partners, but not the national government. The Co-ordinating Committee is responsible for co-ordinating applications and making recommendations to the NAIT with the formal approval of the Regional Steering Committee. This new structure is intended to ensure a more professional treatment of applications by bringing in expert opinions and a better division of labour.

It is worth noting that external bodies also play a prominent role in the implementation of the Objective 2 programme, especially in Nordjylland. Assistance with proposal preparation is available from local business counsellors and advisory organisations at district or regional level, and the ‘framework programmes’ eg. advisory services or various forms of organisational and technical support - are generally delivered by separate organisations such as the North Jutland Development Fund, North Jutland Business Service, two technology Information Centres and a university science park (NOVI).

3.4 Finland In Finland, the Ministry of the Interior has overall responsibility for the design and co-ordination of regional policy, including the preparation of legislation, designation of the assisted areas map, promoting regional policy among other areas of government policy, and co-ordinating the administration of regional policy at regional and local levels. A second central government institution, the Ministry of Trade and Industry, is responsible for the implementation of regional policy through a network of 19 regional business service offices throughout Finland. This structure has evolved for historical reasons; originally the design and co-ordination of regional policy was located within the Prime Minister’s department, later transferred to the Ministry of the Interior. In some respects, it permits a more ‘strategic’ approach to be taken to regional development since regional policy design is not associated within one particular sectoral policy area. As in the UK, a degree of decentralisation is evident with the growing use of integrated regional offices of central government for policy implementation. The activities of the district offices of State ministries with economic development responsibilities (Trade & Industry, Labour, Agriculture and Forestry etc) have progressively been co-

53 Damborg C (1998) The North Jutland Objective 2 Programme: Co-ordination and Evaluation, IQ-Net Occasional Paper No. 2, EPRC Glasgow.

46

ordinated and, since 1997, are housed in regional Employment and Economic Development Centres (T&E Centres).

A further element of the decentralisation process was the establishment of a regional tier of government. In 1994, 19 regional councils were created throughout Finland comprising representatives of the municipalities and with responsibility for regional development (previously the responsibility of 12 provincial State boards). The role of the councils is still developing, especially their relationships with the municipalities and State district offices, bodies with which they need to co-operate in promoting strategic regional thinking and the preparation of regional development programmes for each region.

The implementation of the Structural Funds in Finland is strongly governed by State authorities. Two unique features characterise the Objective 2 programme: there is a single SPD covering all of the Objective 2 areas; and the duration of the programme period was split into two short programming periods - 1995-96 and 1997-99.

At national level, the Ministry of the Interior co-ordinates and evaluates regional development and the Structural Funds, working in collaboration with several sectoral ministries: Trade and Industry, Labour, Education, Transport and Communications, Agriculture and Forestry, and Environment. With the exception of the Ministry of Education, which has no regional structure, most ministries have delegated a considerable part of their decision making to their regional organisations. Objective 2 and the ERDF are administered by the Ministry of the Interior. All ESF assistance (and some ERDF) is routed through the Ministry of Labour, although the Ministry of the Interior co-ordinates ESF actions under the regional programmes and the Ministry of Trade and Industry and Ministry of Education also play a part in their implementation.

There is a single Monitoring Committee for the six Objective 2 programmes, chaired by the Ministry of the Interior, and funding decisions are taken by the regional-level district offices of national government ministries (since 1997, in T&E Centres). Although the regional programme for the Objective 2 areas was based on plans prepared by eight regional councils, the process was co-ordinated by a national working group composed of central ministries and the Association of Finnish Local Authorities, overseen by the Ministry of the Interior. Sectoral ministries have been keen to protect their interests in the implementation of development programmes through their own administrative systems.54

The structure for implementation of the Structural Funds in Finland reflects the historical approach to regional development, which has traditionally been dominated almost exclusively by State authorities - government departments or specialised state agencies. As noted above, a regionally-based approach to regional development was introduced in 1994, and the regional councils direct and coordinate planning and implementation of both national and EU regional development programmes, but the role of the councils remains relatively weak compared to the State authorities and municipalities. (In the autonomous

54 Eskelinen H, Kokkonen M and Virkkala S (1997) Appraisal of the Finnish Objective 2 Programme: Reflections on the EU Approach to Regional Policy, Regional Studies, 31(2), pp167-172.

47

Åland Islands, responsibility for regional development lies with the regional government.)

Nevertheless, the increasing importance of the regions is reflected in the fact that regional Programme Management Committees have been placed in the regional councils in each of the six Objective 2 areas. These PMCs are made up of representatives of the region, the local offices of the national ministries and social partners. Project funding is largely decided by central ministries or their local offices, but the Regional Management Committees examine and formally adopt all projects, giving them some control over how the SPD is implemented at regional level.55

Advisory groups also exist with a pro-active role to assist in project development at an early stage, working on issues such as tourism, technology and training. A notable feature of the Finnish system is that there is a separate decision-making structure for business aid applications; special working groups on business aid exist in each region which discuss business development applications in tandem with the local PMC. This ensures specialist appraisal and confidentiality. The co-funder has final approval - so retaining systems in place prior to the implementation of the Structural Funds.

3.5 France Traditionally, regional policy in France has been highly centralised, with the State as the dominant authority responsible for planning territorial development through a series of national plans and major infrastructure projects promoting industrial decentralisation. As a distinctive example of delegated administrative authority for regional policy, the Délégation à l’aménagement du territoire et à l’action régionale, DATAR (the delegation for spatial management and regional action) was created in 1963 with the specific remit of regional policy making and of co-ordinating regional development objectives across government departments. DATAR was unique in having a strategic rather than an implementation role, being used to set up structures or mechanisms for dealing with particular problems but leaving the subsequent operational management to another part of national or local government. Several co-ordination mechanisms were also established, although with limited effectiveness in practice: an inter-ministerial committee for regional development (Comité interministériel d’aménagement et de developpement du territoire, CIADT); and a national council for regional development (Conseil national d’aménagement du territoire, CNAFT). Decentralised administration of central territorial planning was pursued through the creation of regional prefets in 1964.

Since the 1980s, this top down approach to aménagement du territoire, through a hierarchy of national, regional and sectoral development plans and large-scale projects, has been progressively superseded by policies promoting endogenous local development.56 Decentralisation legislation established the Region as a level of self-government and reinforced the economic development role of local government. A new system of national-regional

55 CEC (1996) op. cit. 56 Balme R and Jouve B (1996) Building the Regional State: Europe and Territorial Organization in France, in L Hooghe (Ed.) op.cit.

48

bargaining for economic development planning and resource allocation was introduced, whereby each Region prepared a regional development plan as a basis for negotiating a Contrat de Plan Etat-Région - a five-year planning and expenditure schedule agreed between each regional council and the central government. This approach is distinctive in the way that it combines a highly centralised system of sub-national policy administration through the hierarchy of Regions and départements (maintaining the State as the predominant actor in regional development policy) and a structure of sub-national government in the form of Regions whose responsibilities include competence for planning and economic development. As in Austria, informal co-operation through personalised networks of political and administrative actors play an important role in policy-making.57

European regional policy is considered to be complementary to French national policy implemented in the same areas, and there is considerable integration of administrative structures. The State leads Structural Fund management and implementation at all tiers. The main co-ordinating authority at national level is DATAR, currently under the Ministry of the Environment. DATAR negotiates Structural Fund issues with the Commission, shapes national and regional approaches to Fund administration, undertakes overall Fund co-ordination and administers the ERDF. The ESF is managed at national level by the ESF Mission of the Ministry of Employment; although the Mission supervises ESF spending in Objective 2 regions, the regional programmes are managed by DATAR and run at the regional level. The ACCT (Agence Comptable Centrale du Trésor) in the Treasury is the national point of arrival of Structural Fund resources, which are then disbursed to DATAR and the relevant ministries.

At regional level, the regional offices of the State (préfectures de Région) have taken the lead in drawing up and managing the implementation of Objective 2 programmes, largely consistent with the contrats de Plan and giving an absolute priority to job-creating projects.58 The préfect is charged with ensuring the coherence of EU and national policies, organising and liaising between the programme partnership and approving projects. In practice, administration and co-ordination are undertaken by a ‘Mission Europe’, housed in the regional préfecture’s overall secretariat - the Secrétariat Général aux Affaires Régionales or SGAR. Among the wider partnership, regional councils have considerable influence, usually being the main programme co-financer, but their funding plans are constrained by the commitments made in their Contrat de Plan. For Structural Fund programmes delivered at regional level, two main committee structures operate: a Monitoring Committee (MC), and a Programme Management Committee (PMC). The MC meets twice a year to monitor overall programme progress and take steering decisions. Chaired by the regional préfect, and prepared by the European secretariat, it includes a wide representation of partners including the regional council, the regional offices of state for technology, business, the environment, training, etc, the

57 Benz A and Eberlein B (1998) Regions in European Governance: The Logic of Multi-Level Interaction, RSC Working Paper, No. 98/31, European University Institute, Florence. 58 Lagrange R (1997) French policy perspectives, in J Bachtler and I Turok (Eds.) op. cit.

49

conseil économique et social, DATAR, the Ministry of Employment Commission services and municipalities. Trade unions are represented indirectly through their membership of the conseil. The Programme Management Committee (PMC) is the principal operational forum, project funding decisions being made here. In most cases, there is a single, regional-level PMC, but in fragmented Objective 2 regions, PMCs may operate at the département level (again chaired by agents of the State, the préfets de département). These committees meet more often than the MC, usually every two or three months. Decision-making responsibility is devolved to these committees for all decisions except those that are regionally significant, very specialised or training-related.

3.6 Germany In Germany, regional development policy is constitutionally the responsibility of the Länder. The Federal Government, through the Federal Ministries of Economies and Finance, participates in a co-financing and coordinating role with regional policy being treated as a ‘joint task’ between federal and state levels (Gemeinschaftsaufgabe ‘Verbesserung der regionalen Wirtschafts-struktur’). The distinctive aspect of this approach is a formal planning system which brings the two levels of government together for consensus-based decision-making with the aim of ensuring that excessive competition between the states in the provision of regional aid does not occur. There is a clear framework for regional development and a specific division of responsibilities between federal and state levels with respect to their roles in the specification of regional policy objectives, the definition of eligible areas, the administration of measures and the provision of funds. The Federal Republic of Germany initially attempted to integrate the implementation of the Structural Funds within the structure of national regional policy insofar as possible, using the Funds to co-finance German regional assistance for productive investment and local infrastructure under the Federal-Land ‘Joint Task’ (GA).59 Over time, this has changed, with a progressive ‘de-coupling’ of ERDF resources from the Joint Task to implementing the Structural Funds increasingly within purely Land programmes.60 Thus, the Länder have been integrating the Structural Funds into their state economic development strategies, using the Funds to co-finance the existing regional policy instruments and administering them via existing departments/agencies.

Although there is a strong degree of devolved implementation, the Federal Government has a significant influence on Structural Fund programmes, particularly those of the smaller western programmes (and the Objective 1 programmes of the eastern Länder). The Federal level is responsible for formal communication with the Commission and the allocation of resources among regions. More so than in other Member States, subnational authorities

59 Nägele F (1996) Regionale Wirtschaftspolitik im kooperativen Bundesstaat. Ein Politikfeld im Prozeß der deutschen Vereinigung, Opladen, Leske & Budrich. 60 Lang J, Naschold F and Reissert B (1998) Reforming the implementation of European Structural Fund: a next development step, FS II 98-202, Wissenschaftszentrum Berlin für Sozialforschung, Berlin.

50

have an influential role in negotiating with the Commission, including the process of area designation.61

The Federal Ministry of Economics oversees the coordination of the ERDF, while the Federal Office for Employment, the labour market agency of the Federal Government, has a co-ordination function for the ESF, albeit limited in the case of Objective 2 programmes. Funding flows from the European Commission to the Federal Ministry of Finance, and is then distributed to each of the Länder treasuries. The relevant Land ministry draws down funding according to its responsibilities in implementing the various programme measures.

State governments are solely responsible for the planning, implementation and evaluation of Objective 2 programmes with the dominant players being the Land ministries of economics.62 The Federal Government role is primarily one of co-ordination. There is a single Monitoring Committee for Objective 2, chaired by the Federal Ministry of Economics and including representatives of relevant Commission DGs, Federal government ministries and Land government departments of economics and labour. Fund-specific meetings, chaired by the appropriate federal government departments, are held immediately before the ‘plenary’ Monitoring Committee meeting. The economic and social partners are not included in the Monitoring Committee itself, but consultative meetings are held in advance of Monitoring Committee meetings to allow trade unions, employers associations, local authority associations etc to give their views on the agenda items. During the 1994-96 programming period, the larger Objective 2 programmes (receiving more than 40 MECU) created regional sub-committees of the national Monitoring Committee; other regions have more informal consultative committees. The Berlin regional Monitoring Committee covers both the West Berlin Objective 2 and East Berlin Objective 1 programmes.

In each Land, there is normally a strict division of ERDF and ESF administration between the two State ministries responsible for economics and labour/social affairs respectively. As in Austria, Objective 2 resources are allocated through existing funding circuits, and project generation and delivery therefore operate primarily through the appraisal and implementation mechanisms of existing Land programmes. For the most part, Objective 2 management has been centralised at Land level, although local involvement has been growing with experience of the Structural Funds: in Nordrhein-Westfalen, some functions such as appraisal and monitoring were already devolved during the first programming period to a network of 12 ‘regional secretariats’ established at local level as well as regional committees of public and private sector organisations and political parties.

61 Conzelmann T (1995) Networking and the Politics of the Regional Policy: Lessons from North-Rhine Westphalia, Nord-Pas de Calais and North-West England, Regional & Federal Studies, 5(2), pp.134-172 62 Wollman H, Lang J and Schwab O (1996) Partnership Arrangements within EU Structural Fund Programmes: Case Study on Germany, Institut für Sozialwissenschaften, Humboldt-Universität zu Berlin.

51

3.7 Italy Despite its description as a ‘regionalised unitary state’,63 regional development has historically been a strongly centralised area of policy in Italy, with a primary role for national government and only limited decentralisation to regional bodies. Traditionally, the focus of regional policy was on north-south disparities, especially the backwardness of the Mezzogiorno, through national government policy and planning initiatives - infrastructure investment, industrial subsidies, the (re)location of state-owned industry and promotion of growth nuclei. Over the past two decades, the special intervention for the South has progressively been replaced with a regional policy addressing regional problems across the country (ie. including the Centre-North) and a reorganisation of the institutional infrastructure for regional policy, including the abolition of special agencies for the Mezzogiorno. Current regional policy responsibilities are divided between the Ministry for the Budget, which plays the primary role with respect to policy formulation (including area designation) and European issues, and the Ministry of Industry, which has overall responsibility for the administration of regional incentives (and the Ministry of Public Works for infrastructure). Other central government departments are also involved in regional development through an interministerial committee for economic policy (CIPE).

The role of the regions in economic development is weak but growing, particularly as a result of EU regional policy. Under the 1948 Constitution, provisions were made for the creation of regions (along with provinces and communes) and for the allocation of special resources for development purposes, but it was not until the 1970s that regions were set up throughout the country. The regions were given competencies for economic development, strengthened through various regionalised initiatives during the 1980s and 1990s, but their capacity for effective policy action remains limited, and central government ministries continue to play the lead role in regional development matters. Nevertheless, the implementation of the Structural Funds has reorganised administrative structures sufficiently for the regional governments to develop and implement regional development programmes and to co-ordinate partnerships of public and private actors as part of the programme management process.64

With respect to the institutional arrangements for EU regional policy, the overall co-ordination of the Structural Funds in Italy, including negotiation with the Commission is carried out by the Ministry of the Budget. The Ministry co-ordinates financial management, disbursing Commission funds to the various implementing authorities as appropriate, and, together with the Ministry of Industry, is responsible for ERDF. The Ministry of Labour has responsibility for ESF, including co-ordination of regional links with the Commission although, ESF actions in Objective 2 regions are managed by the regional authorities (departments for training and employment). A national-level coordinating/monitoring body (cabina di regia) has recently been

63 Grote J R (1996) Cohesion in Italy: A View on Non-Economic Disparities, in Hooghe L (Ed.) op. cit. 64 Desideri C (1995) Italian Regions in the European Community, in B Jones and M Keating (Eds.) op.cit. Bagnasco A and Oberti M (1998) Italy: ‘Le trompe-l’œil’ of regions, in P Le Galès and C Lequesne (Eds.) op. cit.

52

established as part of the Ministry of the Budget, to promote more effective co-ordination between national government departments; with a degree of autonomy from the public administration, the cabina is intended to reduce bureaucracy and speed up the allocation of funds.

The implementation of the Structural Funds in Italy varies by type of programme reflecting the scale of funding and institutional management capability. Whereas the Objective 1 programmes have significant central government involvement (with the multi-regional component being co-ordinated by the Ministry of the Budget), the management of Objective 2 programmes has been largely devolved to regional authorities with a limited role for central government bodies.

The launch of the first Objective 2 programmes in 1989 represented a considerable challenge for regional institutions, which had no prior experience of implementing industrial policy at the regional level.65 The responsibility for Objective 2 programme administration was devolved to regional authorities from the outset, except for SME grant schemes managed by the Ministry of Industry (later decentralised for regional management). As in other countries, programme management has involved a ‘learning by doing process’ over successive phases of programming (eg. introducing innovations such as calls for tenders for projects, and the application of selection criteria), key factors being the support from Commission services (moreso than from central government) and the involvement of sub-regional actors in the drafting and implementation of programmes, deriving from the local nature of Objective 2 problems.66

The lead role in programme management is taken by the Industry Assessorate (Assessorate dell’Industria). In the early programme period, this involved co-ordinating the largely separate activities of sectoral policy departments in the drafting and delivery of programmes. In 1995, regional co-ordinating units were established in each Objective 2 region (subsequently in Objective 1 and 5b areas) at both political and administrative levels to co-ordinate the different departments implementing priorities or measures in each programme.67 Nevertheless, co-ordination and communication between implementers remains weak, limiting the strategic coherence and direction of programmes68 Objective 2 Monitoring Committees are chaired by the president of the regional council, the key partners being the regional council departments, national government, Commission services and (since 1994) representatives of the employers’ associations, trades unions and local authorities.

3.8 Luxembourg As a very small unitary state, regional development in Luxembourg is, not surprisingly, completely centralised. Regional policy is administered by the

65 ISMERI Europa (1996a) Ex Post Evaluation of Objective 2 Programmes in Italy, Report to the European Commission (DG XVI), Rome. 66 ISMERI Europa (1996b) The Contribution of Community Structural Policies to the Economic and Social Cohesion of Italy, Report to the European Commission, Rome. Ernst & Young (1997) op.cit. 67 CEC (1996) The Structural Funds in 1995 - Seventh Annual Report, Commission of the European Communities, OOPEC, Luxembourg. 68 Crescini M (1998) The on-going evaluation of SPDs – Objectives 2 and 5b of the Region of Tuscany, Paper to the European Conference on Evaluation, Seville, 16-17 March 1997.

53

Ministry of Regional Planning (whose responsibility it is to establish the rules for the optimal utilisation of the territory), the Ministry for Economic Affairs (regional incentive administration), the Municipalities Development Department of the Ministry of the Interior (analysing and drawing up development projects) and the Ministry for Small Businesses and Tourism.

As with national regional policy, the implementation of the Structural Funds in Luxembourg is highly centralised. Virtually all aspects of implementation, from negotiation with Commission services to the management of programmes, are undertaken by central government departments. The Ministry of Economic Affairs has overall responsibility for the Funds and administers the ERDF. The ESF is managed by the ‘ESF Mission’ in the Ministry of Labour. The spatial scale also accounts for distinctive partnership. Arrangements for Structural Fund administration have been characterised as the “most extreme example of informal relations” in the EU with a great reliance on consensus, a limited number of key actors who know each other, tight linkages between partners and wide-ranging consultation that identifies most projects.69

3.9 Netherlands Regional development in the Netherlands has been characterised by gradual devolution from central government to the provinces and municipalities over the past decade. Until the 1980s, Dutch regional policies (which were significantly aligned with physical planning and environmental policies) featured strong top-down State control in the design and implementation of instruments and measures. During the decade, policy thinking moved from a centralised top-down approach, though a redistribution of national resources by central government, to decentralised bottom-up economic development, based on regions exploiting endogenous potential.70 Superseding the traditional goal of interregional equity, national efficiency became the overriding aim of regional policy from 1980 onwards, and more resources were allocated to promoting the business environment (especially site development), as opposed to enterprise aid for the northern part of the country.71

Although the more ambitious ideas for creating regional authorities and governments were not realised, the provinces were in a good position to reinforce their role as strategic intermediaries for regional planning and development.72 Increasingly, central government has withdrawn from active regional policy administration in favour of delegating powers for regional economic development to the 12 provinces, which in turn, have delegated some responsibilities to regional at local levels. Currently, the Ministry of Economic Affairs (Ministerie van Economische Zaaken) co-ordinates regional policy in terms of the legislative framework, strategic development issues, the designation of assisted areas and administration of regional financial

69 Stern E (1997) op. cit. 70 Kleyn W H and Bekker M (1997) Integrated Regional Development under EU Regional Policy: The Dutch Experience, in J Bachtler and I Turok (Eds.) op. cit. 71 Oosterhaven J (1996) Dutch Regional Policy Gets Spatial, Regional Studies, 30(50), pp.527-532. 72 Hendriks F (1997) Regional Reform in the Netherlands: Reorganizing the Viscous State, in M Keating and J Loughlin (Eds.) op. cit.

54

incentives. The implementation of regional policy, and broader responsibility for planning, co-ordinating and administering economic development and spatial planning, are carried out sub-nationally by the provinces.

Implementation of the Structural Funds in the Netherlands is co-ordinated by the Ministry of Economic Affairs with respect to Commission negotiation, area designation and financial control. Programme management is highly decentralised, dating back to 1988 when the government decided to devolve responsibility for managing Objective 2 programmes entirely to the regions. In practice, the process was more one of progressive decentralisation over time, with the provincial authorities gradually taking on increasing responsibilities for chairing the Monitoring Committee, managing programmes, delivering projects and financial administration.73

In 1989-91, the development of the first programme was very much led by the Ministry of Economic Affairs which commissioned the drafting of the Regional Development Plans (ROPs) that underpinned the content of the Objective 2 programmes. The Ministry was also responsible for the execution of the ERDF until 1990; similarly, the Ministry of Social Affairs was responsible for the ESF part of Objective 2 programmes. At subnational level, the provinces were responsible for the secretariat of Monitoring Committees and implementing ERDF operations. From 1990 onwards, the management of Objective 2 changed significantly with a decentralisation of programme responsibilities to the provinces (eg. to the Development Company Overijssel in the case of the Twente Objective 2 programme). The provinces became responsible for drafting the programmes, within the framework of nationally determined priorities, and their financial administration, combining a strategic programme management role with enhanced project delivery responsibilities, especially in terms of programme marketing, project generation, and the preparation of project selection decisions. Progressively over the 1992-93 and 1994-96 programming periods, the provinces were able to focus more on programme ‘steering’ as the project implementers at municipal level became more experienced.

In parallel to the decentralisation of ERDF, the administration of ESF was also decentralised as part of a reorganisation of the Ministry of Social Affairs to create a Central Employment Agency (Centraal Bureau Arbeidsvoorziening), subdivided into several Regional Employment Agencies (RBAs). After 1991, ESF implementation was decentralised to the RBAs, with increasing input from new ‘ESF co-ordination points’ tasked with promoting the programme (especially liaison with regional partners), preparing project selection decisions (for approval by the RBAs) and monitoring programme implementation.

During subsequent programming periods, the decentralisation process was reinforced by several developments: regional representatives of the Ministry of Economic Affairs took over many of the regional development functions formerly carried out in The Hague; responsibility for developing and implementing the programmes was totally run by regional authorities (in some cases with considerable externalisation of functions to private consultants);

73 Research voor Beleid (1996) In-Depth Evaluation of Twente: Ex Post Evaluation of the Industrial Core North Objective 2 Regions, 1989-1993, Report to the European Commission (DGXVI), Leiden.

55

and active networking between Objective 2 programme managers – through ‘Programme Management Netherlands’ - provided a supportive process of exchange-of-experience which assisted regional-level learning about programme management functions.

Currently, there is a common management structure for each programme74: a Monitoring Committee chaired by the political representative of the province and with representatives of the national government, EC and social partners; a Steering Group normally chaired by the political representative of the region or local authority and with a wide range of regional/local partners; and a programme management secretariat as a professional bureau to support project development, decision-making and monitoring of the programme (acting either as an independent body or as part of the provincial/regional authorities).

3.10 Spain In Spain, the Ministry of Economy and Finance has overall responsibility for basic economic planing legislation, the co-ordination of regional development, area designation and the administration of regional aid. These functions are exercised in conjunction with the Autonomous Communities which are constitutionally responsible for regional and industrial development (including SME development, communications, tourism, urban planning etc). Under the corresponding statues of autonomy, several regions with national minorities (Pais Vasco, Cataluña and Galicia, Communidad Valenciana, Canarias, Navarra and Andalucía) have amore developed autonomous status, the other regions being covered by general statutes, enabling regional authorities to request further powers. Over the past decade, such requests have narrowed the differences between the two types of region, with ‘full’ autonomous powers progressively being accorded to Andalucía, Communidad Valenciana, Navarra and Canarias.75

Especially in the early days of programming, the implementation of the Structural Funds was dominated by central government. The lead role for all three regional policy objectives is played by central government through the ERDF co-ordination department of the Directorate General for Planning in the Ministry of Economy. The Ministry co-ordinates planning and EC negotiations, especially in respect of Objective 1. Financial monitoring is undertaken at national level through the Ministry of Finance. The Ministry of Labour and Social Affairs is the overall managing body for ESF, the administration of the Fund being divided according to the nature of the programmes, between the seventeen Autonomous Communities and the central departments (the Autonomous Communities manage about 25 percent of ESF assistance).

Uniquely, the Spanish authorities have retained the three-stage procedure for Objective 2 programming (plans, CSFs, OPs) not just for 1989-93 but also for the 1994-96 and 1997-99 programming periods. The lead role has been taken by central government, with regions involved in formulating plans (eg.

74 Broos L (1999) Structural Funds in the Netherlands: a short overview, ERAC, mimeo. 75 Morata F and Muñoz X (1996) Vying for European Funds: Territorial Restructuring in Spain, in L Hooghe (Ed.) op.cit.

56

elaborating first drafts of sub-regional documents) but not in final decisions.76 “The presentation of a unique plan was in response to the government’s desire to have a broader operational margin in order to decide on the territorial allocation of the resources which depended on the central administration….(..40 percent for Objective 2 areas77)”.78

For Objective 2, the administrative arrangements have been quite different between the 1989-93 and the 1994-99 programming periods. In 1989-93, each CSF established an ERDF and an ESF programme for each region, as well as one or two ESF multi-regional programmes which covered the whole Objective 2 territory. There was therefore no co-ordination between the actions co-financed by the two Funds.

At this stage, only the programmes for Cataluña and País Vasco had dedicated Monitoring Committees. The other five programmes were served by the single CSF Monitoring Committee which met in Madrid.

In 1994, the Spanish central authorities were the only Member State to decide to retain the three-stage CSF procedure for Objective 2, against the opinion of the Spanish regional authorities and Commission services. Several improvements were however achieved in the field of co-ordination and partnership. First, integrated ERDF-ESF programmes are now delivered in each of the seven eligible areas. The central administration has nevertheless kept a multi-regional ESF programme consisting of vocational training actions implemented by the Ministry of Labour. Second, although all Monitoring Committee secretariats continue to be sited in Madrid, in the Ministries of Economy and Labour, each regional programme now benefits from its own dedicated Monitoring Committee which usually meets in the region itself. The CSF Committee is only devoted to general and co-ordination issues. Third, it is now the regional authorities which co-ordinate the ERDF projects of local authorities, rather than the central administration as it was in 1989-93. The regional authorities are better placed to understand and follow the co-ordination of local actions. Fourth, at this stage all the documents concerning the Structural Funds (programmes, working documents for Committee meetings, annual and final reports), are drafted by the central and regional administrations.

In terms of the financial weight of the three public administrations concerned by Community aid, the importance of the regional and local levels has increased. In 1997-99, central authority projects accounted for 36 percent of ERDF allocations from regional programmes and 18 percent of the ESF allocations from the multi-regional ESF programme. The central authorities’ share of the ERDF varies between programmes (from 0 percent in Baleares and 14 percent in Navarra to 42 percent in Aragón.

Co-ordination still presents some problems, not only between the two co-ordination bodies (the Ministries of Economy and Labour on one side and the regional departments of Economy and Labour on the other) but also between

76 Quasar (1997) Ex post Evaluation of Objective 2 Programmes in Objective 2 ‘Mediterranean’ Regions, Report to the European Commission (DGXVI), Madrid. 77 40 percent in 1989-93, but only 30 percent in 1994-99. 78 Morata and Muñoz (1996) op. cit.p.207.

57

these co-ordination bodies (which do not have operational responsibilities) and the central and regional beneficiary departments.

3.11 Sweden In Sweden, the Ministry of Industry and Trade has overall responsibility for the design and supervision of the implementation of regional policy. The administration of regional policy is carried out by the National Board for Industrial and Technological Development (NUTEK) with respect to providing financial support to enterprises, developing the competence of regional and local authorities in the field of regional development and evaluating regional policy. Responsibility for implementing and co-ordinating state regional development measures at regional level largely rests with the County Administration Boards (Lansstyrelsen). The CABs are essentially the local arm of national government (the head of the CAB is the country governor, appointed by the government) and are used as a vehicle for implementing national policies, including regional policy, at the local level.

While there is a distinct delineation between EU and national regional policy, Sweden has implemented the Structural Funds utilising the same institutional structure used for national regional policy implementation. At the national level, the Ministry of Industry and Trade oversees implementation of ERDF, while NUTEK provides the secretariat for the Fund. The ALMS (National Labour Market Board) provides the same designated function for ESF on behalf of the Ministry of Labour, although the Objective 2 programmes are overseen by NUTEK. Funding flows from the Commission via the Ministry of Finance to NUTEK and ALMS, who deal directly with project payment. (The handling of payments to the regions is currently being decentralised to the regions).79

Although the administrative arrangements for the Structural Funds have essentially the same framework as for national regional policy, the main difference is the level of local partnership and project decision-making responsibility within EU programmes. The delivery of programmes is co-ordinated locally by the CABs. Each region also has its own Project Management Committee (PMC) which decides on project applications and its own Monitoring Committee. The decision-making of the PMC is supported by various committees and advisory groups, but their importance varies across regions. Implementation of SPDs involves a complex structure combining decentralised decision-making in project selection with centralised financial management. For each project, at least two decisions are required, one by the national body co-financing the project, and one by special, local Decision Groups (established to decide on EU funding)80 and comprising representatives of county councils, municipalities, local partners etc. Central government influences the implementation of the Structural Funds in several ways: the Objective 2 Monitoring Committees are chaired by the Ministry of Industry and Trade, and, given that national government also finance the CABs, it can also influence the ability of CABs to co-finance Structural Funds

79 Brown R and Bachtler J (1998) Business aid and the Structural Funds in Sweden, Report to the European Commission (DG XVI), EPRC, Glasgow. 80 Van der Wee D (1996) The implementation of Objective 2 in Sweden: A preliminary assessment of the situation at the end of 1996, Commission of the European Communities, Brussels (mimeo).

58

projects. The CABs have a major role in the programme management process, co-ordinating the management and providing the secretariat for the Decision Groups. The local authorities have become more involved in regional policy through their representation in the PMCs or the Decision Groups. The exception to this general picture is the Objective 2 programme of Fyrstad which established a small, independent secretariat to manage its programme since the eligible area cuts across two CAB areas (Alvsborg and Bohus), which traditionally had little experience of working together on regional development.

It is anticipated that regionalisation of Objective 2 implementation will increase in future, partly through the decentralisation of financial management, but also because of the decision to adopt new ‘regional growth agreements’ as part of the administration of national regional policy. Mirroring the implementation of the Structural Funds, regions will be expected to undertake regional analyses and develop regional strategies, to be negotiated with central government and then implemented by CABs or other self-governing bodies in close collaboration with regional/local partnerships.

3.12 United Kingdom The United Kingdom has historically centralised much of regional policy management within a single government ministry. The UK Department of Trade and Industry (DTI) has the primary responsibility for designing regional policy legislation, designating the assisted areas map and implementing regional assistance in England, latterly together with the Department of Environment, Transport and the Regions (DETR), through a network of nine central government offices located in the English regions. In Scotland, Wales and Northern Ireland, different arrangements apply. Each has its own central government ‘territorial’ department which also plays a role in the design and implementation of regional policy. Following devolution, Wales and Scotland will have much greater autonomy in their policy and institutional approaches to regional development from 1999 onwards. In the English regions, Regional Development Agencies have been created, reporting to central government (DETR), which will be significant Structural Fund actors.

As in Sweden, EU regional and national regional policy are regarded as separate in the UK, although they are administered through the same institutions. Arms of central government have the lead role in the implementation of the Structural Funds, but the practices of programme management vary between the constituent countries of the UK. With respect to financial management, all EU monies are transferred to the UK Treasury in the first instance and thereafter disbursed directly to relevant spending departments; the Treasury is also responsible for accounting and additionality matters. Policy issues, including negotiation with the European Commission, are led and overseen by the Department of Trade and Industry. In recent years, Fund-specific negotiations have been transferred to the ‘lead’ departments - Department of the Environment, Transport and the Regions (ERDF), Department for Education and Employment (ESF) (and Ministry of Agriculture, Fisheries and Food (EAGGF)).

59

Management of Structural Fund programmes is partly subsumed within existing government structures. There is a Programme Monitoring Committee for each regional programme, overseen by central government (the DTI) and managed by the Regional Government Offices (chaired by the DETR) in the English regions, by the Welsh Office in Wales, the Scottish Office in Scotland and the Northern Ireland Office in Northern Ireland. Co-ordination at central level is exercised through an inter-departmental committee bringing together all of the sectoral ministries, the territorial departments and UKRep, and chaired by the DTI. Similar co-ordination arrangements exist for the regional level, for example through the ‘General Practitioners’ groups within the Regional Government Offices in England and the Welsh/Scottish Offices.

The administration of Objective 2 programmes in England is undertaken by European Secretariats in the Regional Government Offices. These offices have delegated responsibility for local funds but are responsible to government departments in terms of the administrative framework for programmes and financial accountability. Monies are allocated through a process of competitive bidding: project appraisal and selection criteria are drawn up in partnership, applications invited and awards made by decision-making committees, on the basis of the appraisal criteria.

In Scotland, there is a distinctive approach to programme management. The Scottish Office has overall responsibility for programme delivery, chairs the Monitoring Committees and makes Structural Fund payments. However, in the case of the largest programmes, it delegates the secretariat functions (including administration and day-to-day management, co-ordination and servicing of committees, programme publicity and project generation) to dedicated independent ‘programme executives’. Programme Management Committees are in place to make project decisions under each programme, frequently supported by specialist Advisory Groups established under each main programme priority.

For the first two programming periods, Objective 2 programme management in Wales was run by the Welsh Office. In 1997, implementation of all EU programmes in the Principality was devolved to an independent secretariat. Modelled on the regional arrangements for Scotland, the Welsh European Programme Executive Ltd (WEPE) was established with administrative responsibility for European funding for the whole of Wales. With the aim of acting as an independent, impartial and transparent organisation, WEPE markets the programmes, invites and appraises applications, issues grant approvals and undertakes physical monitoring of the programmes. For the Objective 2 programme (Industrial South Wales), new decision-making structures were also introduced for the 1997-99 period, including the setting up of a ‘Core Group’ (as a type of programme management committee, taking a strategic perspective on the delivery of the programme) and five Project Assessment Groups, one for each main priority in the programme, to advise on the scoring and selection of projects.

61

4. REGIONAL DEVELOPMENT PLANS AND STRATEGIES

Over the past decade, the most immediate programming challenges were often associated with the preparation of regional development plans and the development of strategies. In each programming period, new regulations or guidelines governed the process and content of programmes, placing new demands on participants – regional secretariats and partnerships, national authorities and Commission services. The following chapter explores how the response of participants to these challenges has evolved over time. Echoing the theme of other chapters, there is clear evidence of learning and incremental adaptation and improvement in developing programmes and strategies.

4.1 Regional Development Plans and Programmes: 1989-93 With the approval of new regulations by the Council in 1988, the Member States and the European Commission were immediately confronted with the task of implementing the new principles of the reform. However, regional eligibility was only immediately clear for Objective 1 areas: the designation of Objective 2 regions was not officially announced until March 1989, following submission of lists of areas satisfying the criteria laid down by the Council and the Commission’s adoption of a list of eligible areas. Sixty regions (or parts of regions) were finally selected under Objective 2 covering between 16 and 17 percent of the Community population.

The new rules required Member States to undertake a novel ‘forward planning effort’ in submitting multi-annual plans covering their proposed Objective 2 strategies for the subsequent three years, the national resources to be committed and the use of Community assistance. Under a three-stage system, Member States submitted regional development plans as a basis for negotiation (with the Commission) of a Community Support Framework (CSF) for each eligible region. The CSFs detailed the priorities to which subsequent implementation measures, in the form of Operational Programmes (OPs), were to relate. Following submission of the plans in June/July 1989 and subsequent negotiation, 53 CSFs were approved by December 1989, the remaining one for Spain (submitted as a single CSF for the seven Spanish Objective 2 areas) being approved in March 1990. The submission of OPs was undertaken at various times, often coinciding with CSF submission and occasionally pre-empting CSF approval. Most OPs were approved by July 1990.

In order to meet the very tight deadlines, regional development planning was quick and unsophisticated. Early Objective 2 plans were drafted by small groups or committees, frequently within an individual government department, and often in a hasty manner with limited discussion and consultation. In eight of the Member States, plans were submitted at regional level, in some cases being elaborated with the involvement of local authorities in the definition of policies and priorities, but central government authorities dominated the process. Even in countries with devolved approaches to regional development (eg. Germany, Belgium, the Netherlands), local bodies were hardly involved in the preparatory stages of programme development, with marginal influence

62

on the priorities and measures and the allocation of funds.81 The dominance of central government is reflected in the common priorities defined across different Objective 2 CSFs, as in France and the UK.

As part of the plan preparations, analyses of stated needs were relatively basic, often limited to a simple review of the regional economy and rationalisation of the intended spending allocation. The anticipated benefits of the programming approach were constrained by the lack of adequate economic and statistical indicators for forward analyses and the poor response to the requirement for ex ante evaluation of programmes; some Member States (eg. Germany) were dismissive of the feasibility, utility and relevance of ex ante appraisal. As noted in the appraisal of the Dutch, Belgian and German programmes in the ‘industrial core north’, neither the CSFs nor the OPs provided quantified or other measurable targets.82

While the principles of the reformed Structural Funds forced authorities at regional level to consider and adopt strategies more explicitly than before, many of these early programmes had little or no strategic coherence or direction, being viewed essentially as mechanisms for drawing down EU funding. Budgetary and financial accounting criteria were sometimes more important determinants of programme content than the strategic needs of the regional economies (as in Spain).83 Often, the plans were relatively traditional, related to pre-existing priorities and structures and prepared to a standardised format. Existing EU-funded plans were in operation in several regions and served as the base for the Objective 2 development plans eg. the Integrated Action Programme for Limburg (Belgium), the Luxembourg National Programme of Community Interest, the Integrated Mediterranean Programme and the Strathclyde Integrated Development Operation (Western Scotland). Predetermined national plans, such as the Contrats de Plan Etat-Région in France, also provided the basis for the strategic definition and implementation of CSFs.

The greatest strategic coherence was evident in regions with some previous experience of - and commitment to - strategic regional economic development planning. In the Netherlands, the programming approach was supported by the Dutch government as a contribution to effective decentralisation of economic development, and building on existing plans. “As a result, Objective 2 strategy development was soundly based on a thorough analysis of the regional bottlenecks and opportunities”,84 leading to focused strategies. The same applied to some German regions such as Bremen. In other countries, such as the UK, which lacked a formal planning process, programmes were characterised by numerous priorities and measures without strategic coherence.85 Programmes appeared to be made up of selections of eligible

81 Research voor Beleid (1996) Ex Post Evaluation of Objective 2 Regions – Industrial Core North 1989-93, Report to the European Commission (DG XVI), Leiden. 82 Ibid. 83 Quasar (1997a) Ex Post Evaluation of Objective 2 Regions in the Mediterranean Area, Report to the European Commission (DG XVI), Madrid. 84 Research voor Beleid (1996) op. cit. 85 Hall Aitken Associates (1996) Ex Post Evaluation of Objective 2 Regions 1989-93, Report to the European Commission (DG XVI), Glasgow.

63

projects, mainly from the priority schemes of local authorities.86 Likewise in Italy: “as with other structural interventions, planning remains an occasional activity, poorly supported by studies”.87 Inexperience limited the scope for intervention in new fields (eg. the environment) and encouraged project-based or sectoral thinking rather than a programme approach.88

Development objectives and axes were frequently general and vague. As one evaluation study noted, in a comment applicable to many other programmes: “OP priorities are fairly similar and indeed could be applied to economic development activity in most areas”.89 Many of the Objective 2 plans were dominated by support for productive investment (in the form of grants to enterprise) and local infrastructure, especially transport networks, rehabilitation of sites, and industrial estates (especially in Denmark, Germany and Italy). During the negotiations, the Commission disputed this approach: “Many Member States seem to have had difficulty, when preparing their plans, in shifting the emphasis of their applications for assistance to the Funds in line with Commission’s suggestions”….“Member States stated needs which, all too often, reflected a conception of regional policy based on the importance of infrastructures. This approach is not in itself sufficient to solve the new problems of economic development and conversion posed”.90

Negotiation with Commission services proved contentious and protracted in many cases.91 First, in some countries and regions, allocations for the programming period under the new CSFs had already been committed to earlier programmes (which were strongly infrastructure-based) pre-dating the reforms. This applied particularly to the UK, Belgium, the Netherlands and Luxembourg.

Second, the Commission frequently requested a reprioritisation of plans to place greater emphasis on employment creation, SME measures and environmental improvement. Disputes arose from the Commission’s pressure to reduce the proportion of ERDF funding going to large-scale infrastructure, pressure which was not always viewed as justified given regional needs. The shift to measures promoting the competitiveness of firms also entailed delays because of the need to have new or non-notified regional aid schemes examined and approved by the EC competition policy authorities.92

Third, the combination of grants and loans in the plans was often considered unsatisfactory by the Commission, with an excessive proportion of grant funding. Although the Commission made efforts to increase the proportion of

86 EPRC (1997) Ex Post Evaluation of UK Objective 2 Regions 1989-93, Report to the European Commission (DG XVI), Glasgow. 87 ISMERI Europe (1996) Ex Post Evaluation of Objective 2 Programmes in Italy 1989-93, Report to the European Commission (DG XVI), Rome. 88 Quasar (1997a) op.cit. 89 Pieda (1994) Interim Evaluation of Eastern England CSF Programmes, Pieda, Manchester. 90 CEC (1991) Annual Report on the Implementation of the Reform of the Structural Funds, Commission of the European Communities, OOPEC, Luxembourg. 91 Yuill D, Allen K, Bachtler J, Clement K and Wishlade F (Eds.) (1991) European Regional Incentives 1991, 11th Edition, Bowker-Saur, London. 92 CEC (1992) Second Annual Report on the Implementation of the Structural Funds, OOPEC, Luxembourg.

64

loans, Member States contradicted this, partly on the grounds of insufficient demand for loans in the regions concerned.

Fourth, the involvement of sub-national partners in the negotiations was resisted by some Member States as running counter to established institutional frameworks and practices. Within the limits of national institutional frameworks, the Commission “tried to promote its own conception of partnership”93 by associating regions in the planning and negotiation of programmes through partnership meetings in the regions and dialogue with sub-national authorities, an approach which was received with suspicion and even hostility in some Member States.

Lastly, additionality was disputed, with the Commission questioning national compliance with the principle, and Member States resisting the provision of information – a dispute which continued for several years as the Commission sought to verify additionality transparently.

The Objective 2 programmes, with a total allocation over 6 billion ECU, were launched early in 1990. European Commission data underlines the dominance of expenditure on productive investment (see Figure 4.1) with over 40 percent of total programme expenditure dedicated to SME support, the development of business services, innovation and technology transfer initiatives and enterprise investment. A further 36 percent of expenditure was allocated to physical regeneration (transport and communications links, industrial sites) and environmental improvement measures, and 21 percent was devoted to training and retraining measures funded by the ESF. The remaining two percent of funding was largely spent on technical assistance.

Figure 4.1: Breakdown of Objective 2 funding by programme area (1989-93)

0

5

1 0

1 5

2 0

2 5

3 0

3 5

4 0

4 5

P h y s ic a l R e g e n e r a t io n &E n v i r o n m e n t

H u m a n R e s o u r c e s P ro d u c t iv e E n v i r o n m e n t T e c h n ic a l A s s is t a n c e a n do th e r

P r i o r i t i e s

%

Source: European Commission

93 CEC (1991) op.cit.

65

A systematic comparison of the content of programmes, provided by a synthesis of ex post evaluation studies in 1996,94 indicates considerable uniformity in approach: “Objective 2 programmes generally had very similar overall aims and involved a similar range of measures to achieve them, notwithstanding the fact that circumstances varied considerably in the various areas”. However, as Figure 4.1 indicates, there are some clear contrasts between Member States in spending priorities over the 1989-93 period.

The Belgian, Danish, German, Italian and Dutch programmes prioritised support for productive investment. In Belgium, around two-thirds of programme support was allocated to SME business support and training schemes in the five Objective 2 areas focusing on the reconversion of areas dependent on the coal and steel sectors. Particularly high levels of ESF expenditure occurred in Limburg, reflecting major employment losses in mining. In Denmark, intervention in the two Objective 2 regions was concentrated on restructuring in the shipbuilding sector with the specific aim of promoting internationalisation; both the Nordjylland and Vestlolland programmes had some of the highest proportions of programme spending on training and research & technology development (RTD) among Objective 2 regions during this programme period.95 Most of the seven Objective 2 areas in Germany used business support measures to encourage structural diversification from various sectors – coal, steel, shipbuilding etc – by co-financing existing Federal or State (Land) SME and training schemes. Infrastructure measures were also prominent in the German programmes, ranging from flagship projects in Berlin (associated with reunification) to the rehabilitation of sites in Nordrhein-Westfalen, Bremen and Saarland.96

In Italy, all of the nine Objective 2 areas allocated a high proportion of programme expenditure to SME support (69 percent in the case of Marche), as well as general economic activity (Toscana, Lombardia, Lazio), RTD (Piemonte) and training. Some uniformity was again evident: “strategies do not differ significantly between the areas affected by declining large industries and those facing weak SME structures: in both, the main priority is the support and strengthening of SMEs”,97 with a strong preference for co-financing national and regional industrial policy grant schemes. The three Objective 2 programmes in the Netherlands had a similar orientation, dedicating some two-thirds of total expenditure to SME support and human resources. An apparent feature of the Dutch programmes was no direct Objective 2 spending on RTD or environmental improvement (with the exception of Zuid-Limburg where the rehabilitation of sites accounted for 12 percent of the programme).

A different strategic approach was adopted by the French, Spanish and UK programmes. These three countries – which collectively accounted for three-quarters of EU expenditure on Objective 2 – spent much less on SME support (and business development generally), allocating some two-thirds of resources to infrastructure and training. This also applied to the single Objective 2

94 Ernst & Young (1997) Ex Post Evaluation of the 1989-93 Objective 2 Programmes: Synthesis Report, DG XVI Series: Evaluation and Documents, No.4, Commission of the European Communities, Brussels. 95 Hall Aitken Associates (1996) op. cit. 96 Research voor Beleid (1996) op. cit. 97 ISMERI Europa (1996) op. cit.

66

programme for Luxembourg which devoted three-quarters of funding to infrastructure associated with the development of a Pôle Européen de Développement together with border regions in neighbouring countries.

In the case of France, there was a strong degree of commonality among the 17 Objective 2 programmes, most areas allocating c. two-thirds of expenditure to physical regeneration, infrastructure investment and ESF training programmes to reduce dependence on (mainly) coal and steel industries. Common priorities among the French programmes were the development of industrial and commercial zones and related services, support for local business potential through direct aid and training, and financing for the development of tourism. The main exceptions to the general pattern were Auvergne, Bourgogne, Bretagne and Pays de la Loire which instead prioritised business support (and technology assistance in the case of Pays de la Loire) over infrastructure investment.

On balance, the nine United Kingdom Objective 2 programmes had a similar allocation of expenditure, with about 60 percent of programme expenditure allocated to infrastructure projects, environmental improvement and training schemes. This concentration was particularly pronounced in Scotland where 83 percent of programme resources in Western Scotland and 70 percent in Eastern Scotland were dedicated to these priorities, much of the spending going to transport and communication projects and environmental measures to improve the image of the regions. Some of the less urbanised areas (Clwyd, West Cumbria) provided a higher proportion of business development support.

Finally, the CSF for Spain was distinctive in being developed as a single framework of financial commitments within which regional authorities could design their own programmes. Infrastructure investment accounted for 30 percent of national programme expenditure, but collectively the seven Objective 2 programmes had the most balanced spread of spending across the various priorities. While infrastructure projects took up almost half of the Cataluña programme (mainly spent on transport in Barcelona), business development support was significant in Aragón, Navarra, Pays Basque and La Rioja, and there was also considerable spending on training, environment and RTD.

Table 4.1: Distribution of Objective 2 funding by priorities (1989-93) Business Tourism Infra-

structure Training Environ-

ment Techno-logy

Other Total

MECU % MECU % MECU % MECU % MECU % MECU % MECU % MECU %

Belgium 77 28 16 6 34 12 97 34 17 6 36 13 3 1 281 100Denmark 3 14 0 0 5 21 8 38 0 0 6 27 0 0 22 100France 174 16 87 8 412 37 269 24 99 9 58 5 11 1 1109 100Germany 72 12 12 2 147 25 106 18 123 21 66 11 59 10 586 100Italy 94 24 38 10 66 17 99 25 65 17 29 7 1 0 391 100Luxemb. 1 8 0 0 9 75 1 4 1 12 0 0 0 0 12 100Netherlands 53 34 17 11 29 18 49 31 4 3 0 0 5 3 157 100Spain(1) 328 22 0 0 520 35 338 22 163 11 153 10 4 0 1506 100UK 139 6 231 11 687 32 494 23 241 11 88 4 282 13 2165 100EU-12 878 14 401 7 1828 30 1461 24 668 11 356 6 360 6 6063 100Source: compiled from Ernst & Young (1997) op. cit. (1) DG XVI figures incorporating multi-regional programmes.

67

Apart from the balance of spending across priorities, a further feature of the programming documents was the limited internal coherence of programmes. Most priorities were ‘mono-fund’, allocated to either ERDF or ESF, with little concrete programme management action proposed to achieve synergies between them. For the most part, training measures appeared to be planned separately from the rest of the programmes.

In summary, notwithstanding these deficiencies, the reform of the Funds led to significant improvements in the definition and implementation of Community regional development expenditure in industrial regions. Subsequent assessments and evaluation studies make clear that – however imperfectly - advances were made in both programming and partnership by comparison with the methods for allocating assistance used previously. Especially in countries such as Spain, Italy and the UK, the process of elaborating CSF and OP strategies provided a more robust framework for planning regional development than had existed before,98 also initiating progressive improvements in internal and external coherence (with other national/regional policies).99 In the Commission’s view, the process of plan submission and negotiation also enabled greater concentration on a more limited range of priorities, with a definite shift away from infrastructure investment to spending on measures promoting the creation and retention of jobs. In retrospect, the quality of plans, programmes and strategies was influenced by several factors.

First, the characteristics of the eligible Objective 2 area were important determinants of programme content. Strategic coherence in the definition and implementation of programmes increased when the boundaries of the Objective 2 region corresponded with an existing regional administrative unit. Also, local problems were more likely to be taken into account effectively in the operational programmes when the Objective 2 region represented an economically homogenous area.100

Second, the time factor was critical. All parties in the programme preparation process were working in a new regulatory environment under very tight time pressures. There was very little time available for the design of strategies, let alone reflection, regional analysis, consultation with local partners and appraisal.

Third, strategic coherence was related to pre-existing experience of regional planning. As the Commission subsequently reflected, “only those regions with experience of programming were able to draw up measures based on a development strategy acceptable to local agents”,101 although such experience carried with it the risk of a pre-existing template restricting the scope for exploiting Community assistance.

Finally, apart from time pressures, inexperience with the terminology and concepts of the new programming approach led to widespread disregard of the requirements for providing monitoring and evaluation data and undertaking a

98 Ernst & Young (1997) op. cit. 99 Quasar (1997a) op. cit. 100 CEC (1992) Second Annual Report on the Implementation of the Structural Funds, OOPEC, Luxembourg. 101 Ibid.

68

prior appraisal of programme impact. These were new methodological challenges for which experience, resources and systems were lacking.

4.2 Reprogramming of CSFs: 1992-93 Initially, Objective 2 areas had been designated and programmes submitted on the basis of a three-year programming period running from 1989 to 1991. However, the delays in starting up the programmes meant that programme implementation was compressed, and it became generally recognised that a three-year programming period was not compatible with the requirements of an operational industrial conversion programme. In April 1991, it was therefore decided to extend unchanged the list of 60 eligible Objective 2 areas for a further two-year period covering 1992 and 1993, with a view to making a ‘fresh start’ in Objective 2 programming in 1994 (along with Objectives 1 and 5b). The Member States were asked to update their Objective 2 conversion plans and priorities for the new CSFs for 1992-93. The aim was to ‘roll over’ the priorities of the 1989-91 period, adapting programmes to changes in the regional economic situation, concentrating resources on fewer sectors, and improving the integration of ERDF and ESF within programmes. Increased emphasis was to be given to job-creating investment rather than the provision of basic infrastructure, focusing programmes on promoting productive investment, SME development, environmental improvement, exploitation of tourism potential, assistance for RTD and cross-border co-operation measures.102

The new CSFs for 1992-93 were submitted and approved rapidly, followed by the presentation of most OPs in early 1992. As expected, most priorities and measures were a continuation of operations under the 1989-91 programmes (especially in France where the programmes presented in 1989 had been designed to run for five years). Changes partly reflected the Commission’s priorities but also experience with programming and the absorption of funding.

From the annual reports of the Commission103 and the ex post evaluation studies previously cited, several trends in reprogramming can be identified among the Objective 2 regions. First, programmes were adapted to new or emerging regional challenges, for example, greater emphasis on business adaptation to the Single European Market (the Netherlands) or military conversion measures in response to unification and the end of the ‘Cold War’ (Germany). Second, more stress was placed on assistance to the development of productive activities especially among small firms (Denmark, Spain, the Netherlands). Third, increased resources were channelled to research, education, innovation and technology, particularly to promote research-industry links, technological qualifications for entrepreneurs and technology transfer into small firms (Belgium, Denmark, France, Italy, Germany). Fourth, more effort was made to produce integrated programmes (or integrated priorities within programmes) by linking training with economic development measures, especially in Italy (where a multi-fund approach was adopted by all Objective 2 regions) and the Netherlands. Fifth, the environment and regional

102 CEC (1993) Third Annual Report on the Implementation of the Structural Funds, OOPEC, Luxembourg. 103 CEC (1994) Fourth Annual Report on the Implementation of the Structural Funds, OOPEC, Luxembourg.

69

image was promoted to a greater extent (Belgium, UK) and some innovative measures to encourage green tourism (the Netherlands) were introduced. Lastly, several countries/regions invested in technical assistance measures to improve the quality of information available to programme managers; examples included the establishment of a permanent observatory for assessing training needs (Aragón, Spain) and resources to promote better management of programmed measures (Piemonte, Italy).

4.3 Regional Development Plans and Programmes: 1994-96 In 1993, the Structural Fund regulations were revised. Under Objective 2, the eligible areas were redesignated to cover almost 17 percent of the Community population. The list of eligible areas was announced in January 1994 following the submission of lists of proposed areas from Member States. Whereas Objectives 1 and 5b were designated for a six-year period (1994-99), it was determined that regional eligibility under Objective 2 should run for two periods of three years in order to provide flexibility in adjusting to changing conditions of industrial restructuring. Apart from broadening the scope of the Structural Funds with new types of eligible measures (especially in the fields of research, science, technology, education and health), a key change to programming was a simplification of planning and decision-making procedures. In response to intense criticism from Member States regarding the complexity and bureaucracy of programme preparation, the amount of information and the level of detail required in programmes was reduced, and the three phases of programming (regional development plans, CSFs and OPs) were reduced to two. Member States were given the option of simultaneously presenting their development plans and draft programmes within Single Programming Documents (SPDs). Among other changes, a clearer commitment was made to conducting systematically ex ante, interim and ex post evaluation, and greater recognition was accorded to environmental issues.104

4.3.1 Preparing the Plans: 1994-96 In many designated areas, planning for the 1994-96 programmes had begun already in 1992, since the funding period was originally expected to commence earlier, but uncertainty over eligibility constrained many regions from undertaking detailed preparations. Consequently, as in 1988-89, the programming process was conducted within a highly compressed timescale. By comparison with the previous programming period, the development of the 1994-96 plans was more participative, responding both to analysis and wider viewpoints among key economic and social actors. They embodied much more careful preparation and a more open approach in response to a challenging set of regulatory requirements (see Figure 4.2). Nevertheless, evaluation evidence suggested that plan preparation still tended to be dominated by the favoured few. According to the Commission: “despite the experience and expertise acquired during the preceding period, several Member States are still reticent about full and open regional partnership”. In

104 Bachtler J and Michie R (1994a) Strengthening Economic and Social Cohesion? The Revision of the Structural Funds, Regional Studies, 28 (8), 789-796.

70

particular, the participation of economic and social partners was described as “timid”.105

Figure 4.2: Preparing regional development plans: 1994-96

105 CEC (1995) The New Regional Programmes under Objectives 1 and 2 of Community Structural Policies, COM (95) 111 final, 29.3.95.

Analysis of regional disparities and problems

Programme appraisal

Regional development strategy:

• = aims/objectives • = priorities/axes • = measures

Implementation arrangements

Consultation with partners

Previous programme experience

Existing national/EU policies

Use of individual funds

Additionality

Financial parameters

Consultation with partners

Quantification of objectives

Prior appraisal of impacts

Environmental assessment

71

The enhanced status given to appraisal and evaluation in the revised regulations provided the basis for a most systematic, comprehensive and detailed approach to evaluating regional development plans. The draft Objective 2 SPDs (only Spain kept the three-stage plan, CSF and OP system) were submitted in March/April 1994, and the documents were then subject to ex ante appraisals by the Commission (as discussed later in this report). The appraisals were tasked with assessing the rationale, coherence and content of the strategy, the quantification of objectives and the impact of the programme (in terms of aggregate and specific economic effects as well as environmental impact), the internal coherence of the programme with respect to ERDF/ESF integration, external coherence with other policies, and the implementation arrangements.106 These appraisals were used as part of the negotiations to develop the draft SPDs, to take account of Commission perspectives, in particular to ensure the inclusion of indicators of performance and impact, particularly in the area of employment effects.

The comprehensive nature of the appraisal process means that it is possible to gain some insight into the how the draft SPDs were perceived and the focus of the negotiations between Commission services and Member States. Based on comparisons of appraisal studies107 undertaken during 1994, the Commission concluded that the programming documentation had improved:108 “overall, the quality of the documents proposed by the Member States has been distinctly higher than in the preceding period, although the Commission has been a little disappointed in its expectations; this improved quality can be partly ascribed to the intensive preparatory work carried out by the Commission departments with the Member States, with a view to the preparation of the plans, particularly in the area of quantification of development disparities”.

However, several groups of weaknesses can be identified from the appraisal studies (see Figure 4.3).109 First, inadequate information was provided on the context for the strategies, especially with respect to basic socio-economic information on the local economy, analysis of strengths and weaknesses, and the linkages between the EU programme and other interventions. A common criticism of programmes was the weak analytical base to strategies with disconnected or superficial reviews of regional trends and problems, a poor understanding of how the local economy and labour market functions, and lack of sub-regional focus.110 In smaller regions, the designated area was not set in its wider geographical context or the broader policy environment. Establishing the contribution of EU support to economic growth or conversion within a region, therefore, was problematic. The lessons of previous programming experience were not always exploited, inhibiting understanding

106 Bachtler J and Michie R (1995) A New Era in EU Regional Policy Evaluation? The Appraisal of the Structural Funds, Regional Studies, 29 (8), 745-752. 107 For example: Bachtler J (Ed.) (1994) Ex Ante Appraisal of the Objective 2 ‘North Sea’ Region SPDs, Report to the European Commission (DG XVI), EPRC, University of Strathclyde, Glasgow. 108 CEC (1995) op. cit. 109 Bachtler J (1995) Ex Ante Evaluation of the Structural Funds: Experiences from Northern European Regions, Paper to the Conference The Evaluation of European Regional Policy, EC/MEANS, 16 November 1995, Brussels. 110 Bentley G and Shutt J (1997) European Regional Policy in English Regions, in J Bachtler and I Turok (Eds.) The Coherence of EU Regional Policy: Contrasting Perspectives on the Structural Funds, Jessica Kingsley Publishers, London, 123-142.

72

of the degree of innovation or continuity. Mis-timing of evaluation studies meant that interim or ex post evaluations were not always available for programme preparation.

Second, the internal coherence of strategies was questionable. SPDs sometimes appeared to be merely a framework for a collection of measures with little strategic direction or interrelationship between the higher order strategic objectives and proposed actions or measures. Strategies rarely embodied an explicit conceptual basis for achieving their planned objectives of economic growth, job creation etc, and causal links between objectives and outcomes were often asserted but not justified, exacerbated by the loose use of terms such as competitiveness. Linkages between regional analyses, strategies and outputs were frequently unclear or non-existent. Many documents were disappointing and bland; projects were being suggested which would have taken place without the EU support. There was a lack of solid ‘bottom-up’ underpinning of initiatives, a lack of well-thought action programmes, and often largely a repackaging of existing initiatives.111 Evaluation studies continued to cite the lack of internal coherence - the compartmentalisation of ESF human resources measures and the difficulty of integrating ERDF and ESF within strategies.112 Critics noted the tendency for strategies to be developed from the ‘inside outwards’ rather than starting with a broader macro perspective linked to the micro view of operational possibilities.113 It was also argued that more programmes needed to be based on a ‘generic regional strategy’ indicating strategic development of the region and co-ordinating different sectors, strategies and individual plans. The remit and division of competencies within implementation structures was not always specified, nor the way in which compliance with other EU policies (competition policy, equal opportunities, public procurement) would be enforced.

Third, the most difficult planning issue was that of quantification of objectives and the prior appraisal of impact. Indeed, under a strict interpretation of the regulations it is arguable that few SPDs complied fully with the regulations. Few programming documents provided an adequate range of indicators at each level, and many appeared unwilling or unable to provide sufficient prior appraisals. In many cases, the quantification information was only partial, and was provided in the form of output rather than impact indicators. In part, this reflected a strong feeling in certain Member States that detailed quantification and prior appraisal were inappropriate, irrelevant or impossible – either because of the strategic approach to implementation (competitive bidding) or because prior appraisal was considered to be methodologically flawed.114

111 Barrett G (1995) The Art of Putting Effective Programmes Together: An Appraiser’s View, Paper to SAUS Conference The ERDF: Recent Developments, School for Advanced Urban Studies, Bristol, 10-11 July 1995. 112 Taylor S (1997) Structural Fund Synergies: ERDF and ESF in Objective 2 Programmes, IQ-Net Thematic Paper, 2(2), EPRC, University of Strathclyde, Glasgow. 113 Roberts P and Hart T (1996) Regional Strategy and Partnership in European Programmes: Experience in Four UK Regions, Joseph Rowntree Foundation, York. 114 Bachtler J and Michie R (1994b) Planning, Appraisal and Negotiation: The Implementation of the Structural Funds, Report to EPRC Regional Research Consortium, EPRC, University of Strathclyde, Glasgow.

73

Figure 4.3: Appraisals of 1994-96 Programming Documents

Regional analysis ��insufficient information on regional socio-economic characteristics, especially non-labour market data ��absence of comparisons with national or international data eg. EU averages ��lack of sub-regional information and disaggregation of data ��inadequate analysis eg. lack of SWOT analyses Previous programme experience ��lack of adequate description of previous programmes experience ��inadequate reporting of results of internal and external evaluations ��absence of explicit linkage between previous programme experience and proposed operations Regional development strategy ��insufficient coherence within strategies; discontinuity between higher and lower order objectives ��lack of clear linkages between the regional analysis and strategy ��unclear linkages between priorities and development axes; insufficient use of multi-fund programmes ��inappropriate detail at the measure level ��inadequate information on the relationship between the strategy and other policies ��lack of ranking of priorities in terms of importance Quantification of objectives ��inadequate specification and quantification of target indicators ��lack of justification of indicators with respect to validity, applicability, data availability etc Prior appraisal ��absent or limited prior appraisal; qualitative not quantitative appraisal; aggregate figures only Environment ��partial or inadequate level of detail on the environmental situation ��insufficient linkage with specific directives and their applicability ��lack of environmental impact assessment ��no clear indication of the involvement of relevant environmental authorities Financial information ��insufficient detail on national sources of co-financing ��absence of information on EIB and other EU funding instruments ��limited justification of allocation of financial resources between priorities or sub-programmes ��lack of detail on the proposed use of the funds ie. the assignment of tasks to funds Policy coherence ��absence of information on enforcing compliance with EU policies eg. competition policy ��inadequate integration of plans with other EU policy support ��lack of clear linkages between structural actions and the Member State policies Additionality ��specification at the national rather than regional level ��no demonstration of stated compliance with additionality requirements Management/administration arrangements ��unclear divisions of competence within committee structures ��lack of information on the integration of specific groups eg. social partners

Lastly, many draft SPDs had difficulty with the environmental criteria in the regulations, as discussed in more detail in Chapter 7. Whereas most plans provided data on the environmental situation in their areas, the level of detail

74

was often partial or inadequate as a basis for policy. Environmental impacts were frequently not addressed, and in most cases there was no clear indication of the involvement of relevant environmental authorities either in plan formulation or in future implementation.

4.3.2 Content of Strategies: 1994-96 The balance of expenditure in the 1994-96 programmes shifted significantly away from the pattern of allocations over the 1989-93 period (see Figure 4.4). With a total volume of Structural Fund expenditure of 6,977 MECU for Objective 2 regions in 1994-99, the main focus was once again on the productive environment (productive investment), which accounted for 45.2 percent of Objective 2 allocations. Included under this priority are measures concerned with business and industrial competitiveness, focusing primarily on SMEs and measures which aid the process of economic diversification.115 This represents a significant rise in comparison with the level of expenditure on this area (40 percent) in the 1989-93 programming period.

Figure 4.4: Breakdown of Objective 2 funding by programme area (1994-96)

0

5

1 0

1 5

2 0

2 5

3 0

3 5

4 0

4 5

5 0

P h y s ic a l R e g e n e ra t io n &E n v iro n m e n t

H u m a n R e s o u rc e s P ro d u c t iv e E n v iro n m n e t T e c h n ic a l A s s is ta n c e a n do th e r

P r io r i t ie s

%

Source: European Commission

Countries allocating noticeably high levels of funding for these type of activities included Italy and, in particular, Spain. The high Spanish allocations on activities of this type were mainly accounted for by high levels of funding for infrastructure projects, which accounted for 27 percent of Spanish resources. Likewise, infrastructure projects accounted for a large proportion (23 percent) of the Belgian allocation. Support for the productive environment also covers SME projects, which accounted for a large proportion of expenditure under this category heading; Italy (22 percent), Spain (26 percent) and Germany (22 percent) all spent large proportions of their Objective 2 allocations on activities supporting the SME sector.

115 CEC (1995) op. cit.

75

Human resources was the other expenditure category with an increased expenditure allocation compared to the previous programming period. The increase was comparatively large, rising to 34 percent from the 1989-93 expenditure level of 21 percent. Allocation levels were fairly consistent across countries, not varying widely from the average, with the exception of Denmark, which allocated over half of its Objective 2 funding for training purposes (57 percent), most of which took place in the area of RTD.

Policies relating to physical regeneration and environmental protection showed a large reduction in expenditure allocation, but some countries still allocated relatively large proportions of funding to activities of this sort, notably Luxembourg (29 percent), Germany and France (18 percent each).

More detailed insights into the structure and content of strategies are available from comparative work undertaken by EPRC, which analysed the strategy and policy information in programme documentation for all Objective 2 areas.116 According to this analysis, the approach to programming in Objective 2 SPDs commonly proceeded through a hierarchy of strategic elements, beginning with an overall statement of strategic intent for the SPD as a whole, which was then progressively disaggregated or refined into strategic objectives, and translated into action through development priorities (or axes) and specific measures. The strategic approaches, as expressed in strategy statements, varied widely in terms of structure, substance and terminology and contained differing degrees of geographical and sectoral orientation.

Job creation/maintenance was the most common aim, mentioned in almost half of SPDs, followed by economic diversification, environmental protection, wealth creation and achievement of growth. Strategic aims were expressed in many different forms. A contrast can be drawn between those focusing on absolute improvements in certain socio-economic criteria such as employment, population and income, and those which instead sought a relative improvement in these criteria in relation to other parts of the EU. Other strategic statements focused on the conversion processes (eg. adaptation, diversification, modernisation) or the target priorities of the strategy (eg. human resources, RTD, services and tourism). A commitment to environmental sustainability in economic development was a complementary element of many of the overall strategic statements.

Repeating the experience of many 1989-93 strategies, most strategic aims could be applied to almost any programme and any Objective 2 area with relatively bland but universally relevant statements such as “to reindustrialise and regenerate the area” or “to improve the structure of the economy”. Among the exceptions were programmes for Basse-Normandie, Haute Normandie, West Berlin, Hessen and Arnhem-Nijmegen whose area-specific strategic aims gave the impression of having been developed from the bottom up ie. the strategic aim was intended to provide a summary statement of the whole programme. Several were highly focused, and two SPDs incorporated specifically quantified targets as objectives: “An annual rate of employment growth 0.3 percent higher than the national average for 1994-99” (Nord:Pas-

116 Bachtler J and Taylor S with Kearney C (1996) Extended Synthesis of Agreed Single Programming Documents in Objective 2 Areas, 1994-96, Report to the European Commission (DG XVI), EPRC, University of Strathclyde, Glasgow.

76

de-Calais); and “To create jobs (more than 1,750) and to increase income per capita (by 0.8%)” (Marches).

There are clear national cultures in the expression of strategic aims. The strategy statements of UK SPDs tend to be fuller - and in some cases more ambitious - than those of other countries, and several attempt to provide an aspirational or inspirational statement of common purpose, for example: “to achieve a dynamic, innovative and competitive regional economy where people and businesses are well equipped to face the challenges of new and emerging European and international markets” (Eastern Scotland). In turn, the Dutch SPDs cite wealth creation and sustainable development: most of the German strategies focus on diversification; and many of the French and Italian SPDs give primacy to the generation of employment opportunities.

At a second tier within strategic hierarchies, the overall aim was generally sub-divided into ‘strategic objectives’ which added detail to the overall statement of strategic intent, describing more fully what needed to be achieved. It is at the strategic objective level that aims usually became more specific to the region.

Almost all Objective 2 programmes had clearly presented, explicit strategic objectives On average, the SPDs each contained four strategic objectives, although the numbers ranged from one to ten per programme. The number of strategic objectives did not necessarily relate to the size of the programme, but was instead usually more a measure of the complexity or number of dimensions of a programme. Thus, whereas the programme for Alsace was allocated 19 MECU in Structural Funds and had nine strategic objectives, Rhône-Alpes received an allocation of 99 MECU but had just three strategic objectives.

The 1994-96 round of SPDs saw a new trend in the expression of priorities. The traditional approach to organising measures was to employ descriptive priority headings such as ‘business development’ or ‘training’, grouping measures of the same type. In 1994, this approach began to be superseded by the use of priority headings which grouped measures in different fields by their objective. For example, the priority heading ‘enabling the growth of the logistics sector’ might encompass measures proposing basic infrastructure improvements, the development of logistics firms and relevant training opportunities. Two-thirds of programmes expressed some or all of their priorities in this way, an approach giving primacy to the purpose of measures and lending additional focus to the project application and selection process.

In terms of programme structure at the priority and measure levels, the SPDs for the Italian and UK regions appeared to be more complex, containing a higher number of priorities (as well as measures), than in countries such as Belgium and the Netherlands. In part, this was related to the resources available, with the larger UK programmes offering more potential for a higher number of economic development themes to be pursued. The approach taken by different Member States to defining measures also has an impact: Italian programmes disaggregated activities which in other countries might be interpreted as similar enough to be placed under the same measure heading, and this increased the number of measures in comparison with other countries’

77

programmes. (In Piemonte, programme size and national approach combined to produce a programme consisting of over 30 separate measures.)

Regions that were eligible for Objective 2 funding in the 1989-93 programming round generally maintained strong continuity in the overall direction and substance of their programmes. Two-thirds of the SPDs displayed a strong connection between the two programming periods. Nevertheless, there were clear shifts in the emphasis given to different priorities, some of which were the consequence of experience (both negative and positive) during the previous programming period, and some of which emerged through the negotiation phase with the Commission. Identifiable trends indicated increased attention being given to business development, RTD, technology transfer and environmental issues, mainly at the expense of investment in economic and transport infrastructure.

At the priority level, the majority of programmes contained some sectoral targeting. This was particularly explicit in the case of the Dutch strategies which identified key industries as a focus for the priorities: transport and distribution (logistics), producer services, tourism. Many of the UK strategies also contained sectoral priorities, sometimes cast as ‘drivers for change’ (West Midlands). By contrast, the French and Italian SPDs had little or no explicit sectoral targeting.

Although many of the Objective 2 areas are highly heterogeneous regions, and in some cases comprise geographically discrete sub-areas, relatively few of the SPD strategies contained a spatial dimension. Only in the UK was there a fairly consistent geographical orientation incorporated into some of the priorities. Here, the focus of targeting was on need rather than opportunity, with additional resources being directed at the areas of greatest disadvantage.

Figure 4.5: Objective 2 allocations by action area (1994-96)

611

497 495

346

264 251

199

147 143

431

383

276

111 101

215

149

23

92

34

0

100

200

300

400

500

600

700

Hum

an re

sour

ces

Busi

ness

deve

lopm

ent

Econ

omic

infra

stru

ctur

e

R&T

D

Reg

ener

atio

n

Basi

c in

frast

ruct

ure

Tour

ism

Com

mun

ityec

onom

icde

velo

pmen

t

Envi

ronm

ent

Tech

nica

l ass

ista

nce

Stru

ctur

al F

unds

allo

catio

n - M

ECU

per

yea

r Allocation per year (MECU) -excluding 'overlap' measures

Allocation per year (MECU) -including 'overlap' measures

Source: Bachtler J and Taylor S with Kearney C (1997) op. cit.

78

At the lowest level within the strategic hierarchy of an SPD are the measures through which expenditure will be committed. It is only at this level that the true substance of the programme becomes apparent.

Human resource development was a strategic element of all the Objective 2 programmes, accounting for 26 percent of the total Structural Funds allocation (1,500 MECU), with most programmes dedicating between 15 and 35 percent of their programme resources to this area. Notably, training was better integrated into economic development strategies (at a superficial level at least) rather than being presented as an independent strand of activity. In the Plymouth programme, for instance, training was an element of all four priorities, complementing industry and business development, the development of knowledge-based industries, tourism and community economic development. The main exceptions to the integrated approach were the smallest programmes, where human resources tended to be presented as a separate priority, and the German SPDs (with the exception of West Berlin) and Spanish OPs which listed ERDF and ESF measures as separate priority headings administered by separate organisations. In a further trend, training was treated less amorphously than in the past, focusing on particular areas such as training or re-training of the employed workforce, promoting the use of training infrastructure and targeted human resource development - for tourism, innovation, inward investment and the environment.

Business development through direct aid to firms received 21 percent of Objective 2 support (1,225 MECU - excluding training support). There were distinct national differences among the Objective 2 programmes: strong concentration on aid to firms in both Denmark and Italy (40-55 percent of programme expenditure) compared to only 10 percent of German programme support. The focus was predominantly on enhancing indigenous growth potential through a wide range of opportunities, especially targeting SMEs eg. consultancy and financial support for technology transfer, new product development, marketing initiatives and environmental projects (in areas such as waste management and energy efficiency), and advice on joint ventures, marketing and modern management approaches. Other foci included the development of specific sectors, the establishment of networks and support to firms of strategic importance, including the embedding of inward investors into their host region’s economy. Proposed measures were increasingly shaped by the specific development needs of sub-groups of firms within any given locality. In Emilia Romagna, for instance, one priority was dedicated to new firm formation and another to the development or reconversion of existing businesses; Lombardia took a similar approach, targeting the reinforcement of existing businesses with growth potential separately from the reconversion of firms in declining sectors. North East England developed the approach in a more comprehensive way, tailoring its business development activities to four main sub-groups of companies: strategic businesses and inward investors, SMEs, firms in knowledge-based industries and companies in the cultural and tourism sectors.

Support for economic infrastructure was the third largest category of Objective 2 expenditure, accounting for 17 percent of total spending (1,002 MECU), again, a key element of almost all Objective 2 strategies. Measures in this category included the development or improvement of the ‘business

79

environment’ including sites and premises for economic activity. In most programmes, these activities played a secondary, complementary role to more central business development or human resource actions. Exceptions included Schleswig-Holstein, where the restructuring of obsolete, industrial and port-related waterfront sites was fundamental to a regional modernisation strategy focusing on development of the producer services sector through provision of prestige property facilities.

Research, technological development and innovation measures accounted for some 13 percent of the Objective 2 allocation (784 MECU). With over 10 percent of the Structural Funds allocation in over half of programmes, exceptionally high percentage allocations to research and development (20-40 percent) were made in Zuidoost Brabant, Nord-Pas-de-Calais, Auvergne, Nordjylland, Emilia Romagna, West Berlin and Arnhem-Nijmegen. Typical actions were the promotion of increased co-operation between business and research and development facilities, and the reorientation of existing facilities to serve better the innovatory requirements of business. Other common forms of support included finance for the development of new products or processes and enterprise aid to develop inter-firm innovation networks. In the field of RTD training, the most frequent measure involved advanced training in technology and innovation, often for already highly skilled staff. A growing area of research support was environmental technology, notably in the UK (East Midlands, South Wales, North East England, Plymouth, West Midlands, West Cumbria and Furness) and Belgium (Limburg and Turnhout). An interesting example was Groningen-Drenthe (Netherlands) where a key element of the regional strategy was the further development of the so-called ‘MERA’ cluster - an existing group of firms in the environmental, waste, energy and recycling sectors - through support for the implementation of new technologies.

Development of the tourism sector attracted almost 10 percent of the Objective 2 Structural Funds allocation (572 MECU), and appeared in four out of five strategies. Six (small) programmes dedicated more than a fifth of their programme to tourism - Gibraltar, Val d’Aosta, Umbria, Plymouth, Marches and West Cumbria - allocations for the three years ranging from 5 to 35 MECU. In some cases, the development of the tourism industry was a priority in its own right, very often combined with the development of related cultural industries, for example in Toscana. In others, such as Midi-Pyrénées, it was among a range of industries targeted to assist economic diversification. Urban regeneration, with some 9.6 percent of the Objective 2 total (564 MECU), included actions to improve the physical environment and remove the traces of former industry deterring investment and qualified people from remaining in an area. Urban regeneration or requalification du territoire was most frequent in the French SPDs, with 17 out of 19 programme areas involved. It was present, less prominently, in all the Belgian programmes, and two-thirds of Italian SPDs.

Physical infrastructure (mostly for transport) formed just 7 percent of the total Structural Funds allocation (414 MECU), a significant reduction by comparison with the previous programming periods and reflecting changed Commission priorities. Over half of physical infrastructure spending was in

80

the UK, with a high level of expenditure also in the Netherlands. Outside the UK, where spending was primarily directed at intra-regional public transport links, the focus of transport infrastructure measures was on the improved exploitation of existing networks eg. through the development of multi-modal transport nodes such as terminals (Twente, Toscana) or regional transport interchanges (Friuli-Venezia-Giulia). Community economic development measures accounted for 6.4 percent of the Structural Funds (374 MECU), but were concentrated on less than one-quarter of the Objective 2 programmes. They were a particular feature of the UK regional programmes, and, to a lesser extent, those in France. They involved integrated economic and social regeneration measures targeting additional resources to specific geographic zones of the greatest economic and social disadvantage in order to decrease intra-regional social and economic disparities.

Lastly, environmental concerns were integrated into Objective 2 strategies in two major ways. First, every eligible area was required to incorporate the environment as a horizontal dimension of strategies, with an assessment of the regional environmental situation as a part of the regional profile, consideration for the potential environmental impact of measures and the inclusion of relevant environmental criteria to guide project selection. Second, two-thirds of programmes defined specific environment-related measures. Such measures only formed more than 15 percent of the programme in four cases: Picardie, West Berlin, Saarland and Luxembourg. The relevant measures were in two categories: those addressing the legacies of the past in order to assist the process of reconversion (eg. through physical regeneration in Luxembourg); and forward-looking measures, seeking to exploit ‘eco’ products, services and technologies as a potential source of competitive advantage and so economic regeneration (as in Saarland).

In summary, taking forward the conclusions of earlier programming periods, the process of formulating programmes gave an impetus to strategic planning and development at regional level, supported by extensive consultation of regional institutions and other partners. The approach to strategic planning in Objective 2 areas varied widely in terms of structure and substance. It reflected, in particular, the various traditions of development planning across the EU and the extent to which there were established regional institutions capable of exercising or leading a strategic development function. Nevertheless, it is evident that many Objective 2 areas utilised the EU development planning process to undertake a serious and systematic assessment of regional development needs and strategic priorities.

Many regions responded effectively to the requirements of Objective 2 programming. It is not only the most experienced which performed well, although long experience, as in Western Scotland, clearly had the potential to impact on the complexity of strategies and the effectiveness of their implementation arrangements. Some of the regions which were newly designated in 1993 also performed well in strategy and programme development, bringing fresh approaches. The resulting plans comprised, in the best instances, coherent, focused and area-specific strategies grounded in a detailed analysis of the local economy, and with identification of both

81

strengths and weaknesses. Clear and concise ‘mission statements’ provided a unifying role for organisations and individuals working within an area.

A further feature is that many of the strategies embodied considerable complexity and sophistication. Compared to the previous programming round, the 1994-96 strategies covered a wider range of priorities, with increased attention being given to business development, RTD (often on an experimental basis), technology transfer and environmental issues, mainly at the expense of physical and economic infrastructure projects. The various types of action area (eg. business development, training, economic infrastructure) appeared to be better integrated than previously, with measures being combined to enable the achievement of specific ends - for example, an increase in the regional innovation capacity through support for relevant training, technology transfer mechanisms and consultancy initiatives targeted at business. A further aspect of enhanced integration was the effort to improve synergies through the fuller co-ordination of ESF and ERDF actions.

The sophistication of strategies was also reflected in the targeting of measures. Training was treated less amorphously than in the past, with specific training measures tailored to the requirements of the employed workforce, specific sectors such as tourism, specific fields such as environmental skills and RTD, and the reintegration into the workforce of the unemployed or disadvantaged. Business development measures distinguished between the needs of new entrepreneurs, craft firms, firms in declining sectors but with potential for reorientation, enterprises with strong growth potential and inward investors. Some strategies contained a strong sectoral orientation, sometimes cast as ‘drivers for change’, whereby key industries were identified as a focus for priorities - logistics, knowledge-based industries, producer services, tourism, environmental products and services, and high-technology sectors. The heterogeneity of some Objective 2 areas increased the pressure to incorporate a spatial dimension within strategies. This was particularly apparent in areas allocating resources to community economic development where specific geographic zones of severe economic and social disadvantage were targeted with additional resources in the form of ‘sub-strategies’.

Lastly, the range of Objective 2 strategies collectively embodied a significant innovation. This included the presentation of strategies (eg. demonstrating the inter-relationship between regional problems and actions), the design of organisational mechanisms for co-ordinating the administration of different Funds or representing the interests of partners, and examples of innovative types of measures, especially in the environmental field (eg. environment-related training, exploiting ‘eco’ business opportunities and improving the environmental performance of enterprises). Comparative analysis of the Objective 2 strategies also reveals several questionable aspects of regional development programmes.117 First, whilst most programmes appeared to be appropriate for the regional problems prevailing in their areas, strategies rarely had an explicit conceptual basis for achieving their planned objectives of economic growth, job creation etc. Some strategy documents lacked a clear rationale, and contained discontinuities between higher order strategic objectives and proposed actions

117 Bachtler J (1995) op. cit.

82

or measures. In addition, causal links between objectives and outcomes were often asserted but not justified, exacerbated by the loose use of terms such as competitiveness.

Second, it was difficult to establish the coherence of EU-funded strategies with the broader policy context. The relationship or linkages between EU, national, regional and local economic development policies were sometimes obscure or even contradictory, and the degree of real partnership (as required by the Structural Fund regulations) in implementation structures could be suspect. This problem was compounded by the small size but diverse coverage of many programmes, Structural Fund expenditure being spread over a wide range of ‘action areas’ and measures. Inevitably, there were problems in determining the anticipated impact of such programmes; prior appraisals rarely provided meaningful projections of the impact of planned interventions.118

Third, programme designers appeared to have faced problems in effectively incorporating the environment as an integral part of strategies. The majority of strategies provided data on the environmental situation in their area, but often this description was not adequately integrated into the strategy proper. Further, the potential environmental impacts of measures were often not satisfactorily addressed. However, in this field at least, experience has varied between programmes, with some potentially able to provide effective approaches.

Finally, it was evident that some strategic documents continued to be little more than a ‘paper exercise’ designed to legitimise EU co-financing of predetermined national or regional projects and schemes that had little common purpose.

4.4 Reprogramming: 1997-99 When the Objective 2 programming periods were approved in 1993 it had been anticipated that due to the transition process in industrial regions there may be a necessity to make adjustments to the areas covered by the programmes after three years. A few very minor changes did take place. As mentioned in Chapter 1, they involved some urban areas of Spain (Zaragoza, Barcelona, Palma de Mallorca and Madrid), Italy (Pisa, Milan, Varese, Reggio Emilia, Modena, Rome, Trieste, Florence and Venice) and the Netherlands (Overig Groningen), although the level of population coverage under Objective 2 remained as it was during 1994-96 within each Member State.

Preparatory work for reprogramming was undertaken during 1996. Draft programmes were submitted to the Commission between early September and December 1996 and approved for the most part between March and August 1997.

During 1996, in the preparatory phase, the Commission circulated guidelines relating to ‘horizontal themes’ which, it recommended, should receive

118 Bachtler J and Michie R (1995) op. cit.

83

emphasis in the 1997-99 programmes.119 Member State authorities and Programme Monitoring Committees were encouraged to promote greater thematic concentration on the objectives of: employment creation; RTD; environmental concerns; equal opportunities; and entrepreneurship.

This thematic focus stems, in part, from a series of European Council decisions on the role of EU policies in fostering Community competitiveness. The Essen Council in 1994 announced that combating unemployment and promoting equal opportunities is the priority task of the Community and its Member States, a principle reaffirmed at subsequent Summits. In the context of the 1993 White Paper, the relationship between economic growth and employment was also explicitly addressed at Council level, highlighting the role played by SMEs as a source of jobs, growth and competitiveness, especially through better access to information, training and RTD, the promotion of internationalisation and better access to capital markets. Sustainability in economic development has been increasingly emphasised since the mid-1980s, partly through global initiatives such as the 1992 UN Conference and the successive EU Environmental Action Plans, progressively promoting environmental integration in all EU policies.

Several programmes undertook extensive early preparatory work for the 1997-99 programmes in order to avoid the problems which had been experienced in previous rounds because of compressed programme development timetables. Many ensured that their interim evaluation studies would be completed to coincide with the preparation of the new programmes, ensuring the results could be fed into them. Aquitaine, Nordjylland and Nordrhein-Wesfalen began interim evaluations in early 1996, resulting in their timely completion for contributing to the new programmes. Not all of the regions achieved this degree of efficiency. Nevertheless, improvements in the programming documents were evident in the 1997-99 round. All the regions had been involved in at least one round of programming, so experience had been gained. The process of plan preparation in the 1997-99 programmes became increasingly interactive, involving a wider range of partners in many regions. Among the results were that the analysis, consultation and feedback process amongst the partners improved the quality of programme documentation in many regions and also increased local ‘ownership’ of programmes.

As in previous programming rounds, the experience of negotiation was highly variable. For many regions, the negotiations seem to have run relatively smoothly, focusing more on technical than substantive issues with limited change involved in the majority of programmes. In several Member States (eg. France), this represented a major contrast to the 1994 programme negotiations which were extremely difficult - once memorably described as a “dialogue between the deaf”. This may be attributable to the fact that reprogramming frequently involved a rollover of existing programmes as well as sensitivity among Commission services to the intense criticism they received during the previous negotiations.

119 CEC (1996) Note for Guidance concerning Operations in the Declining Industrial Areas (Objective 2) for the Second Programming Period 1997-1999, Commission of the European Communities, Brussels.

84

This relatively positive experience was not universal. A frequent criticism concerned the erratic schedule of the negotiations, characterised by poor timetabling and communication, and intermittent bursts of frenetic activity. Commission services again appear to have been overloaded, unable to deal with the workload associated with the volume of programmes in a short space of time. The insistence on including additional commitments to equal opportunities and the environment was not always welcome. The Commission’s agenda was sometimes unclear: in some cases (eg. Nordjylland), the Commission would not allow the region to change its basic structure of priorities and measures, even to facilitate inclusion of horizontal themes such as equal opportunities and the environment. By contrast, in Western Scotland, the Commission advocated a strategic reorientation of infrastructure development to improve the integration of the poorest areas within the regions; the question of ‘strategic sites’ evolved into a major dispute with the Commission requiring the designation of five growth poles on which development was to be targeted, but which was considered locally to be politically unacceptable and economically risky (and subsequently rejected).

A further criticism of the 1997-99 negotiations was that the process in some cases took insufficient account of the increasingly inclusive nature of programme development at the regional level. Partners have become more experienced in and committed to participation in the programme design phase. In some cases, draft programmes which had been elaborated in partnership were fundamentally changed during the final negotiations undertaken between a restricted group of key players. Where this is accompanied by insufficient communication and transparency, it risks the collaboration of horizontal partners being perceived as ‘token’, undermining partner trust and goodwill and leading to problems at the implementation stage. More sensitive management of the negotiation phase, tailored to the partnership context, has consequently become crucial.

With respect to programme content, the 8,147 MECU (at 1997 prices) dedicated to Objective 2 programmes in the 1997-99 programming period represented a real increase of 13.8 percent in comparison with the previous allocation, even though the level of population coverage remained at 16.4 percent.120

Figure 4.6 provides an overview of the shift in the level of allocated resources between four distinguishable priorities receiving support. Throughout the three programming periods between 1989 and 1999, there have been distinct changes in the prioritisation of support areas. The most dramatic changes have taken place in the physical regeneration and environment category, which after a massive decrease in allocation between 1989-93 and 1994-96, fell slightly for 1997-99. The largest expenditure shifts took place between the 1989-93 and 1994-96 programming periods, reflecting changing attitudes towards priorities to be supported and the learning process involved, whereas the 1997-99 programmes appeared as a broad continuation of the previous programming round with only slight adjustments in expenditure allocations. The

120 CEC (1997) The New Regional Programmes 1997-1999 under Objective 2 of the Community’s Structural Policies – Focusing on Job Creation, Commission of the European Communities, COM (97) 524 final 14.11.97, Brussels.

85

programmes remained substantively the same. The limited change reflects the fact that strategies were felt to be working well, and that the Commission’s main focus was on future reforms, preferring a rapid revision of the existing strategies in most cases.

Figure 4.6: Comparison of Objective 2 funding by programme area (1989-93, 1994-96, 1997-99)

0

5

1 0

1 5

2 0

2 5

3 0

3 5

4 0

4 5

5 0

P h y s ic a l R e g e n e ra t io n &E n v iro n m e n t

H u m a n R e s o u rc e s P ro d u c t iv e E n v iro n m e n t T e c h n ic a l A s s is ta n c e a n do th e r

P r io r i t ie s

%

1 9 8 9 -9 31 9 9 4 -9 61 9 9 7 -9 9

Source: European Commission (various documents), Note: 1989-93 figures represent actual expenditure, 1994-96 and 1997-99 figures represent allocations.

Within these expenditure categories there were fairly noticeable changes between countries. Although expenditure allocations on the productive environment rose only slightly, supporting infrastructure allocations changed significantly in some countries. Proportionally, Denmark overtook Spain on its infrastructure allocations. Whereas Spain in the previous programming period spent 27 percent of its allocations on infrastructure activities compared to 16 percent in Denmark, this changed to 14 percent for Spain and 25 percent for Denmark in the 1997-99 programme. In Germany, infrastructure projects also rose from receiving a 7 percent allocation of resources in the previous programmes to 22 percent. Overall, contrary to Commission aims, infrastructure allocations rose from 12.5 percent to 13.4 percent.

In the training category, the largest rise was in RTD allocations which rose from 9.9 percent to 13.4 percent. This was an area included as one of the five horizontal priorities to be included in the programmes and the rise in allocations reflects this increased importance. The highest national allocation to this area was in Finland where 27 percent of its total Objective 2 allocations were dedicated to RTD, but the largest increases were in Spain and Belgium where the figures rose from 5 and 11 percent respectively in 1994-96 to 17 and 18 percent in 1997-99.

SMEs and indigenous development were once again the focus for many of the programmes. In Belgium, allocations rose from 8 to 10 percent. However, overall SME allocations dropped from 16.7 percent to 15.8 percent. German

86

allocations, for example, dedicated to this sector dropped from 22 to 7 percent. In terms of the targeting of measures, as with the previous round of funding, they were typically aimed at SMEs and developing the local potential of regions. This round of programming saw not only more measures included in programmes, but also a wider range of measures.121 Regions appeared to be more willing to promote a wider range of activities to create economic diversification, potentially with a loss of synergy between actions.

A comparison of programming documentation for Objective 2 regions for the 1994-96 and 1997-99 periods122 yields several impressions. As already noted, the new Objective 2 programmes were of higher quality, on paper at least, than their predecessors, a view shared by Member State authorities and Commission services. Furthermore, although the structure and organisation of programmes may have been modified, the strategies appear to have remained substantively the same. There was considerable continuity in the strategic orientation between the 1994-96 and 1997-99 periods. In Belgium, for example, the 1997-99 programmes were for the most part a continuation of the previous programmes. Small adjustments took place, such as in Liège where allocation towards productive investments rose. The most significant changes in the programme strategies was at measure level, as the new programmes contained increasing numbers of measures, covering a wider range of actions.

Specifically, several issues are evident. First, presentation of programming documents improved, with more systematic and in-depth socio-economic analyses (including SWOT analysis), clearer presentation of strategies with focused objectives, fuller quantification of indicators, specification of targets and use of benchmarking and demonstrable synergies between programme elements. Programme managers and partners evidently had a better understanding of the ‘rules of the game’. However, Commission documentation makes it clear that, “there was still scope for improvements in linking this instrument to the translation of the strategies into priorities and measures”.123

Second, the core strategic aims and objectives and expenditure allocations did not change markedly. Reflecting the trend since 1989, the emphasis on employment creation has continued to increase, such as in the Netherlands, where employment actions received increased attention in the new programmes. Regional and corporate competitiveness was also mentioned more frequently. Major changes were attributable to significant shifts in national policy, as in France where there was a greater focus on defence-related and urban problems.

Third, many regions took the opportunity to simplify or rationalise the layout and structure of the 1997-99 programmes. There were numerous cases of priorities being merged, reduced or restructured. The trend, already evident in 1994, of moving away from generic priorities (eg. support for small business), towards more sophisticated thematic priorities, combining targets such as RTD, SMEs and human resources under one heading, is continuing. The links

121 CEC (1997) op. cit. 122 Bachtler J and Michie R (1997), The Evolution of Objective 2 Programmes, IQ-Net Thematic Paper, Series 3, No. 1, EPRC, University of Strathclyde, Glasgow. 123 CEC (1997) op. cit.

87

between programme components - objectives, priorities, measures - were made more explicit; the relationship of individual measures to overall programme goals was demonstrated more clearly. ESF measures were better defined, enhancing the management and coordination of training actions.

Finally, reflecting Commission pressure, environment and equal opportunities goals were more prevalent within programming documentation. Commitment to equal opportunities ranged from assessment of female employment issues within the socio-economic analysis, to inclusion of equal opportunities within programme objectives, to specific measures or actions targeted at women. The approach to environmental issues was similar, although there appeared to be more examples of genuine good practice such as detailed prior appraisals of environmental impact and the involvement of environmental authorities (and criteria) in project-level decision-making and monitoring. The Italian programmes were among those which took the environment and equal opportunities more into consideration in the 1997-99 programmes. Among other ‘horizontal themes’, almost one-quarter of Objective 2 regions appear to be using RIS/RITTS strategies to improve the strategic approach to RTD promotion, although there is wide variation in the degree of integration with mainline programmes.

4.5 Conclusions and Recommendations The presentation of regional development plans and Objective 2 strategies has progressed significantly over the past decade. Early plans had little or no strategic coherence, being viewed mainly as mechanisms for accessing EU funding. Analyses of stated needs were basic, and requirements for quantification or prior appraisal were neglected or ignored. Development objectives and axes were frequently general and vague and programmes sometimes seemed to be little more than selections of eligible projects.

Successive phases of programme preparation have seen a progressive increase in sophistication among Objective 2 strategies. Some of the best examples have systematic and in-depth socio-economic analyses (including SWOT assessments), and embody a clear, unifying mission statement, quantified objectives, specification of targets, use of benchmarking and a coherent set of measures, targeted at specific economic development problems and related to other national and regional initiatives, collectively supported by a monitoring system capable of determining physical as well as financial outputs and impacts. The links between programme elements – objectives, priorities, measures – are becoming more explicit in the presentation of strategies, with the relationship of individual measures to overall programme goals being demonstrated more clearly.

The Funds have clearly generated greater interest in strategic thinking and planning when this was on the decline or did not exist in some countries. This has been of particular relevance to Objective 2 regions, which typically have a greater number of agencies or initiatives operating across a range of sectors. The programmes have also widened the range of organisations able to exert an influence on economic development - now in many cases including the voluntary sector and trades unions - in economic development thinking and decision-making. The Funds have required a more integrated approach to

88

planning and implementing projects than has been typical in the field of economic development.

One of the major impacts of the Structural Funds in Objective 2 regions has been to encourage dynamism and innovation in economic development. Examples of new ideas being disseminated through EU programmes include regional technology policies, community economic development and integrated environmental and economic measures. Perhaps the most important innovation promoted by the Structural Funds is in the field of community economic development. Some of the most deep-rooted economic and social problems lie in areas within urban/industrial agglomerations suffering from generational unemployment, poverty, and marginalisation or exclusion from labour markets and economic activity. Objective 2 interventions have been used for pilot programmes to enable excluded social groups to re-enter the labour market and to change attitudinal culture through community responses to skills development and local service provision.

Virtually all Objective 2 regions have seen a shift away from promoting ‘traditional’ economic development activities such as physical infrastructure towards softer, more innovative measures. In the RTDI field, for example, investment in RTDI infrastructure is being superseded and complemented by a range of imaginative measures to exploit RTDI facilities through advice and consultancy support, enterprise and employee training, environmental technology promotion, network building and awareness raising. The promotion of technology and innovation policies appropriate to the needs of firms which are neither technologically oriented nor innovative has filled a particularly significant gap in the policy framework of many areas undergoing industrial reconversion (see Chapter 6 below).

Reflecting Commission pressure, environment and equal opportunities are becoming more prevalent within programming documentation. Commitment to equal opportunities ranges from assessment of female employment opportunities within programme objectives, to specific measures or actions targeted at women (see Chapter 8 below). A shift towards integration of economic development and environment has been most necessary - and most evident - in Objective 2 areas, where a clear momentum has been building up to strengthen the environmental dimension of SPDs as a horizontal rather than compartmentalised feature of programme design and implementation (see Chapter 7 below).

Given the significance of the programme development stage in establishing the foundations for successful programming, there is still considerable scope for better programme preparation and strategy development, building on the improvements that have been made already, and for the wider dissemination of established good practice.

89

4.5.1 Programme Development Recommendations

In developing their programmes, regions should be encouraged to exploit the next programming period to create meaningful regional development strategies with scope for experimentation and innovation.

While there have been long-term shifts in expenditure allocation among Objective 2 programmes, there is generally considerable continuity between successive programme periods. This is likely to be the case again in reprogramming from 1997-99 to 2000-06, but there is a strong case to promote new thinking and policy directions.

For many Objective 2 regions, 2000-06 may be the last programming period in which they receive substantial resources under the Structural Funds. They should therefore maximise the opportunities that this period presents to experiment with new strategic ideas for economic development. As this study has shown, an enormous reservoir of programming experience has been built up in Objective 2 regions over the past decade; more partners than ever before are engaged as experienced actors in different parts of the programme development process. It is important that regions harness this expertise and experience, and that they are prepared to initiate new or improved policy orientations. At the very least, the next programming period should be used for experimentation and testing new initiatives that may contribute to more effective strategies.

In most Member States, there is considerable potential for enhancing the strategic orientation of programmes. Indeed, in several countries, Objective 2 strategies still exist only on paper with little meaningful co-ordination or integration between priorities and measures in practice. Building on the progress that has been made in some regions in some countries, efforts should be made to widen the use of the SPD as a robust and coherent framework for planning regional development.

In terms of specific policy orientations, the conditions under which grants can be allocated to enterprises should be made more demanding, particularly by restricting the passive channelling of financial support to companies. In addition, the more widespread use of financial engineering methods should help to improve the focus and value for money of these initiatives. In parallel, resources should continue to be allocated to ‘business environment’ measures – notably in research, education, innovation and technology – experimenting with new and unconventional measures that could overcome obstacles or socio-cultural resistance to diversification and change. These measures can be more durable and of much wider benefit that initiatives undertaken within single firms.

90

The most challenging goal is for strategy development to be undertaken in a more holistic fashion, whereby the various elements of the strategy are developed as an integrated whole.

Currently there is a widespread tendency to assemble Objective 2 strategies as a collection of ‘prefabricated units’, with the regional analysis, assessment of previous programme experience, strategy, environmental profile, prior appraisal, etc, being prepared separately and then collated within a programming document. Better interconnection between these building blocks, from the ‘bottom up’, would ensure an improved linkage between the analytical base and the strategy, the integration of sustainability and gender issues, and synergies between the Funds. Strategy development also needs to give more consideration to the conceptual basis for objectives and outcomes. Four aspects of programme design are particularly important.

��First, regional analysis requires comprehensive information on regional socio-economic characteristics, especially labour market data (including information on equal opportunities). Comparisons should be made with national and EU averages, with trend as well as current data. Where relevant, information should be disaggregated to indicate sub-regional characteristics and special problems. Regional data should be analysed adequately, making use of techniques such as SWOT analysis to provide a reasoned and meaningful foundation for the regional development strategy.

��Second, a review of previous programming experience should describe the recent history of economic development interventions under Structural Funds and other relevant programmes in the area, together with review and assessment of the results of internal and external evaluations. In particular, there should be an explicit link between previous programme experience and the proposed operations.

��Third, a regional development strategy should, above all, be coherent. Building on the regional analysis and previous programming experience, it should present a structured hierarchy of aims, objectives and priorities with clear linkages between the development axes. The strategy should present some conceptual view as to how it will impact on economic development in the area. It should also contain links between structural actions and other Member State policies and provide some ranking of the importance of objectives and priorities. Objectives and actions should be fully quantified, with detailed specification of results, outcomes and impacts – including indicators, targets, data sources and data collection methods. Prior appraisals should have both quantitative and qualitative dimensions. Sustainability and gender mainstreaming should respect the guidelines, in each case aiming for the ‘third order’ levels of integration (described in Chapters 7 and 8). For the strategy, this requires: detailed information on the current situation; links with specific relevant laws, regulations and directives and their applicability; impact assessment (using EIA and SEA tools in the environmental case); and a clear indication of the involvement of relevant authorities to design, implement and monitor the commitments.

91

��Lastly, the implementation arrangements should be adequately described, indicating how the plan and strategy has been constructed and the involvement of different partner interests and how the programme will be managed and administered, stating the role and responsibilities of the managing authority and other key structures. This description should illustrate clearly the divisions of competence within committee structures and the integration of specific groups. The role of individual partner groups at each stage of the programme management cycle should be specified, preferably providing some historical context to illustrate how their involvement in the programme has evolved over previous programming periods. Of critical importance in the next programming period is a detailed account of monitoring and evaluation arrangements, taking account not just of the regulatory requirements but also the lessons of the penultimate chapter of this report.

The programme development process is too complex to be undertaken easily in the time available for it in the formal programming calendar. To resolve this, the process of strategy development needs to be seen as an ongoing process linked to other parts of the programming cycle.

The lesson from every programming period over the past decade is that the time factor is critical. All parties in the programme preparation process have had to work in a new regulatory environment under very tight time pressures. Frequently, there has been little time available for the design of strategies, let alone reflection, regional analysis, consultation with local partners and appraisal. Furthermore, even in 1996-7, the mis-timing of evaluation studies meant that interim or ex post evaluations were rarely available to inform programme preparation.

The experience of previous programme preparation suggests that the strategy development process should not be left until the detailed regulatory position has been finalised and then compressed into a period of 3-6 months. Instead, strategy development needs to be seen as a continuous process, linked into other parts of the programming cycle, with constant monitoring and refinement, and periodic review. As part of this process, the content, form and exploitation of evaluation studies needs to be organised to take account of reprogramming requirements, for instance targeting thematic or process issues of concern.

The design, management and delivery of Objective 2 programmes needs to be integrated more fully into the wider regional economic development context. In particular, this may become the only way to consistently achieve coherence with parallel human resource development programmes.

Objective 2 areas frequently form only part of a regional economy or labour market area, and Objective 2 strategies are only part of the total economic

92

development policy framework. Consequently, the links between individual Structural Fund strategies and the wider regional context are crucial for the effectiveness of programmes. Many of the more coherent regional strategies are those which have been able to build on existing, region-wide strategies, as in the Netherlands. In other regions, initiatives have been undertaken to engage in strategic analysis and consultation about the wider regional context within which a new Objective 2 plan will fit, complemented by diagnostic exercises about the programme itself. The establishment of common strategic frameworks and understandings is to be recommended.

As part of this process, strategy preparation needs to be highly interactive, involving as wide a range of relevant partners as possible; research for this report has shown how the active participation of partners has improved the quality of programme documentation in many regions.

A particular challenge in the next programming period will be to ensure that Objective 2 ERDF and Objective 3 ESF policies and interventions can be designed which are complementary and mutually supporting, in spite of their being pursued through separate programming mechanisms. Establishing explicit and active consultation and information channels which are appropriate to both the programme development and implementation stages and which are at least as good as - but preferably better than - current channels within Objective 2 programmes must be a priority.

Among the potential benefits of establishing greater coherence and stronger networks between all regional economic development actors is that it will help to prepare programming areas (even at this stage) for the post-2006 period when Structural funding may not be maintained at current levels.

As programme development becomes more inclusive at the regional level, the final programme negotiations need to become more sensitive to this, for example by being undertaken with greater transparency and consistency.

The quality and relevance of programmes can be increased through wide partner participation in the programme development phase. It is important that programme dimensions which emerge through consultation and have wide partnership support feed through where possible into the final programme, and are not unnecessarily diluted during the final, restricted negotiation phase with Commission services. This risks undermining trust, partnership cohesion and the partners’ sense of ownership of the programme, which in turn can establish unfavourable conditions for implementation. To ensure this, the negotiation phase will need to be managed in a way which is more sensitive to the more complex and inclusive partnership context, notably by becoming more transparent and more consistent.

93

5. JOB CREATION

5.1 Introduction As unemployment has become an increasingly severe and persistent economic problem facing the EU, durable job creation has grown in prominence as a central objective of economic and social cohesion policy. This focus will intensify in the next round of Structural Fund programming. This chapter reviews how Objective 2 programmes have responded to the challenge of ‘doing more and doing it better’ for employment. After setting the context of unemployment in the EU, the chapter charts the rise in prominence of job creation as an Objective 2 priority, and assesses how it has influenced strategic direction, policy choice and content, and approaches to implementation. It goes on to consider ways in which practice could be further refined in future.

5.2 Context

5.2.1 Employment Action in the EU EU unemployment trends have worsened significantly and consistently over the past two decades. Between 1975 and 1985, unemployment rose across the 15 Member States from just 2 to 10.5 percent.124 While economic recovery in the mid-1980s meant that by 1989, many Member States were enjoying their lowest unemployment rates for a decade, the average rates did not fall to levels even approaching those of the mid-1970s. By the early 1990s, the situation had deteriorated again, with overall unemployment rising to a high of 11.2 percent in 1994 - nearly 20 million people. While a subsequent upturn has led to unemployment of just under ten percent – or 16.5 million people - underlying trends over the last two decades allow little optimism. The pattern has been one of unemployment rates ‘ratcheting up’, repeated falls during cyclical upturns not compensating for greater rises during the downturns. Even the overall growth in employment which has taken place over the decade since 1987 has failed to have an impact on unemployment figures, instead being ‘absorbed’ by labour market expansion.

The problem of overall unemployment has been compounded by its increasingly uneven distribution across space and sub-groups of the population. Significant disparities are present in regional unemployment rates – reaching a ratio of one to ten. These disparities are tending to widen over time, although the trend was halted temporarily during the 1980s upturn. Among the affected population, structural unemployment and labour market exclusion are widespread: one in two jobless people in the EU has been out of work for more than a year, while more than 20 percent of the workforce under 25 years old is unemployed. Unemployment rates are higher among the less skilled and frequently among women. There is also an effect of concentration, with more pronounced problems of female, long-term, youth and unskilled unemployment in areas with the highest overall unemployment rates.

124 CEC (1999) Sixth Periodic Report on the Socio-economic Situation and Development of the Regions of the European Union, Commission of the European Communities, Brussels.

94

Unemployment trends in Objective 2 areas have broadly mirrored the above trends. However, compared with EU averages, they have been relatively favourable over time, in part due to policy interventions. Commission analysis has shown that since 1989, the gap in unemployment rates between Objective 2 areas and the rest of the EU has narrowed appreciably. Absolute performance has also been favourable: in 1997, the average rate in such regions was less than in 1993 whereas in the EU as a whole it remained constant.125 In part, this reflects the combined impact of Objective 2 programmes and other indigenous economic development efforts – although it also derives from other factors, notably labour supply developments (with reductions in the working population because of ageing, early retirement and outmigration).126 Nonetheless, unemployment rates among Objective 2 areas remain high: in 1997, the average rate was 11.9 percent, 1.2 percent higher than the EU average. This average masks significant differences between Objective 2 areas in different Member States, with the lowest rates in Luxembourg and Austria (at 3.4 and 4.4 percent), followed by Denmark (5.5 percent) and the UK and Portugal (6.6 percent each). Member States with the highest rates of unemployment in early 1998 were France (12.1 percent), Finland (12.5 percent) and Spain (20 percent).127 In addition, unemployment is structural and persistent: the primary employment challenge in Objective 2 areas is compensating for severe job losses in traditional industries, by widening and modernising the industrial base and, in parallel, shifting the skills base to fit new demand profiles.

The persistence and gravity of unemployment trends across the EU as a whole, combined with consciousness of the negative impacts of the single European market on regional convergence, made the fight against unemployment ‘a first priority for the European Community and its Member States’.128 The need for a concerted effort to develop new solutions was reflected in the 1993 European Commission White Paper ‘Growth, Competitiveness, Employment’, which provided a European perspective on unemployment, giving a detailed analysis of its incidence and setting out the Commission's thinking on possible ways forward. The White Paper was further developed by the European Council during its meetings in Essen, Cannes and Madrid. The Essen European Council in December 1994 was the most important meeting in this series, leading to the ‘Essen Track’ proposals, which called on Member States to take action on five fronts:

��raising employability by increasing investment in vocational education and training;

��boosting the effects of economic growth through reorganised working time, moderate wage growth and the opening of new areas of employment;

��lowering indirect labour costs in order to encourage demand for labour; 125 CEC (1999) op. cit. 126 CEC (1994) Competitiveness and Cohesion: Trends in the Regions, Fifth Periodic Report on the Social and Economic Situation and Development of the Regions in the Community, Commission of the European Communities, Luxembourg. 127 Eurostat Press Release, April 1998. All figures relate to February 1998. 128 CEC (1995a) Community Structural Policies and Employment, Paper presented to the Informal Council of Ministers responsible for regional policies and spatial planning by Mrs Wulf-Mathies, 30 November – 1 December 1995.

95

��enhancing the effectiveness of labour market policy, for example by means of activation policies which prompt the unemployed to look for work; and

��intensifying measures to benefit underprivileged groups in the labour market.

As part of the policy of giving priority to employment generation, Member States undertook to develop more concerted approaches, including the development of multiannual employment programmes, setting out the measures which were planned or in place to address the above aims. In addition, pursuing the resolution to open new areas of employment, the Commission began to encourage ‘bottom up’ Local Development and Employment Initiatives (LDEIs) in fields which respond to growing and changing needs for local services and have the potential to boost the employment intensity of growth.129

The commitment to employment was further reinforced in the Amsterdam Treaty, which includes a major new chapter entitled ‘An Effective Europe: Jobs’, with the overall aim of achieving a high level of employment in the EU. Following this, and the Luxembourg Jobs Summit in November 1997, the Commission adopted the first employment policy guidelines for 1998, followed by the preparation by EU Member States of annual National Action Plans for employment (NAPs), built around the four agreed pillars of employability, entrepreneurship, adaptability and equal opportunities. The NAPs further develop the Essen employment programmes, setting out commitments and how they will be achieved. An Employment Committee was set up at EU level to support the initiative, advising on and reviewing NAPs and deriving recommendations from Member State experiences on how to increase employment levels.

Member States submitted their first NAPs at the Cardiff European summit in June 1998. Although with shortcomings, the plans demonstrated progress towards a comprehensive, multiannual approach, building on existing policy foundations, and benefiting from high-level political support and an explicit local dimension.130 Differences in the approach taken to tackling unemployment in individual Member States are explained by several factors. First, as mentioned above, the severity of unemployment varies considerably between Member States. Second, major differences exist between the type of joblessness in each country. For example, unemployment amongst young people is relatively more acute in Spain than in the Netherlands. Third, the stage of economic development attained by Member States differs, as do their social, cultural and institutional characteristics, and their existing systems of social security and labour regulation. Fourth, the political principles shaping policy vary between countries eg. prioritisation of labour market deregulation in the UK.

The NAPs are wide-ranging, encompassing both microeconomic and macroeconomic interventions (the latter including benefit reforms to underpin active labour market policies and tax and indirect labour cost reforms to help

129 CEC (1995b) A European Strategy for Encouraging Local Development and Employment Initiatives, Commission of the European Communities, COM(95), Brussels. 130 CEC (1998) From Guidelines to Action: the National Action Plans for Employment, DG V, Commission of the European Communites, Luxembourg.

96

foster entrepreneurship). They currently focus predominantly on supply-side policies, in large part as a result of recent positive trends in employment. In terms of the Structural Funds, this means that while they address the role of the ESF in helping to deliver commitments, they do not highlight the important role played by the ERDF in increasing the demand for labour – one of this Fund’s primary objectives.

5.2.2 Employment and the Structural Funds Considerable expectations have been focused on the role of the Structural Funds in reducing disparities in the EU, not only in terms of wealth (GDP) but also (un)employment. Impacting on patterns of employment has always been an important objective of the Structural Funds, but the prominence of this aim has increased in recent years, in line with the economic and policy trends described above. As unemployment has worsened, Community resources dedicated to economic and social cohesion policy have increased. In addition, the way they are deployed has been honed. Recently, those utilising the Funds have been exhorted to do more and do it better for employment, and, latterly, to contribute in a more concerted way to the realisation of the National Action Plans for employment.

Of the Structural Funds, the ESF, operated by the Directorate General for Employment and Social Affairs (DG V), has been the main focus of attention in Europe-wide debates about employment. The ESF was originally created in 1957 by the Treaty of Rome with ‘the task of promoting…employment facilities and the geographical and occupational mobility of workers’ (Art.123). Reforms and revisions to the ESF in 1971, 1977 and 1983 not only adjusted its fields of intervention, but also increased the regional dimension of its role in employment by introducing a greater commitment to differential assistance for regions with employment problems.

Over the last decade, the main role of the ESF has been to respond to supply-side issues in the labour market, primarily increasing the employability and adaptability of the workforce, rather than increasing demand for labour per se. Under Objectives 3 and 4, it has targeted specific populations EU-wide, combating long-term unemployment and ensuring the occupational integration of young people (since 1989) and helping to adapt the workforce to industrial change (since 1994). Since 1989, the Fund has also provided a complementary flank of the geographically defined ‘Objective’ programmes for the worst affected parts of the EU.

In Objective 2 programmes, the ESF operates alongside the ERDF. Overall, the ERDF has been more oriented to direct job creation through diversification and modernisation, while ESF interventions are used to support this reconversion process by adapting the skills base to a changing profile of demand for human resources.

In addition to refocusing mainstream policies, the Commission has encouraged reflection on innovative ways to generate jobs using the Structural Funds, and so implement the Essen commitment to ‘boosting the effects of economic growth through…the opening of new areas of employment’. The Territorial Employment Pacts (TEPs), which were given official recognition by the European Commission on the eve of the European Council in Amsterdam, are

97

one of the vehicles for developing such innovative actions. They consist of integrated projects whose objectives are first to attack the roots of unemployment by creating real employment opportunities at the local level, and second, to generate new ideas by sharing experience. TEPs work from the bottom up, mobilising enlarged partnerships of local actors with a role in employment generation to develop new and integrated approaches based on a detailed analysis of local situations. The present 89 TEPs, which qualify for support from the Structural Funds until the end of 1999, cover a total population of 35 million. For the most part, they are centred in areas particularly affected by unemployment – including Objective 2 areas.

Selected Objective 2 programmes responded to the TEPs and LDEIs during the 1997-99 cycle, providing a valuable source of potential co-financing for the project ideas generated by these initiatives. The Turnhout SPD includes LDEIs, which are being pursued to create 100 new jobs, half of them for women. In Italy, Emilia Romagna included provision to fund the proposals of a local TEP. Some French programmes also included TEPs (eg. Pays de la Loire, where proposals cover personalised service provision, environment-related jobs, job-sharing, micro-companies and new occupations) or agreed to pursue the Pact methodology for other proposals (Picardie, Haute Normandie). Further programmes stated that, should a TEP be established in the programming area, the programme would give special priority to providing funding for the projects it generated.

Some Objective 2 programmes have adopted the principle of boosting the employment intensity of growth, by proposing actions integral to their programmes which will help to support development in the identified growth areas. For example, some German programmes propose training measures for various types of service job, including enterprise-related services, social and neighbourhood services and the environment. Such approaches could in future become more widespread, drawing on the lessons distilled from the experience of TEPs and LDEIs.

5.3 Objective 2 Programmes and Employment

5.3.1 Trends in Objective 2 Programmes The core eligibility criteria for Objective 2 areas relate to patterns of employment.131 Since the launch of the first programmes in 1989, the creation and/or safeguarding of jobs has always been high among their ultimate objectives and the primary indicator for assessing their achievements. In every reprogramming exercise since 1989, the emphasis placed on pursuing enhanced job creation outcomes has grown steadily, so that by the time of the

131 Council Regulation (EEC) no. 2081/93 of 20 July 1993 amending Regulation (EEC) No. 2052/88 on the tasks of the Structural Funds and their effectiveness and on coordination of their activities between themselves and with the operations of the European Investment Bank and the other existing financial instruments (Article 9, p11), published in Official Journal of the European Communities No. L193; 31.7.93.

98

Objective 2 reprogramming exercise for 1997-99, the Commission’s guidance presented job creation as the ‘paramount priority’.132

A comprehensive overview of the employment achievements of Objective 2 programmes is not available, given the timings of relevant programmes. However, the ex post evaluation of the 1989-93 programmes estimated that some 845,988 gross jobs had been created (based on extrapolation of incomplete monitoring data), just over half of these being net jobs (447,351). The current 1997-99 programmes are expected to create, save or redistribute almost 880,000 gross jobs.

Intensifying the focus on employment outcomes in Objective 2 programmes has involved three broad challenges.

��First, the policy content of programmes has evolved in response to the call to do ‘more for employment within existing programming and [do] it better’.133 Both the Commission and individual programmes have aimed to increase the value for money of Structural Fund expenditure by enabling more, higher quality, and more durable jobs to be created, and to promote a more successful fight against labour market exclusion.

��Second, programmes have been making employment more ‘visible’ as a theme, adjusting their documentation to make explicit an issue which was formerly considered (to a certain extent) a self-evident consequence of programming.

��Third, programmes have increasingly been employing more mature and effective methodologies and mechanisms comprehensively and reliably to predict and then demonstrate, the employment impacts of Structural Fund co-financed interventions.

Trends in each of these three areas are discussed in the sections which follow.

a. Trends in policy content In addition to changes in strategic focus and programme presentation, the balance of expenditure on different fields of intervention has evolved over time in Objective 2 programmes to prioritise employment outcomes. Whereas the note on Community Structural Assistance and Employment134 describes Objective 2 programmes as ‘seeking the most immediate possible impact on employment’, the typical policy mixes that have emerged in Objective 2 as a result of this process of evolution reflect broader and longer-term policy objectives. These encompass interventions with impacts over varying timescales and affecting a wide spectrum of the workforce: helping to establish the conditions for long-term economic growth by raising competitiveness; increasing the job intensity of this growth; and taking a pro-active approach towards social solidarity. These are three of the four overriding goals emphasised in the aforementioned EC Communication written to guide Structural Fund interventions.

132 CEC (1996a) Note for Guidance concerning Operations in the Declining Industrial Areas (Objective 2) for the Second Programming Period 1997-1999, Commission of the European Communities, Brussels. 133 CEC (1996b) Community Structural Assistance and Employment, Commission of the European Communities, COM(96) 109 final, Brussels. 134 CEC (1996b) op. cit.

99

Measures which have proved to be of poor job-creating potential have not been continued in some cases. Two types of policy have been marginalised in Objective 2 programmes by the increased focus on employment impacts. The first comprises those which promise limited economic impacts over any timescale relative to expenditure. The primary example here is basic infrastructure where this has already reached a ‘critical mass’, beyond which considerable investment is required to make an appreciable difference to the functioning of the regional economy. The orientation away from infrastructure was underlined in the Commission guidance on formulating the 1997-99 programmes, which stated that:

‘co-financing of basic infrastructure investments…requires specific justification in terms of regional development needs and must directly contribute to the synergy and objectives of the programme, in particular to the safeguarding of existing and creation of new jobs’.135

During the 1996 negotiations, the Belgian and French programmes were among those which reduced their emphasis on basic infrastructure in favour of business development, and incorporated the proviso that in order for infrastructure projects to be considered, they should have direct economic impacts. Programmes now requiring physical regeneration projects to demonstrate their relevance to employment outcomes include: Western Scotland, Greater Manchester, Picardie and Bourgogne. A similar approach is also being taken increasingly for environmental improvement projects, which have sometimes in the past been purely cosmetic and insufficiently related to clear economic objectives.

The second type of intervention to be sidelined are interventions bringing about direct negative employment growth eg. through industrial automation. Although such policies may enhance competitiveness by increasing productivity, and so eventually bring about employment growth, they have become politically increasingly difficult to support within Structural Fund programmes. Such policies are typically still pursued, but in parallel, outside the programme.

At the same time, and following Commission guidelines, programmes have universally given increased priority to developing indigenous business potential by providing direct support for the productive sector. This tends to have significant, fairly easily measurable and often quasi-immediate and durable employment effects, and results in favourable ‘cost-per-job’ ratios. The Nordrhein-Westfalen programme has moved in this direction only fairly recently. Having addressed the principal physical deficits which resulted from the decline of heavy industry, the region is now building on this foundation and shifting the emphasis towards measures which will create more numerous jobs in a more direct way (eg. start-ups and business development). At an earlier stage of diversifying away from heavy industry, the programme placed a heavier accent on infrastructure projects which were not ‘employment-rich’, but were designed to create the conditions for later employment generation.

As experience has been gained in SME development measures under Structural Fund programmes, these measures have increasingly been honed

135 CEC (1996a) op. cit.

100

(eg. in Germany and the UK) to ensure more careful targeting and project development, including by establishing business needs more carefully, to ensure greater benefits are derived. In addition, programmes have more recently experimented with financial engineering measures helping to ensure improved value for money from the projects supported.

Direct business development projects are often job-rich, but they do present limitations. First, they fail to enhance the collective business environment. Second, without their implementation being rigorous, these measures tend to comprise a high proportion of deadweight as the most job-rich projects are the most likely to have gone ahead anyway. Third, unless firms are targeted whose markets are predominantly outside the regional economic space, they can have high displacement, growth in one firm leading to the contraction of another within the same region. For these reasons, focusing a large proportion of programme resources on individual firms is arguably both a short-term and an iniquitous use of resources, helping specific parts of the private sector only, and leaving no structures which are able to support businesses more broadly in the long term to diversify, innovate and enter new markets. For this reason, investment in the collective business environment has remained a strong focus of Objective 2 programmes, although it is less employment-rich in the short term than direct interventions.136 Such measures may even be increasing again in significance, as in Denmark where they became stronger elements of the 1997-99 programmes during the negotiation phase. Among the examples of measures improving the conditions in which SMEs operate are the creation of business-oriented RTD and innovation facilities, which will raise regional competitiveness in the longer term, but which have limited ‘value’ in the short term if the only measure of quality is employment impacts. Such projects entail a degree of risk and represent a high cost-per-job over the programme implementation period. However, they do promise to make a region more competitive on a medium to long-term basis, and, in the short term, the few jobs they generate directly tend to be of high quality. Aquitaine is among the programmes which have placed significant emphasis on structuring the business environment, and in particular raising regional technology and innovation levels. In South Karelia, too, the Structural Funds are being focused on ‘smoothing the process’ of regeneration by removing bottlenecks. ‘Drivers for growth’ have become the focus, which should lead to solid new jobs in the longer term.

ESF policies in Objective 2 programmes have also evolved over time. The 1989-93 period was the first phase in which the ERDF and ESF (established in 1975 and 1957 respectively) were ‘required to act in a closely co-ordinated manner and with a common set of objectives’.137 A strong articulation between ERDF demand-side measures and ESF supply-side ones is essential under Objective 2, where the primary issue is one of reorienting economies – and the people employed in them – to growth sectors, but this has been difficult to achieve, especially at first. Initial deficiencies in ensuring that the ESF was driven by economic development imperatives in part stemmed from the regulations under which the ESF operated. During the 1989-93

137 Armstrong H (1993) Community Regional Policy, in J Lodge (Ed.) The European Community and the Challenge of the Future, Pinter, London.

101

programming period, the ESF was governed more by the application of detailed eligibility criteria than by a clear definition of problems to be solved. It was confirmed later that there was almost universally poor coherence between the activities supported by the two Funds.138 To help resolve the problems, the ESF was revised as part of the 1993 reform of the Funds: ‘The new ESF regulation [was] designed to make intervention of the Fund more policy-driven, thus avoiding the restrictions of the former system in which detailed criteria often prevented the funding of suitable schemes’.139

Over time, there has been increasing tailoring of ESF policies, not only to specific target groups, but also to economic sectors, and including interventions able to tailor training to the requirements of specific firms or individuals. A trend which promises to improve the match between supply and demand in the labour market has been the use of ESF resources to improve labour market information systems, increasing knowledge of the situation of skills supply and demand in order to facilitate better targeting of interventions. Industrial South Wales used Technical Assistance in its 1994-96 programme to support labour market analyses. Both the Fyrstad and Ångermanlandskusten SPDs (1995-99) stated that an in-depth analysis of the local labour market would be undertaken in order to identify skills gaps in local companies and so target ESF expenditure more effectively. The Ångermanlandskusten SPD also stated that the ‘education and training profile of the local unemployed should be examined...to target ESF training where it will have the most positive effect’.140 In the 1997-99 round, further progress was made following Commission recommendations, with systematic and ongoing analysis of skills mismatches being given a higher profile (eg. in some French programmes). In Western Scotland, the review process identified a need for clearer analysis of SME training needs, a requirement likely to be provided (in part) by a Scottish Enterprise review of skills in SMEs. The 1997-99 Finnish SPD likewise stated that the development of human resource projects would be based on evaluations of labour market strengths and weaknesses.

A suite of policies which has become increasingly prominent in spite of apparently poor employment performance - in the short or longer term - are community economic development interventions, justified by the need to counter economic and social exclusion by targeting additional support to regenerate the worst affected parts of eligible areas. This type of intervention was an element of some early programmes (eg. in the Western Scotland), and became a part of most UK programmes in the 1994-96 round following this experience, although its implementation proved difficult during this phase because of the lack of established local partnerships able to develop and deliver sufficient appropriate projects. A targeted learning process was undertaken addressing some of the issues experienced, with the theme being continued, more successfully, into the 1997-99 programmes, not only in UK but also in many French and some German programmes. The integrated

138 Ernst & Young (1997) Ex Post Evaluation of the 1989-93 Objective 2 Programmes – Synthesis Report, Commission of the European Communities, Brussels. 139 EIU (1993) European Trends, 4th Quarter 1993, Economist Intelligence Unit, London. 140 Ångermanlandskusten Single Programming Document, 1995-1999, Commission of the Euroean Communities, Brussels.

102

groups of measures proposed, including training to improve the employability of the long-term unemployed or to increase entrepreneurship in zones of high urban deprivation, tend to have extremely high displacement effects and may not be durable. However, they prioritise the quality and distribution of impacts rather than their quantity.

b. The ‘visibility’ of the employment objective Over time, the visibility of employment creation as an aim of Objective 2 programmes has increased, not only in the way programmes and associated materials are presented, but in the discourse surrounding them at all levels.

Many of the most striking changes in programming documents took place between the 1994-96 and the 1997-99 periods, following the Commission guidance on programme reformulation. The increased profile of employment in programming documents has been reflected in most of the main sections. First, regional profiles and SWOT analyses have come to include more detail, often disaggregating employment trends by sectors and sub-groups of the population, going beyond a broad justification of need in the eligible area to providing insights helping to focus policy. Second, compared to previous programmes, strategy sections more frequently highlight employment creation as a central objective. This is also reflected clearly in other materials publicising programmes. At the priority and measure levels, the importance placed on job creation is reflected in policy rationales, selection criteria and targets.

At an operational level, programme managers and partnerships have become increasingly focused over time on the employment imperative in policy design, project assessment and programme monitoring. Participation in the regular menu of programme evaluations over the last decade has also helped to bring home the need for a continuing explicit, critical focus on programme impacts, especially in terms of employment, and to familiarise more actors with the techniques and terminology required to address the theme with confidence.

At Commission level, a wide range of policy documents and initiatives has also helped to raise the profile of employment as an objective of Structural Fund programmes generally and of Objective 2 programmes in particular. The review of 1997-99 Objective 2 programmes, focused particularly on how the employment theme had been taken up.141 A higher profile initiative under Objective 2, whose results were widely disseminated, was the ‘Jobs Challenge’, an international competition launched in 1997 to identify projects which had been particularly successful in creating jobs and to distil transferrable lessons from their experiences. The international competition was followed up with a publication describing the winning projects, and, in March 1999, by a Helsinki seminar on best practice.

c. Trends in quantification The proportion of SPDs specifying and quantifying their employment targets has increased steadily over successive programming periods. Across the

141 CEC (1997) The New Regional Programmes 1997-99 under Objective 2 of the Community’s Structural Policies – Focusing on Job Creation, Commission of the European Communities, COM(97)524 final, Brussels, 14.11.97.

103

1989-93 programmes, only one in five programming authorities included explicit employment targets in their programming documents.142 Few details are available to asses the quality of targets at this time, but the ex post evaluation found that, compared with likely actual outcomes, most were underestimates. The situation improved in terms of coverage in the 1994-96 SPDs and in those which were approved in 1995 for the new Member States,143 most making reference to their envisaged employment effects. Nonetheless, during this phase, just under half of all SPDs still had incomplete, unquantified or in some cases absent employment targets. Where quantified targets were set, there were also quality issues in most cases, including:

��some inconsistency and inaccuracy in the terminology used to describe impacts;

��failure to specify the nature of anticipated employment impacts (temporary, permanent, direct, indirect);

��the absence of an explicit methodological basis for targets – except in one Spanish programme, and here only for construction-related employment; and

��a concentration on gross targets, with little or no effort to identify possible net effects, once displacement, deadweight, multipliers and double counting had been taken into account.

These shortcomings were reflected in considerable diversity between programmes in terms of their anticipated ‘cost per job’ – even between ostensibly similar programmes in the same Member State.

Moving into the 1997-99 programming period, both the coverage and the quality of employment targets improved across the board, with each programme engaged in a process of learning and building on past experience. The greatest progress was made in programmes and groups of programmes where a dedicated initiative was undertaken to improve target setting, either using internal programme resources or consultants, and drawing on and adapting established methodologies, including those of the Commission and guidance generated by the MEANS programme. Case study examples are provided later in this chapter. A small number of Member States continue to resist detailed prior quantification because of the uncertainties inevitably involved in target setting. However, in general, the principle is accepted and expertise is accumulating in how it can be applied.

5.4 Programme Implementation and Job Creation Having reviewed the progress made by Objective 2 programmes in enhancing the job creation focus of strategies – through improved policy content, presentation and quantification – the following section considers how the implementation of Objective 2 programmes has been adapted to the challenge of creating employment. It begins by discussing the extent and quality of

142 Ernst & Young (1997) op. cit. 143 Bachtler J and Taylor S, with Kearney C (1996) Extended Synthesis of Agreed Single Programming Documents in Objective 2 Areas 1994-96, Report to the European Commission (DG XVI), EPRC, University of Strathclyde, Glasgow.

104

target setting and then addresses project appraisal, decision-making and monitoring issues from a jobs perspective.

5.4.1 Target Setting Parallel with the need for Objective 2 programmes to have enhanced employment impacts is the requirement to anticipate and verify them more accurately and comprehensively. This has systemic implications for the way programmes are delivered, and in particular how targets are established and their achievement monitored. More rigorous prior quantification is required, based on sound methodologies, verification of individual project targets and close monitoring of their achievement during implementation.

In quantifying and delivering employment outcomes accurately and comprehensively, Member States and programming authorities are still to an extent ‘feeling their way’ in a field which has planning, resource, skills and information implications for every stage of programming from initial programme development to final evaluation. Recognising the complexities involved at programme level in promoting and quantifying job creation under Structural Fund programmes, the Commission has accelerated its endeavours to provide technical guidance in the field, principally through the ‘MEANS’ programme (Methods for Evaluating Actions of a Structural Nature), which prepares handbooks on monitoring and evaluation issues, and through the organisation of regular training and exchange seminars across Europe.

Objective 2 programmes are implemented across zones which are significantly smaller than those for Objective 1. They are sometimes very fragmented, and rarely coincide with standard statistical areas. ‘Top-down’ macroeconomic evaluations based on change from baselines tend to generate results which pose significant limitations. Given this context, the need has been particularly strong for Objective 2 programmes to supplement observation of baseline trends with the bottom-up ex post quantification of impacts, establishing frameworks consisting of quantified output and impact targets at measure level, and working up from these to setting targets for each priority and then for the programme as a whole.

In a detailed examination of the 1994-96 programmes (and 1995-99 programmes for the new Member States),144 it was possible to calculate the gross number of jobs created per MECU of total expenditure for 66 out of 82 programming documents by aggregating impact data from the measure level. The results showed considerable variation, with values ranging from high estimates of 172.2 jobs per MECU in Turnhout and 101.1 jobs per MECU in Plymouth to under 2 jobs per MECU in some instances (eg. Aragón, Groningen Drenthe, Hessen and Picardie). The range of variation in cost per job projections within Member States was only slightly less extreme in most instances. Only the figures for Sweden and the UK displayed anything like the sort of consistency which would indicate that broadly similar techniques and comprehensiveness of coverage had been applied.

The wide variation in projected impacts reflected not just the real differences in impact which arise from differing policy choices, but probably more

144 Bachtler J and Taylor S, with Kearney C (1996) op. cit.

105

significantly, differences in the techniques used to make the projections and variations in the comprehensiveness of information provided. In most cases where extremely low values were obtained, quantification was incomplete, with documents supplying job creation data for only some of their measures. However, even among those documents which supplied apparently complete projections of job creations, there was a marked variation in the cost-per-job ratio. For example, in Arnhem-Nijmegen the ratio was 13 jobs per MECU while for East London/Lee Valley the ratio was 41 jobs per MECU.

Table 5.1: Cost of job creation in Objective 2 areas, by country* Member State Number of

Areas Gross jobs per MECU

total expenditure Corrected Average

1997-99 figures

Austria 4 2 to10 - Belgium 4 14 to 172 32 43 Denmark 2 11 to 13 13 17 Finland 8 35 35 44 France 16 1 to 74 30 33 Germany 5 1 to 78 57 63 Italy 10 3 to 24 16 54 Luxembourg 1 34 34 24 Netherlands 5 2 to 25 16 38 Spain 6 1 to 11 - 44 Sweden 5 17 to 30 23 UK 8 23 to 101 48 66

* Figures include only jobs created and not jobs which are described as ‘temporary’, ‘maintained’ or ‘maintained and/or created’. Overall, when the ratio of jobs created per MECU spent was below 30, the figures provided tended either to be incomplete or to refer only to those jobs created directly by the measures concerned and/or in the course of the implementation of the programme. When the ratio was above 30, indirect job creation also seemed to be included and a longer evaluation period used. The lack of methodological information included in SPDs precluded the confirmation of these impressions.

The Commission review of the 1997-99 programmes found that ‘substantial progress has been made in the quantification of employment effects in the 1997-99 SPDs’,145 with detailed measure-level estimates in 55 of the 73 new programmes. The SPD negotiation phase prior to approval sometimes helped to improve or increase target setting (eg. across the German programmes). In addition, the targets set by some programmes were improved by detailed partnership-based target setting exercises subsequent to SPD approval, for example in Aquitaine, where Commission methodologies were exploited.

Detailed examination of programmes demonstrates that while coverage has improved, significant shortcomings remain. First, quantification is still patchy, if not entirely absent in several cases, including some Belgian programmes (although this was improved during negotiations, but possibly not

145 CEC (1997) op. cit.

106

fully reported in the finalised SPDs). With regard to the levels of the programme at which targets are set, few programmes provide targets at all levels. Of those which do, the best examples are in selected UK, Austrian and French programmes, where employment objectives are quantified for the whole programme as well as for individual measures. Some programmes only provide total figures for the SPD as a whole, such as in Finland (because of its status as a national programme). Other regions provide more detailed quantification down to measure and even sub-measure level, this occurring especially in the Netherlands and some French regions.

With regard to the targets set, there has been progress in terms of how they are described, but this is uneven. For example, while targets are usually expressed in terms of jobs created/maintained, some programmes only focus on job creation (eg. Finland). On a positive note, occasionally other measures are used to describe employment impacts – eg. percentage reduction in unemployment rate in some Finnish regions, or number of ‘workplaces’ in Dutch regions. Regions in the Netherlands, Denmark and Spain also tend to make the distinction between temporary and permanent employment.

The quality of targets varies considerably. Estimates vary from highly precise figures – in a minority of cases - to rounded-up numbers. Little information is provided in the majority of SPDs on how targets were derived: figures are simply stated. There are very few instances where issues such as deadweight are taken into account (two French regions being exceptions). In addition, displacement effects are neglected in the calculations, as is the possibility of double-counting of jobs at the programme/measure level, although again, there is not enough information on the calculation process to assert this with certainty.

BEST PRACTICE IN QUANTIFYING EMPLOYMENT OUTCOMES

More programmes are specifying what is included in employment targets, detailing some of the assumptions on which these targets are based, and indicating the timescales over which they should be realised.

The type and quality of jobs created is being specified with increasing care:

• = differentiating between ‘temporary’ employment generated by implementing projects (eg. in construction, often referred to in terms of ‘person years of work generated’) and ‘permanent’ posts (eg. created in the longer term by the use of infrastructure).146

• = specifying the nature of the jobs created (full-time, part-time, seasonal, etc).

There has been increasing precision in the use of employment-related terminology and concepts, indicating greater familiarisation with the field and increased reflection. Definitions are being used which more accurately reflect the true nature of employment impacts:

• = no longer describing people entering employment as a result of enhanced skills as ‘jobs created’ (except where they are entering self-employment). A more appropriate term for those entering employment is ‘jobs obtained’, while a useful term for those who were already employed, but whose jobs have changed materially because of training is ‘jobs transformed or upgraded’.

146 The terms ‘temporary’ and ‘permanent’ are problematic generally, as all jobs are in reality temporary.

107

• = distinguishing between new and safeguarded jobs, and quantifying each type separately. (The setting of targets for jobs ‘created or safeguarded’ is unhelpful for predicting a programme’s net effect on employment levels.)

• = not defining every job in every firm directly affected by an intervention as safeguarded/maintained.147 This definition is only appropriate when a job would otherwise have been lost. This is often difficult to demonstrate (and, indeed, rarely the case). A preferable approach is either to state that a set percentage of the jobs touched have been safeguarded (eg. in the case of a town centre regeneration project148), or to choose a more accurate description, such as ‘jobs transformed or upgraded’. This definition is a valuable addition, bringing value to some important employment impacts of Objective 2 programmes which might before have gone unmeasured, or been measured misleadingly. It also recognises the impact of projects in companies which were not necessarily threatened with closure, but which have become more competitive as a result of an intervention (eg. through technology or new management approaches).

In terms of method, few if any programmes appear at this stage to have no basis for the targets set. The predominant current approach involves the exploitation of previous experience. This ranges from informal methods, which rely on economic developers having a ‘feel’ for realistic figures (although this represents a complex processing of experience, combined with detailed local or regional knowledge), to more formal approaches, using quantitative and sometimes qualitative data generated from previous interventions to establish gross and often only direct targets. Clearly, the more formal, evidence-based approaches are to be preferred, but these depend on data being available. Among their advantages is that they avoid overestimation driven by political priorities (as in Ångermanlandskusten, in establishing the 1995-99 programme).149

It is useful if the exercise goes beyond gross impacts to include net impact estimates, taking into account the negative effects of deadweight150 and displacement151 (which vary between measure type), as these can radically affect which policies offer the best value for money in employment terms. A further useful addition is to find some means of assessing and comparing the quality of jobs likely to derive from different interventions (in terms of skill levels, value added, the degree of permanence, or the populations which will be affected by them). In practice, such systematic exercises are frequently only undertaken when the policy mix and resource allocations have already been decided, in order to establish the monitoring and evaluation framework.

147Bernard Reverdy Consultants (1997) Appréciation externe des DOCUP pour les Régions Objectif 2 1997-99 - Rapport de Synthèse, February 1997, Report to the European Commission, Brussels. 148See MEANS (1997) Counting the Jobs. How to Evaluate the Employment Effects of Structural Fund Interventions, DG XVI Series: Evaluation and Documents No 1, January 1997, Commission of the European Communities, Brussels. 149Eurofuture AB (1997) Mål 2 Ångermanlandskusten - Slutrapport från halvtidsutvärderingen. 150 ‘Deadweight’ is the term used when the effects or impacts of Structural Fund interventions would have been realised even if the intervention had not received Structural Funds support. 151 ‘Displacement’ occurs when the employment effects of Structural Fund interventions displace employment elsewhere in the programme area. This leads to a reduced net employment effect.

108

INVESTING IN A SYSTEMATIC, NATIONAL FRAMEWORK FOR QUANTIFICATION – THE UK

A national exercise has been undertaken in the UK to improve the quality of employment targets and their subsequent monitoring.152 Consultants conducted a review and critique of the indicator information contained in the English and Welsh SPDs, and then designed a standard set of output and impact indicators for all the activities being supported by the programmes. It then sought to provide methodologies to quantify output and impact targets for these activities, based on widely available data and previous experience, as measured in monitoring and evaluation materials. This approach is comprehensive, structured and methodical, but also pragmatic: it accepts that perfect quantification of outcomes and impacts is not possible nor even necessary. If programmes focus on the most quantifiable elements, then they will be able to quantify the likely outcomes of the large majority of programme expenditure.

The approach drew heavily on DG XVI methodologies developed through MEANS, and on the similar Scottish Enterprise ‘Output Measurement Framework’ designed, published, applied and subsequently revised and improved for the benefit of the Scottish network of Local Enterprise Companies.

It is hoped that the introduction of a national system for target setting will improve the capacity of responsible programme personnel and so deliver more realistic targets. It should also help applicants to set their own targets, so enabling effective monitoring. At national level, it will permit programmes to be more easily and consistently compared.

EX ANTE APPRAISALS OF EMPLOYMENT EFFECTS FOR SINGLE PROGRAMMES: AQUITAINE AND BASSE NORMANDIE

To improve the coverage and quality of their employment indicators, the 1997-99 Objective 2 programmes in Basse Normandie and Aquitaine were both the subject of systematic ex ante impact evaluation exercises based on Commission methodologies. The programmes worked on a measure by measure basis, drawing on the knowledge and expertise of their partnerships, to derive assumptions from which gross and then net output and impact targets could be set (broadly, the methodology set out in ‘Counting the Jobs’).

In Aquitaine, the Evaluation Committee (a sub-group of the Monitoring Committee) drew on their respective data sources, to go as far as possible towards developing realistic output and impact targets on a measure by measure basis. The exercise began by defining terms, explaining the key definitions used in employment monitoring, and then explaining the methodology with which employment targets would be calculated under the various headings.

The most problematic measures were those which had the most indirect or long-term employment impacts, including ESF actions and RTD interventions. Where there was doubt, modest targets were set, and an explanatory text provided, giving the assumptions on which targets were based and emphasising that the value of such measures lay in their ability to support the success of other actions, or to have longer term structuring effects. The exercise will provide a useful framework for future monitoring and evaluation, and its results will be refined at regular intervals as available information improves.

152 EKOS Ltd (1998) Baselines and Quantification in the United Kingdom Objective 2 Programmes, Volume I: Guidance on Baselines and Quantification, Study undertaken on behalf of DG XVI of the European Commission, Brussels.

109

PRAGMATIC APPROACHES TO TARGET SETTING – NORDJYLLAND

In order to be able to compare the potential employment impacts of proposed measures at the 1997-99 programme drafting stage, and then set realistic employment targets, the Nordjylland programme supplemented the instinctive feel for realistic targets which had been gained from previous programming experience with a modest quantitative review.

This study, carried out by the programme secretariat, worked out the cost per job of 285 projects across three broad types of intervention (investment, knowledge and infrastructure), based on information supplied by implementers at the close of projects. Partly involving the analysis of information which the secretariat already had, this analysis generated cost per job information in a cost-effective and rapid way. The results were then adapted using multipliers to show potential medium and longer term employment impacts, on the basis that the number of jobs created at project end could be expected to have doubled 1-2 years later.153 This enabled the effectiveness of different measure types to be compared, and also assisted the Nordjylland secretariat in conducting more informed negotiations with the national authorities.

Limitations of the Nordjylland exercise include the reliability of the unverified information supplied by project implementers. Generally, however, such exercises give a firm foundation, supplying assumptions on which transparent direct job creation targets can be set and both policies and projects realistically appraised.

5.4.2 Project Appraisal and Decision-Making – Communicating the Jobs Message Making the job creation imperative explicit to applicants makes it more likely that it will be followed through into projects. Expectations have to expressed clearly, and then tools provided to respond to them. There are several channels through which the employment message can be communicated to applicants, including the programming document itself, other publicity materials and client-based pre-application advice services. In many cases, the prominence given to the employment problem and to anticipated employment impacts through each of these channels has increased over successive programming periods. Indeed, the significance of effective, co-ordinated programme marketing is such that some programmes have invested additional resources in personnel with a publicity and public relations remit (eg. Industrial South Wales).

In Member States where the SPD or OP is the core document guiding Objective 2 implementation, several sections can be shaped to help to raise applicants’ awareness:

��the regional profile, where the dimensions and character of the employment problem (including any gender differentiation) are described;

��the strategy section, which sets out programme objectives and the rationale behind them; and

��the measure descriptions - which detail which actions will be supported and why, expected impacts and the selection criteria which will be used to assess individual proposals.

In general, the presentation of the job creation priority to applicants of current programmes is poorly developed, although stronger than in previous rounds.

153Nordjylland Single Programming Document 1997-99.

110

a. Communicating the jobs message – the strategy level Increasing numbers of programmes use their strategy section to state that employment is the overall objective of all interventions (Aquitaine), or the primary selection criteria (eg. Nordjylland, the Finnish programmes and Bretagne, where job creation will be prioritised, but especially if it is in industrial sectors).

In the 1997-99 SPDs, many strategic statements give employment creation primary importance. The focus on job creation issues is most visible where it features at the level of the overall aim (as in Finland, and many French and Spanish programmes). Some programmes reinforce this message by providing quantitative targets in their strategic aims (eg. one French and two Finnish programmes). In other countries, most regions explicitly mention the issue in either the strategic aim and/or the sub-objectives which further develop the aim. The only countries where little mention is made of the issue at the strategy level are Belgium and Italy. In some cases, job creation is implicit in the processes which programmes state that they will encourage and which will generate employment, for example diversification of the economic structure and support for growth sectors. This is a striking feature of UK SPDs.

There is some variety in the way employment objectives are expressed. Several terms are used largely interchangeably (‘job creation’, ‘job maintenance’, ‘creating sustainable jobs’, etc), with a distinction being made between job creation and maintenance in only one in three SPDs. In only a very few cases is the strategic aim phrased in terms of reducing unemployment numbers and rates (eg. Nordrhein Westfalen). In some cases, the objectives are vaguely worded, including reference to improving employment ‘prospects’ (Dutch SPDs) and ‘opportunities’ (Swedish SPDs). Some programmes move beyond numbers of jobs to consider their quality or other characteristics: perhaps one-third of SPDs with employment objectives refer to the creation of ‘sustainable’, ‘good quality’ or ‘permanent’ jobs, or highlight target groups to benefit from job creation. For example:

��targets for qualified, quality jobs (Steiermark, Vorarlberg)

��the creation of sustainable, quality jobs in the commercial and service sectors (Schleswig-Holstein)

��prioritising industrial jobs (Bretagne)

��highlighting the creation of seasonal, qualified and environment-related jobs (Picardie)

��widening employment opportunities for men and women (Bergslagen)

��improving the flexibility of the overall workforce (Fyrstad)

Some objectives are aspirational – setting an overall, rather generalised goal, in which job creation may be explicit or implicit – or list priority areas without any real sense of how these work together as part of a growth strategy. In particular, there is not a sense of overall economic models embodied in the strategies. A minority of regions are clearer on the growth models they are pursuing. For example, several regions intended to exploit their geographical advantages to act as international ‘gateways’ along major/emerging economic trade corridors (among them, regions in Finland and the Netherlands).

111

Similarly, Nordrhein Westfalen has a growth strategy focused on developing SMEs, while that of Nordjylland prioritises internationalisation.

Many objectives mention processes by which jobs will be created. They are usually associated with raising competitiveness and include modernisation, diversification, development of key sectors, SME development, human resource development, and/or increasing technology and innovation levels.154 Part of the overall objective of Industrial South Wales is to ‘maximise…job creation…through the creation of a diversified industrial base and the successful exploitation of innovation’. Job creation objectives are often placed in the context of sectoral diversification, with frequent explicit references to manufacturing and tourism. In at least one case, the target sector is even more focused (the car industry in Varsinais-Suomi). There is usually very little to suggest how these different processes operate together and how they will directly influence employment. There is an assumption that support for business development – either directly or through the business environment – will generate growth and automatically result in new jobs, but the potential for this is rarely addressed in a rounded way (for example acknowledging that measures may involve low quality job creation or create opportunities for people from outside the Objective 2 area).

While some policies contribute both to raising economic competitiveness and creating jobs, some influence only one of these objectives. This establishes a certain tension, which is managed in some programmes by the parallel expression of objectives. Industrial South Wales has such an approach, aiming not only to maximise job creation but also to encourage ‘balanced and sustainable economic growth’, and ensure that opportunities contribute to intra-regional economic and social cohesion. While many programmes have raised the prominence of employment creation, there is limited evidence that their policies have actually been reoriented in line with this. Instead, programmes largely remain driven by objectives of competitiveness.

The ability of Structural Fund programmes to impact on employment depends on the policies they pursue. Different types of economic development intervention deliver a different quantity, quality and durability of employment outcomes, over different timescales. It follows that the balance of policies in an SPD, and the way they are combined, help to determine the nature and volume of their potential aggregate employment effects. Not enough is known about the exact relationship between different policy types and their employment impacts to predict optimal combinations, although some commentators believe that the variations in ‘cost per job’ between programmes allow patterns to be derived.155

154Bachtler J and Taylor S (1996) Regional Development Policies in Objective 2 Regions: A Comparative Assessment, Regional Studies, 30(8), 723-732. 155Leon P (1995) The employment impact of regional policies, Paper to DG XVI Conference, Evaluation of European Regional Policy, Berlin, 16 November 1995. Other evidence indicates that much of the variation in employment targets is explained by inconsistent quantification methods (eg. Bachtler J and Taylor S, with Kearney C (1996) Extended Synthesis of Agreed Single Programming Documents in Objective 2 Areas 1994-96, Report to the European Commission (DGXVI) EPRC, University of Strathclyde, Glasgow).

112

b. Communicating the jobs message – measure level At the measure level, selection criteria and target indicators also provide guidance about what is sought. At this level, employment impacts are often listed among other selection criteria, without an indication of their relative importance. Specifically, limited use of employment criteria has been made in Spain and the Netherlands, with detailed employment impact indicators being used instead. The situation is more mixed in Austria, France (with about half the regions lacking employment-related selection criteria) and the UK. In the case of some UK regions and some elsewhere, selection criteria were to be decided by Monitoring Committees subsequent to approval of the SPD. In terms of the types of project where employment selection criteria are stated, this is most widespread in measures which are likely to have direct employment effects, notably business development/productive investment. Finland illustrates a wider commitment to employment criteria, in that these are not only present in measures involving direct company support, but projects in other areas as well (eg. infrastructure).

The table below sets out the employment-related selection criteria which are used by programmes across a variety of main types of intervention.

Table 5.2: Employment-related selection criteria given at the measure level in Objective 2 SPDs Field of Intervention

Selected employment-related selection criteria

Core selection criteria for the whole programme

Overall priority given to job creation (East Midlands, Aquitaine) Every project must fulfil at least one of five job creation criteria and employment effects must be measurable in advance (Finnish programmes) Number and quality of jobs will be taken into account (Ångermanlandskusten) Important but in equal measure to SME networking, internationalisation and reduction in dependency on large firms (Fyrstad) One of six ‘priority considerations’ (Western Scotland and Industrial South Wales) Prominent in the SPD, but in practice judged alongside other important criteria (Nordjylland)

Business development

Creation and maintenance of jobs (Vorarlberg, Lolland, North East England, Cataluña, Bayern) - Including quality of jobs (Niederösterreich, Fyrstad, Ångermanlandskusten) - New permanent jobs, including for target groups (Norra Norrlandskusten) - Including sectoral targeting (Finland) - Including for women (Alsace, Aquitaine) - Including by supply chain development (West Cumbria & Furness) Increased start-ups and/or expansions (Steiermark, Vorarlberg, Alsace) - Including by female entrepreneurs (Aquitaine)

Research and technological development (RTD)

Job creating projects perferred (Alsace, Bergslagen, Catalonia, Niedersachsen) Effect on the quality of jobs available (Niederösterreich) Effect on the number and quality of jobs (Blekinge, Ångermanlandskusten) New and upgraded jobs (Finland) Information society projects opening opportunities for marginalised populations (Rheinland-Pfalz)

113

RTD human resource development

Temporary technology jobs through placements (Centre) New technology jobs in firms through graduate placements (Aquitaine) New jobs, including for target groups (Norra Norrlandskusten) Guaranteeing the viability of firms (ie. jobs maintained) (Fyrstad) Job creation by increasing availability of skills to meet unmet demand (Nordrhein-Westfalen)

Environment Employment impacts (Vorarlberg, Norra Norrlandskusten, Bourgogne) Some programmes state that direct employment impacts of environmental interventions are not the primary objective (eg. in Germany)

Tourism Priority to projects delivering new and/or upgraded jobs (Finland, Alsace, Aquitaine, West Cumbria and Furness, Bergslagen, Champagne-Ardennes) New SME jobs/SME growth (Western Scotland) New permanent jobs for women (Blekinge)

Economic and basic infrastructure

Job years created in construction (Niederösterreich, Bergslagen, Navarra, Baleares, Hessen, Rheinland-Pfalz) Enabling job creation through the use of the infrastructure created (facilitating new SMEs in Aragón, jobs for the disabled in Steiermark, jobs for women in Alsace, general job creation in Zuidoost Brabant and Rheinland-Pfalz) New and upgraded jobs (Finland)

Human resource development

People moving out of unemployment/into employment and/or jobs safeguarded (Lower Austria, Vorarlberg, Luxembourg, Alsace, Basse-Normandie, Hessen and Nordrhein Westfalen (here by provision of skills meeting unmet demand)) Contribution to competitiveness, leading to jobs (Vorarlberg, Navarra) Jobs created/upgraded (Finland) Permanent jobs (Ångermanlandskusten) Employability of trainees increased (Bourgogne) - Including of those threatened by unemployment, esp. women (Aquitaine) Higher priority to projects able to demonstrate employment effects (Western Scotland, Zuid-Limburg) Higher priority to projects where employment is guaranteed following training (Arnhem-Nijmegen) Projects preventing the relocation of existing jobs (Pluriregional Spanish ESF programme)

Community Economic Development

Local job creation (Aquitaine, Champagne-Ardennes, Basse Normandie) Potential for job creation (Western Scotland)

Regeneration Projects able to demonstrate employment effects (Western Scotland, Greater Manchester, Picardie, Bourgogne)

Initiatives in support of territorial employment pacts

Generation of employment through innovative projects exploiting new or alternative opportunities (Groningen-Drenthe)

In terms of the way employment impacts are highlighted as selection criteria, most programmes simply state that specified projects’ employment effects will be taken into account in the selection process. A small number of programmes are more specific, stating that additional preference will be given to projects which:

��improve the quality of jobs (usually by increasing skill levels or the degree to which technology is exploited); or

114

��target sub-sectors of the population to benefit from job creation (eg. ensuring women benefit proportionally from job creation or entrepreneurship schemes, or targeting the unemployed, local residents of disadvantaged areas or the disabled).

Others, as in the box below, aim to provide more detail as to what will be required.

Where Structural Funds are being used to supplement existing national and regional business aid schemes (eg. in Austria and, to an extent, Germany and France) these offer robust mechanisms for promoting and then following through employment creation impacts. In Austria, the Regional Innovation Premium (RIP) and the ERP Regional Programme, both of which absorb an important proportion of Objective 2 Structural Funds (23.3 percent in Steiermark and 16 percent in Niederösterreich), both place job creation high among their own project selection criteria. This priority is followed through into the design of the schemes, which in turn brings considerable positive benefits for the Objective 2 programme’s employment impact. The RIP, in particular, offers an additional ‘jobs premium’ which can double the award rate. The ERP Regional Programme, in turn, introduced a new job incentive in 1997, whereby firms increasing employment by ten percent or more, and maintaining this over three years, receive an additional interest rate reduction of one percent over this period. A further means to ensure quality job creation under the ERP Regional Programme has been to raise award rates by ten percent for firms undertaking training and employing apprentices. The schemes apply a quality screen, not counting new seasonal jobs for example.

EMPLOYMENT-RELATED SELECTION CRITERIA – HIGHLY DEVELOPED APPROACHES

In some cases, the SPD provides a consistent framework helping applicants to think through their projects. In Finland, for example, where employment impacts are now the main selection criteria, projects are informed that they have to fulfil at least one of five specified employment-related criteria to be eligible:

• = Significant direct effects on employment at start-up stage

• = Indirect effects at start-up stage

• = Jobs created during the operating stage

• = Corporate investment creating new jobs or maintaining existing ones

• = Training, research or development projects improving the employment potential of participants

In the Nordjylland programme, the selection criteria flesh out what is sought from projects in terms of employment. The SPD states that while projects should at the very least safeguard jobs, the highest priority will be given to job creation (in the short, medium and/or long term). It also stipulates that the jobs created should be ‘real’ new jobs and not relocated jobs – ie. there should be a net increase in local employment.156

Promotional summaries of SPDs are also an important means of communication. In some cases, greater emphasis has been placed on job

156 Nordjylland Single Programming Document 1997-99.

115

creation in publicity materials than in SPDs themselves, in order to influence applicants. The Nordjylland 1997-99 programme summary is a case in point, giving more prominence to job creation than either the previous programme’s brochure or the current SPD. Further channels of communication are programme newsletters (eg. in Western Scotland, Industrial South Wales, Arnhem-Nijmegen and Limburg), press articles (Aquitaine) and project application forms.

Overall, while some programmes already perform well, most could give more emphasis to job creation in all the sections of their programming documents and in their wider publicity materials. Post-2000, this means ensuring that the theme is integrated throughout the programming supplement, and not only visible in the main programming document. Reminders that potential employment outcomes increase a project’s prospects of being approved and could lead to higher rates of intervention offer a powerful means to influence applicants.

5.4.3 Equipping Applicants to respond The second stage after awareness-raising is to assist applicants to respond – in terms of the detailed design of their projects or the way they are presented. Some public sector organisations are reasonably well-equipped to quantify the potential employment effects of their projects, even if they do not always have robust methodologies systematically in place.157 However, it remains worthwhile for programmes to supply guidance, even where applicants’ abilities to set targets are reportedly improving. The benefits are threefold. First, targets derived from methodologies known to programmes will be consistent with others derived in the same way. This enables projects to be compared at the application stage, and reliable information to be gathered across whole programmes during implementation. Second, the use of known methodologies can counter applicants’ tendency to overestimate employment effects, or to present estimates which have no methodological grounding. Both tendencies are still widespread. Transparency in the assumptions from which targets are derived is extremely important, enabling divergences between targets and actual outcomes to be explained by a discussion of the accuracy of the assumptions. Third, encouraging applicants to think through their employment targets and how they will be produced makes them active agents rather than passive observers, alerting them to the direct impacts of their project as they happen.

Strategic aims and selection criteria within SPDs can, if sufficiently developed, provide guidance to applicants. Beyond the SPD, many programmes offer individualised technical support. This can help applicants to think through why and how their project might impact on employment, and thereby to generate targets which are realistic and transparent.

In some cases, the quality of advice which can be given to applicants is limited by advisors having insufficient expertise or insufficient information on the performance of similar past projects. Improved monitoring systems, and detailed evaluation studies are helping to generate data to address the latter

157 Turok I (1997) Evaluating European support for business development: evidence from the Structural Funds in Scotland, Entrepreneurship & Regional Development, 9, 335-352.

116

deficiency. However, the former needs to be addressed through training and continuing professional development within the economic development community.

ASSISTANCE TO APPLICANTS IN TARGET SETTING

In Austria, many applicants consult their local ‘regional management office’ when considering making an application under Objective 2. These offices see job creation as a key part of their economic development role and make this central to discussions. In Steiermark, for example, following an initial consultation with potential applicants, the regional management office gives them a form to fill in which helps to develop their ideas further. The form collects information on the nature of the project, the financial framework, goals, timeplan and jobs to be created. When the form is returned, the employment-related issues are explored in more detailed discussions to ensure that employment outcomes will be positive, or at the very least neutral. Where jobs could be lost through an investment, a rethink of the project plan is encouraged.

Nordjylland provides a further good practice example. During the advice phase, this programme’s secretariat informs applicants that a higher rate of intervention might be made available to particularly good job creation projects (rising from 18 to 20 percent for investment projects and from 40-45 to 50 percent for knowledge projects). At the same time, they avoid exaggerated targets by verifying that they have been thought through in detail - for example, ensuring that the additional employment costs of job creation have been taken into account in budget forecasts. If the final presentation which is submitted is too shallow, then the programme secretariat asks for further details. In Finnish programmes, too, applicants are asked not only to quantify potential employment impacts, but also to substantiate them.

5.4.4 Job Creation and Project Selection At the project appraisal and selection stage, no systematic study has assessed how job creation is actually applied as a selection criteria alongside others. Judging projects solely on the number of jobs they will create or safeguard is widely agreed to be short-sighted. In order to deliver decisions which ‘make sense’, programmes frequently nuance the criteria to prioritise jobs which are durable, of high quality, highly skilled, or which are made available to disadvantaged sections of the potential workforce. In addition, they recognise the indirect contribution of some apparently employment-poor projects to longer-term improvements to the employment situation through their contribution to modernisation or diversification, or their ability to raise innovation rates and the exploitation of new technologies.

Some programmes (eg. Ångermanlandskusten) have sought to revise down the current degree of emphasis placed on employment impacts, in recognition of the fact that the projects creating jobs are not necessarily those which contribute most to long-term regional development. Over-reliance on employment targets as indicators of project quality can have some perverse effects: projects of dubious quality but which promise significant employment impacts can in some cases be politically difficult to refuse, while essential modernisation projects which will lead to rationalisation can be difficult to justify. At the same time, it is also becoming increasingly difficult in some cases to fund projects which, whilst worthwhile, are unlikely to have employment impacts (Päijät-Häme).

117

One appraisal system which is able to bring out the value of projects with limited direct employment impacts but which strengthen other employment-generating actions, is that which is being introduced under the English ‘action plan’ and ‘package’ approach to programme implementation.158 Groups of projects under the same priority heading and targeted at the same part of the eligible area are appraised together by sub-groups of partners, at the early stages of programme implementation. This enables potential synergies between projects to be demonstrated, and thus the value of individual projects to be assessed in a more rounded way than if they were submitted individually, over a longer time period and on a more ad hoc basis, to the programme-wide appraisal process.

To make project selections on the basis of employment outcomes, the realism of targets on which the quality of a project is judged needs to be verified. It is helpful if applicants state the assumptions on which they have based their estimates, and are encouraged to employ methodologies and terminology approved by the programme (see above). If targets are unrealistic or unclear, internal estimates should be made by project appraisers, or additional consultations held with the applicant to derive improved figures.

Two further issues which project appraisal systems need to be able to take into account are deadweight and displacement. These are much more than peripheral concerns: some commentators argue that robust approaches to project selection which minimise deadweight and displacement are critical to raising the potential employment impact of Structural Funds interventions. This was highlighted, for example, in the recent Yorkshire and Humberside interim evaluation.159 Minimising deadweight and displacement maximises the resources spent on projects which are truly additional to the regional economy, and would not otherwise have been undertaken. Failing to do this means that with a displacement and deadweight effect of, for example, 30 percent, three ECU in every ten have no net impact and so are wasted. To spend these three ECU on genuinely additional projects would ensure more net economic impacts immediately - and would effectively expand the programme’s available resources for job creation by almost a third. Appropriate information, including a clear understanding of the circumstances in which displacement and deadweight effects are likely to be strong, is required to enable them to be identified.

5.4.5 Monitoring of Employment Outcomes The monitoring process provides an invaluable means to record employment outcomes on an ongoing basis, generating statistics which can feed back, in turn, into enabling the employment targets of whole programmes and future projects to be more realistically gauged. However, there are several impediments to monitoring systems achieving this. First, many programmes’ monitoring systems focus on recording financial rather than output and impact data. Second, even where impact data is also collected, there is a mismatch

158 Taylor S and Downes R (1997) Meso-Level Partnerships and Structural Fund Implementation, IQ-Net Thematic Paper, 3(3), EPRC, University of Strathclyde, Glasgow. 159Centre for Training Policy Studies (Eds.) (1997) Review of the Yorkshire and the Humber Region Objective 2 Programme 1994-96: Final Report of the Programme Wide Appraisal, Report to the Government Office for Yorkshire and Humber.

118

between the long timescale over which the full employment effects of many projects manifest themselves, and the duration of monitoring, which usually finishes on paying the final tranche of project funding. Ex post evaluation can, of course, be used both to retrospectively collate impact data, and to ensure that impacts achieved beyond the project’s end date are identified. However, such data is generated too late to help hone the delivery of the programme which supported the relevant projects or the formulation of its successor.

To optimise the utility of the monitoring system in supporting and reporting on job creation, the following recommendations can be made.

��Feedback should be requested from implementers about project outcomes while the project is being implemented. Linking its submission to payments helps to ensure it is returned consistently, while advice on target-setting provided at the application stage can help to improve its quality.

��The reliability of the data should be verified using project visits.

��Projects could routinely be followed up one year after completion to capture any further medium-term impacts, including on employment. This could become a universal practice or could be restricted in the first instance to those project types where the main employment impacts are likely to appear beyond the end of project implementation.

Ex post studies are the most important in revealing what the overall gross and also net impacts of programmes are likely to have been on employment. They are limited in this by some inevitable methodological problems (eg. separating the impact of the programme itself from the effect of the wider policy environment and other influential factors), but nonetheless are able to indicate in broad terms what outcomes have been, and at a sufficient distance in time for more impacts to have been manifested than monitoring can capture. ‘Top-down’ approaches dominated the evaluation of European Structural Fund interventions until the late-1980s, assessing the aggregate effects of expenditure on the regional economy, in comparison with the counterfactual (what would have happened in the absence of intervention). Progressively, however, these evaluations have been supplemented or superseded by ‘bottom-up’ studies, often based on individual project data, supplemented by interview evidence and economic analyses. This has taken place as ‘top-down’ studies have become more complex and smaller programming areas have made such studies less feasible. The bottom-up approach aims to build up a picture of programme impacts from the micro level. Many bottom-up evaluations have also taken place in the course of interim or thematic evaluations. One of their most valuable aspects is that, in investigating employment impacts on a project by project basis, they can also generate qualitative insights into the reasons for any patterns observed.

The ability of any evaluation study to answer employment-related questions depends on several factors, including the skills of the evaluator. The better informed those commissioning evaluation studies are about the practice of evaluation, the better able they are to ensure that robust and useful information is generated by it. In some cases, appropriate specialised input across a range of fields can best be ensured by awarding broad-based evaluations to a series

119

of evaluators, each with appropriate skills in specific tasks. There are several examples where this has happened.

A second factor helping to ensure that employment impacts can effectively be evaluated is if programmes have taken a systems approach from the beginning, thinking through the impacts of measures from the programme design stage onwards, undertaking a careful and transparent target-setting exercise based on standard definitions and as many solid assumptions as possible, and following this through into the generation, appraisal and then monitoring of projects.

5.5 Conclusions and Recommendations Objective 2 areas are defined primarily on the basis of negative employment change. As such, employment creation has always been one of the main aims of these programmes, and the primary indicator for assessing their achievements. In every reprogramming exercise since 1989, the emphasis placed on job creation has grown steadily. This objective has become increasingly prominent as high-level political commitment to addressing persistent and worsening European joblessness at EU and Member State levels has grown. This commitment was reflected in: the so-called ‘Essen Track’ proposals which called on Member States to take action in defined fields to increase employment; a major new chapter of the Amsterdam Treaty; and, most recently, the introduction of National Action Plans for Employment, built around the four pillars of employability, entrepreneurship, adaptability and equal opportunities, which further develop the Essen employment programmes. By the time Objective 2 programmes were being developed for the 1997-99 period, job creation had become the overriding Commission priority.

Over the last decade, the content of Objective 2 programmes has been adapted in three main ways to respond to the employment imperative. First, their policy content has evolved. While the overall aim of Objective 2 programmes is to improve the employment situation, most have moved towards balanced combinations of interventions which will help to achieve three separate employment-related objectives: to establish the conditions for long-term economic growth by raising competitiveness; to increase the job intensity of growth; and, to take a pro-active approach, where appropriate, to social solidarity.

In terms of specific policy orientations, investments in basic infrastructure which offer relatively poor value for money either in terms of permanent employment created or impacts on competitiveness have been reduced over time. These policies have largely been substituted by measures releasing the potential of indigenous businesses, either by intervening directly in firms (especially SMEs) or - more durably – by improving the collective business environment (eg. by establishing incubator units, technology brokers, etc). Business development schemes have also progressively been refined as experience has been gained to ensure value for money, eg. by requiring them to more clearly establish demand and to monitor supported firms more closely. The ESF by its nature creates relatively few jobs directly, but serves an important complementary role in getting people into the new jobs being created or preventing their passage into unemployment. ESF interventions

120

have also been refined and adapted in response to the need to address unemployment in more focused ways, with increased resources being dedicated to tailoring human resource development to the needs of specific firms and individuals, and investments in labour market intelligence services ensuring that the skills and qualifications delivered under programmes are genuinely those in demand. Community economic development policies have also become more prevalent, responding to the need to consider not just the number but also the quality and distribution of jobs where there are concentrated problems of labour market exclusion.

The second response of Objective 2 programmes to the employment imperative is that they have almost universally made this theme more explicit. In programming documents, its importance is now spelt out in the regional profile, the strategy statement and, at priority and measure level, in rationales, selection criteria and impact targets. The most striking presentational changes took place between the 1994-96 and 1997-99 programmes as a result of Commission guidance. In addition, Commission level policy documents and initiatives have also helped to raise the profile of employment, especially its role in Objective 2, for example through the ‘Jobs Challenge’, an international competition to distil transferable lessons from projects which have been particularly successful in creating employment.

The third response has been action to improve the ability of Objective 2 actors to predict and then demonstrate employment outcomes, using enhanced prior quantification and monitoring. Across the 1989-93 programmes, only one in five programming authorities included employment targets in programming documents. Coverage and quality have both improved successively since then, with each programme engaged in a process of learning and building on past experience, helped by the (albeit erratic) dissemination of methodological frameworks elaborated with the support of the EC. A small number of Member States still resist detailed prior quantification because of the uncertainties inevitably involved, but in general the principle is accepted and the necessary expertise to apply it is accumulating.

By the 1997-99 programmes, three-quarters of SPDs were providing detailed, measure-level estimates of employment effects, the approaches taken to supplying them varying from relatively sophisticated, comprehensive exercises commissioned to address the issue across groups of programmes, to very pragmatic solutions addressing primary weaknesses in single programmes. In addition, the definitions used in setting out targets have become more consistent and robust. In a small number of instances, targets have become more transparent, with SPDs setting out the assumptions on which they are based.

In addition to the three changes in the presentation and content of Objective 2 programmes, their delivery has also been adapted in an effort to ensure that employment impacts can be achieved and demonstrated: first, improved methods have been being used to communicate the jobs message to potential applicants; second, support has been given to applicants at the project development stage to help them to maximise potential employment impacts; and, third, monitoring systems have been improved to provide more feedback about the actual employment outcomes of projects and measures.

121

The need to improve the employment impacts of Structural Fund programmes will only grow in future. The following recommendations set out some principles by which they can respond.

5.5.1 Job Creation Recommendations

All programmes should at this stage be required to achieve a threshold level of quantification of targets, including for employment - both in terms of their quality and their coverage.

Establishing robust output and impact targets is important to the overall success of programmes, helping to lay a strong foundation for target-setting by individual project applicants, sound and consistent project appraisal, the effective monitoring of project outcomes and the subsequent evaluation of overall programme performance.

Progress in target-setting is still uneven. However, practice is by now sufficiently established that every programme should in future be able to achieve complete coverage of good quality gross targets which employ consistent definitions and are transparent in their methodological basis and underlying assumptions. To respond to the gender mainstreaming obligation, employment targets should also be disaggregated by sex, and preferably informed by current baselines and the previous outcomes of economic development interventions. Ideally, SPDs would go beyond gross targets to set out estimated net employment impacts, when deadweight, displacement and double counting are taken into account.

Where target-setting practice is not yet well-established, programmes can benefit considerably from undertaking a dedicated exercise to improve their responses, using explicit methodologies to derive targets and exploiting past monitoring data to the full to establish realistic assumptions. This can be undertaken at either programme or national level, the latter method promising efficiency benefits and enabling common approaches to be established which then simplify the design of monitoring systems and enable easier comparison between programmes.

In designing and implementing future programmes, employment outcomes should be prioritised, but in a balanced way, taking into account potential impacts over the short, medium and longer term and the quality, durability and distribution of the jobs to be created.

The results of past monitoring and evaluation mean that programme actors are now better informed than ever about the employment potential of different types of economic development intervention. This information should facilitate a fuller debate during the programme development process about policy priorities. In promoting job creation, it will be important that programmes take a balanced approach, considering direct and indirect impacts arising in the short, medium and longer term, and including in the assessment

122

not just the number of jobs likely to result from interventions, but also their quality, durability and, where appropriate, distribution. In selecting policy priorities, there are specific trade-offs to be made, for example for RTDI policies which, while they often have limited employment impacts in the short term will have a greater impact over the medium to long term by raising regional business competitiveness.

To ensure that programmes achieve their employment potential, a rounded approach to employment outcomes should be followed through into subsequent project appraisal and selection systems.

The ERDF primarily addresses demand-side issues in the labour market, creating or safeguarding jobs. Programmes must continue to seek coherence with ESF interventions – whether they are under Objective 2 or operated in parallel – which respond to the supply side, by training people in appropriate skills to take up the vacancies being created.

Objective 2 reconversion programmes involve the modernisation of existing sectors and diversification into new ones. It is essential to equip the workforce to respond to the new opportunities this creates. In ensuring this, human resource development policies are an essential complement to the job-creating interventions of Objective 2 ERDF measures.

Whether ERDF and ESF interventions are undertaken in the same programmes or in parallel, it will be necessary to make specific provisions to ensure that wide and ongoing communication, co-operation and co-ordination take place between regional development actors in both spheres. Among the most efficient solutions may be to establish all economic development programmes operating in a given region within the context of an overarching region-wide strategic economic development framework. The development of improved labour market surveillance systems at the regional level is another concrete action which can help to share information between all regional development actors.

Among some human resource development partnerships, whether it be within or outside Objective 2, there is still too little emphasis placed on ensuring that training meets real unmet demand for skills, and so is economically relevant. Instead, ‘off the shelf’ qualifications are sometimes delivered which may not lead to employment outcomes. In order for the ESF to contribute to the full to the pillars of the National Action Plans for employment (NAPs) – especially employability and adaptability – the emphasis in some cases has to change.

123

Programmes could significantly improve value for money in terms of their job creation outcomes by filtering out policies and projects likely to have limited net employment impacts because of high deadweight or displacement effects. They should place a premium on developing the necessary systems to do this.

Many projects which, while their gross employment impacts may be considerable, promise negligible net employment outcomes once the impact of deadweight and/or displacement are taken into account. It has been estimated that up to a third of programme resources are currently used to support projects which either could have gone ahead with less or no support, or which only serve to displace existing activities within a regional economy. Ensuring that such projects are excluded, both through policy design and project appraisal, would significantly increase the resources available to other projects promising greater net employment effects. This in turn would increase the overall employment effectiveness of the programme. Such filters may have to be adapted to allow projects which, while they are likely to have high displacement effects, contribute to other programme objectives, for example entrepreneurship schemes which enable the reintegration of excluded groups into the labour market.

Further efforts should be made to equip programme applicants to respond to the jobs challenge.

To underline the significance of employment outcomes, SPDs and related applicant information should place greater emphasis on this, following the theme through consistently at all strategic levels (objectives, priorities and measures). In the project development stage, other specific advice to applicants should follow the message through. The principal means of providing advice to applicants are: documentation (eg. application form notes and partner ‘manuals’), presentations and seminars, and individual consultations with programme secretariats or other relevant advisors. Providing assistance to applicants in setting employment impact targets helps to improve their quality and consistency, so making projects more comparable at the selection stage. It also ensures that project implementers are better equipped to report accurately on the impacts of their interventions.

The employment-related lessons of monitoring and evaluation should feed continuously into improving all stages of programming.

Ongoing programmes should actively generate information about the employment outcomes of previous interventions, not only through evaluation but also by adapting monitoring systems to capture as much information as possible from ongoing and recently completed projects. The lessons derived

124

from this information should then be fed back into project development, appraisal and selection systems, to ensure that outcomes are maximised.

125

6. RESEARCH, TECHNOLOGICAL DEVELOPMENT AND INNOVATION

6.1 RTDI Policy in Regional Development The role of research, technological development and innovation (RTDI) in the promotion of balanced and self-sustaining development is becoming increasingly widely recognised. Fully exploiting existing technologies and being at the leading edge of new advances are critical elements for individual economies to become and remain competitive in the global economy. The importance of policies promoting technology and innovation at national level is widely accepted in EU Member States and by the EU itself. The rationale was summarised by the UK 1993 White Paper ‘Realising our Potential: A Strategy for Science, Engineering and Technology’, which stated that: ‘a close partnership and better diffusion of ideas between the science and engineering communities, industry, the financial sector and Government are needed as part of the critical effort to improve…national competitiveness and quality of life’.160

The rationale for excellence-driven RTDI support at national and EU levels is clear, particularly as Europe has one of the lowest levels of R&D expenditure relative to its GDP. In 1996, the EU Member States collectively invested 1.8 percent of their GDP in R&D, contrasting with 2.8 percent in Japan and 2.5 percent in the USA.161 At the same time, there is also a strong case for RTDI support to be spatially targeted as an element of regional development policy, helping to address not only the economic gap between the richest and poorest regions and Member States of the EU, but also their widening technology gap.162

While the cohesion gap is estimated to be 2:1, the technology gap between Member States, measured as expenditure on R&D as a proportion of GDP, has been calculated as 5:1 between Sweden at one extreme and Greece at the other.163 Between regions, the difference is much more marked, at around 98:1 (Berlin: Dytiki Makedonia).164 Strong clusters of RTD-driven firms and supply facilities have emerged displaying a high degree of specialisation and expertise, and collaborating within exclusive networks which leave little opportunity for outsiders.165 The virtuous circle of self-perpetuating development found in such ‘islands’ is in stark contrast with the vicious circle which can establish itself in the less dynamic regions. Since mainstream public and private RTDI expenditure is concentrated by definition in the most technologically advanced regions, both at Member State and EU levels, this not only maintains but further widens these disparities.

160 HMSO (1993) Realising our Potential: A strategy for Science, Engineering and Technology, Cm 2250, HMSO, London. 161 CEC (1997) Second European Report on Science & Technology Indicators, Commission of the European Communities, Luxembourg. 162 CEC (1997) op cit. 163 The Second European Report on Science & Technology Indicators uses the equation ‘gross estimated expenditure on R&D/gross domestic product (GERD/GDP)’ to calculate the technology gap. 164 CEC (1997) Second European Report on Science & Technology Indicators, Commission of the European Communities, Brussels, p.342. 165 Hilpert (1992) Archipelago Europe.

126

Various other indicators can be used to measure technology or innovation gaps, including patents per inhabitant and numbers employed in R&D. However, each only represents a facet of the real disparity, which relates to the functioning of an ‘innovative milieu’ and its ability to lead to economic outcomes in terms of marketable products and processes. The most innovative regions benefit from a dynamic of interaction between firms, RTD suppliers, policy organisations and the market which facilitates product and process innovation. Technology diffusion is efficient, and demand for technological innovation high and sustained. Such regions also benefit from human and social context in which the process of learning and innovation is driven by collective and tacit – therefore unique - knowledge. These patterns mean that the technology gap in reality relates to ‘the level of what has been described as social capital – the network of relations between key policy makers, policy delivery organisations and business’.166 This places the focus as much on relationships and the operation and nature of systems and networks as on individual actors or single indicators of activity. The ultimate challenge for RTDI policy in support of regional development is to encourage the behaviours and capacities which enable a self-sustaining dynamic of world-class innovation to emerge which is fully translated into regional and business competitiveness.

At a more basic level, less ambitious but also important objectives in regionally driven RTDI policy include ensuring the diffusion of established technologies to backward firms in lagging regions – innovation to the firms in question if not in a wider context. A further issue is how to ensure the widespread adoption of information and communication technologies, changing the way business operates and competes, and potentially eliminating aspects of geographical disadvantage.

6.2 RTDI Policy and the EC RTD was one of the original economic areas earmarked for support from the EC, first mentioned in the Treaty of Rome (Article 130, f–p). Although various RTD initiatives were undertaken by the EC before the mid-1980s, it was not until the launch of the ‘Framework Programmes for Research, Development and Demonstration’ in 1984 that the first co-ordinated and strategic approach to RTD in the Community was established. The Framework Programmes provide a structure and priorities driving transnational EU RTD programmes. The first ran from 1984-1987, and the fifth is currently underway.

From the second one, these Framework Programmes have included a dimension encouraging RTD activities in less-favoured regions. This recognised that, without support to promote RTD and innovation, less-favoured regions would continue to diverge in prosperity from the core regions.167 The regional dimension was particularly strong in the Fourth

166 ADE, ZENIT, Enterprise (1999) Evaluation of Research, Technological Development & Innovation Related Actions under the Structural Funds (Objective 2) – Final Synthesis Report, DG XVI of the European Commission. 167 Goddard J, Charles D, Howells J, and Thwaites A (1987), Research and Technological Development in the Less Favoured Regions of the Community (STRIDE), Commission of the European Communities, Brussels.

127

Framework Programme (1994-98), following an EC report published in 1993 which highlighted the relationship between regional and RTD policies.168 The Fourth Framework Programme included:

��support for intermediary agencies which can assist SMEs in less favoured regions to integrate innovation;

��support for the creation or evaluation of Regional Innovation and Technology Transfer Strategies and Infrastructures (RITTS);

��assistance for science parks; and

��a pilot action enabling the co-financing of technological demonstration and transfer of know-how to Objective 1 and 2 regions.

In parallel to the growing regional dimension of EU RTD policy, the EU’s regional policy has been emphasising research and innovation. The Structural Funds have now become a key instrument for promoting regional RTD support, particularly since the introduction of Structural Fund programming in 1989. Before 1988, under 200 MECU of Structural Funding was allocated to research and development, an allocation which rose to four billion ECU for the 1989-1993 programming period.

In addition to support under CSFs, RTD co-financing was directed to under-performing areas through the Community Initiative STRIDE. This was a specific RTDI initiative, funded by the Structural Funds, with a Community budget of 460 MECU for 1998-93. Both Objective 1 and Objective 2 regions were eligible, the Objective 2 allocation being 20 percent of the total (around 90 MECU).169 The aim was to increase participation in RTDI networks, as well as to promote links between research centres and industry, reflecting a new approach to increasing innovation by establishing an interaction-rich local or regional milieu. Among the regions which took part were Berlin, Bremen and North East England. Some STRIDE projects led to durable change: for example, one of the technology transfer networks set up through STRIDE in North East England still exists, and participants of the project are still involved in RTDI initiatives in the region.

The eighth Annual Report on the Structural Funds dedicated a special focus to research and technological innovation, pointing to the persistence of technology-related disparities in the EU and the continued requirement for RTD support in the face of rapid technological change. The form of this support, and the context in which it is deployed, have been the subject of considerable enquiry, with the realisation that a process of strategic RTD thinking helps to target interventions more effectively.

“In the absence of a regional strategy to promote innovation that can help to create an ‘innovative environment’ through more and better cooperative links between those working in innovation in the region, an injection of public funding into the system will not result in a

168 Commission of the European Communities (1993) Cohesion and RTD Policy – Synergy Between RTD Policy and Economic and Social Cohesion, COM (93) 203, Brussels. 169 CEC (1995) Research and Regional Development, Commission of the European Communities, Brussels. (Available at: http://www.inforegio.org/wbdoc/docgener/presenta/rd/rdhome/rd_en.htm)

128

substantial increase in the contribution made by that funding to regional economic development.”170

The following sections review the increasing and evolving role which RTDI policies have played to date in Objective 2 programmes, going on to highlight some principles of best practice in terms of policy design and delivery.

6.3 RTDI and Objective 2 Objective 2 programmes operate in economies which have faced industrial decline often in a narrow range of outmoded sectors. They need to confront complex problems in raising their innovation and technology levels, and establishing more knowledge-based, innovative and interactive environments. Since these areas frequently include metropolitan centres, they do tend to have a potential supply of innovations in the form of endowments of R&D infrastructure (eg. in higher education institutions), but this is usually more strongly oriented to pure than applied research. Where there are linkages with firms, they tend to be in the context of leading-edge projects for prestigious companies. Such collaboration is more cost-effective than resource-intensive projects with small, technologically backward firms, and is more likely to fit with the achievement of organisational core objectives.

In such regions, levels of demand for technology transfer and innovation tend to be low. The business sector typically comprises followers not leaders. Business sector expenditure on RTD is typically small with few firms (especially SMEs) having their own RTD capacity. Inward investment, even from technologically advanced firms, may only bring in low-value ‘assembly’ functions. A large proportion of firms tend to function without the benefit of established technologies, such as computerisation of administrative functions, that are already taken for granted by more advanced firms. They may be unaware of technological trends, and lack knowledge about the types of support available to help raise their technology levels.

The lack of dynamism and belief characterising the business community is often replicated in the political arena. A culture of dependency is characterised by a pervasive belief that the region will never be able to compete nationally or globally on equal terms. Other behavioural issues impeding a dynamic of innovation include firms’ inexperience with collaboration and networking, which can be reinforced by protective stances vis-à-vis markets and knowledge and hierarchical or unresponsive internal structures. Agencies may experience difficulties in engaging meaningfully with the business community to promote RTDI, this process being resource intensive and requiring wide-ranging expertise in a rapidly changing environment.

In this context of mutually reinforcing barriers to innovation and technology transfer, Objective 2 programme managers and partners have had to undergo a learning process, accumulating knowledge and experience of the RTDI field and adapting their approaches to innovation over successive programming periods. The following sections give an overview of the development of RTDI

170 Landabaso, M (1997) The promotion of innovation in regional policy: proposals for a regional innovation strategy, Entrepreneurship and Regional Development, Vol 9.

129

priorities and measures during the three Objective 2 programming periods, 1989-93, 1994-96, and 1997-99. They illustrate how, as understanding of the innovation process has increased and evolved, the approaches taken and the frameworks informing them have become more strategic, sophisticated, diverse and targeted.

6.4 The 1989-1993 Programming Period The 1988 reform of the Structural Funds moved support away from a project-based approach towards strategic programming. Increasing the competitiveness of Objective 2 areas was linked to increasing innovation and technology levels, and led to a broad range of RTD funding being included in programmes. Interventions included training, infrastructure for RTDI (laboratories, research centres and science parks), the creation of incubator and resource centres, technology advisory services, university and industry collaborations and linkages promoting innovation and technology transfer.

Objective 2 programmes made significant funding available for RTDI promotion. The overall allocation of Objective 2 funds to this field during the 1989-93 period has been calculated as nine percent of total funding at priority level.171 As Table 6.1 illustrates, the overall pattern of RTDI allocations and expenditure masks significant variations between and within countries. Higher allocations on RTDI priorities are evident in Denmark, Belgium, Germany and Spain (averages of 27, 13, 11 and 10 percent respectively), range down to no (priority level) expenditure in the Netherlands and Luxembourg, just four percent in the UK and five percent in France. The table shows that variations between individual programmes were also wide. RTD was not consistently a part of programmes: Belgium and Denmark were the only countries where all Objective 2 programmes allocated some priority expenditure towards RTDI, in comparison with Germany where only two of the seven Objective 2 areas, Bremen and Nordrhein-Westfalen, included RTD priorities. Overall, participation was low compared with later rounds of programming, with 23 of the 60 Objective 2 regions having no RTD priorities. The highest level of RTD expenditure was in Vestlolland, where a total of 51 percent of Objective 2 expenditure went to RTDI related priorities, exceeding the second highest RTD allocation of 22 percent in Pays de la Loire.

Midway through the 1989-1993 period, programmes were reviewed, resulting in a shift in funding allocations in some programmes, in some cases favourable to RTDI (eg. Lorraine, where allocations dedicated to new technologies and communications were increased) and sometimes not (eg. the Belgian region of Hainaut, where RTD allocations were cut in favour of support for SMEs).

The reasons given for the inclusion of RTD priorities within programmes and the contexts in which they were deployed were as diverse as the amounts allocated between regions. Some information is given on the context in which RTDI measures were being implemented, but it is generally difficult to speak of a ‘strategy’ or even clear objectives driving the RTDI content of the 1989-93 programmes. Frequently, the range of interventions simply reflected the policies and initiatives already operating in regions, rather than being custom-designed as an integrated group during the Objective 2 programme preparation

171 CEC (1995) op. cit.

130

process.In Spain, RTD measures aimed to redress the extremely low levels of technological innovation amongst firms located in Objective 2 areas and to meet a need for further investment in RTDI infrastructure. In the UK on the other hand, where levels of foreign direct investment were high in comparison with other areas of Europe, RTD measures in the North-East England and South Wales Objective 2 programmes operated in a context of limited local RTD investment by foreign-owned companies and restricted local linkages. Toscana was addressing similar problems.

Table 6.1: RTDI Actions at priority level in Objective 2 programmes, 1989-93 (%) Member State

Average % RTDI allocation

Maximum Objective 2 allocation to RTDI

Minimum Objective 2 allocation to RTDI

Region % Region % Belgium 13 Liège 21 Limburg Denmark 27 Vestlolland 51 Nordjylland 21 France 5 Pays de la Loire 22 Bourgogne, Haute-

Normandie, Langedoc-Roussillon, Lorraine, Midi-Pyrénées, PACA, Rhône-Alpes

0

Germany 11 Nordrhein-Westfalen

18 Berlin, Emden, Reinland-Pfalz, Saarland, Salzgitter

0

Italy 7 Piemonte 17 Lazio, Marche, Toscana, Umbria, Valle d’Aosta

0

Luxembourg 0 n/a 0 n/a 0 Netherlands 0 n/a 0 n/a 0 Spain 5 Aragon & Madrid 8 La Rioja 0 UK 4 Midlands 10 Eastern England, West

Cumbria 0

Source: Derived from Ernst & Young (1997)172

Measuring the impact of 1989-93 RTDI policies on levels of innovation and technology was is difficult. Assessment was impeded in the ex post evaluation by the lack of appropriate baseline indicators and targets, and the partial availability of even basic monitoring data. It is likely that some difficulties were experienced in implementing some of these policies, since less was spent over the period than was allocated (5.8 percent spent at priority level as opposed to the nine percent allocated). In addition, some of the weaknesses of the 1989-93 programming period,173 such as poor articulation between the ERDF and ESF, would have impeded the success of RTDI priorities.

While it could not give a rounded appraisal of impacts, the ex post evaluation of the 1989-93 Objective 2 programmes was able to highlight some lessons arising from the inclusion of RTDI support in these programmes. First, it concluded that one of the most successful methods for raising innovation and technology transfer levels in many Objective 2 areas was the creation of

172 Ernst & Young (1997) Ex-Post Evaluation of the 1989-1993 Objective 2 Programmes: Synthesis Report, DG XVI Series: Evaluation and Documents, No 4, Commission of the European Communities, Brussels. 173 Ernst & Young (1997) op. cit.

131

supporting infrastructure, such as technology centres. This worked best where it built on existing regional RTD strengths, successful examples including Nordrhein-Westfalen and Limburg. A second area of transferrable potential comprised the provision of business support for new technology based start-ups, where commercial acumen rather than technological expertise was lacking. Nordrhein-Westfalen was one of the programmes to establish this link. Third, networking initiatives were also concluded to have been valuable actions for technology and innovation exchange. In the UK, both South Wales and North-East England (due largely to their earlier participation in the STRIDE programme) created successful RTDI networks.

6.5 The 1994-96 Programming Period In the 1994-96 programming period, the emphasis on RTDI increased. There was a reorientation of policy priorities, recognising that many regions receiving Objective 2 support had achieved a satisfactory level of basic research infrastructure (research centres and universities), and that more applied interventions had to be prioritised in order to ensure stronger economic impacts.

According to Commission documentation,174 allocations during the 1994-96 period amounted to around 10 percent of the total. Another study, examining the allocations at measure level,175 calculated the figure to be 14 percent. There is a strong possibility that the true figure could be even higher, as many measures include elements relating to RTDI promotion, especially mixed business support measures.

As with the previous period, there was considerable variation among programmes between and within Member States, with little in the way of identifiable funding patterns. At national level, Commission figures (see column two in Table 6.2) showed that Denmark was once again prominent in the importance it attached to RTDI measures, allocating 36 percent of its Objective 2 resources to this area of support. Large allocations were also made by the Swedish and Austrian programmes when their 1995-99 programmes were approved (28 and 20 percent respectively). Of the other EU Member States, allocations were also fairly high in France and Luxembourg, at around 14 percent. The differences in the national allocation figures between columns two and three in Table 6.2 reflect the fact that the first was derived from analysis at the priority level and the second from measure-level data.

174 CEC (1995) The new Regional Programmes under Objectives 1 and 2 of Community Structural Policies, Commission of the European Communities COM (95) 111 final, 29.3.95, Brussels. 175 Bachtler J, Taylor S with Kearney C (1996) Extended Synthesis of Agreed Single Programming Documents in Objective 2 Areas, 1994-96, Report to DG XVI of the European Commission, EPRC, University of Strathclyde.

132

Table 6.2: RTDI Actions in Objective 2 programmes 1994-96 (% of Structural Funds) Maximum Objective 2 allocation to RTDI at measure level

Minimum Objective 2 allocation to RTDI at measure level

Member State

National % RTDI

allocation (Priority

level calculation)

National % RTDI

allocation (Measure

level calculation)

Programme % Programme %

Austria (1995-99)

20 16 Oberösterreich 52 Niederösterreich 0

Belgium 11 16 Liège 20 Aubange 0 Denmark 36 24 Nordjylland 26 Lolland 17 Finland (1995-96)

9 5 Finland 5 n/a 0

France 14 15 Nord Pais de Calais

32 Bourgogne; Languedoc; Lorraine; PACA; Rhône-Alpes

0

Germany 11 13 Saarland 27 Bremen 6 Italy 9 13 Emilia Romagna 25 Lombardia; Val

d’Aosta 0

Luxembourg 14 12 Luxembourg 12 n/a 0 Netherlands 8 18 Zuidoost Brabant 40 Twente 2 Spain 5 15 Baleares 38 Pluriregional 0 Sweden (1995-99)

28 16 Norra Norrlandskusten

33 Fyrstad 3

UK 9.0 14.2 West Cumbria & Furness

17 Gibraltar 0

Sources: Bachtler, J, Taylor, S with Kearney, C (1996) op. cit. Except: Priority level figures for the EU 12, from CEC (1995) Communication of the Commission on the New Regional Programmes under Objectives 1 and 2, COM95 111 final, Brussels, and for the new Member States, from CEC (1996) Communication of the Commission on the Implementation of EU Regional Policies in Austria, Finland and Sweden, COM (96) 316 final.

Analysis of measure-level data from the 1994-96 SPDs (and the first programmes of the new Member States when they were approved) showed that nine out of ten programmes included some form of RTDI measure, but that variations between individual programmes continued to be wide.176 In the Netherlands, for example, RTDI expenditure in Zuidoost Brabant amounted to 40.3 percent of Objective 2 funding, compared to 2 percent in Twente. The highest RTDI expenditure was in Oberösterreich with 52 percent - the only programme to spend more than half of the Objective 2 allocation on RTDI measures. Usually, RTDI was a part of integrated and wide-ranging programmes rather than their focus. Although RTDI received more prominence generally in the 1994-96 programmes, there were still some regions which failed to include explicitly RTDI-focused measures, among them Languedoc Roussillon, Lombardia and Gibraltar. The programmes with no measures dedicated to RTDI were dispersed across the Member States receiving Objective 2 funding rather than being concentrated in any one country or group of countries. Some may have included relevant elements, but were only parts of measures and thus difficult to quantify. In a few cases, the

176 Bachtler J, Taylor S with Kearney C (1996), op. cit.

133

lack of RTDI measures could be explained by programmes taking a specific focus, of which RTDI measures were not an obvious component. An example is Gibraltar, whose 1994-96 programme was heavily oriented towards developing the existing capacity for tourism. In addition, some regions which considered they had limited potential to become technology and innovation leaders (including some programme areas without a major city, and little upon which to build) focused resources in other areas more likely to help regenerate their economies (eg. Bourgogne and Franche-Comté). In some cases, RTDI was not made a high priority because it had not yet become a prominent issue among the economic development community.

6.5.1 Strategies and Policy Orientations A critical innovation in the 1994-96 round of programming was that wider and deeper strategic exercises were carried out by some regions to inform policy design, some of which were undertaken outside the context of Objective 2. These exercises were responses to the increasingly accepted idea that innovation results from complex and intensive interactions between a spectrum of actors on both the supply and demand sides, and that raising innovation levels requires a more holistic, wide-ranging and co-ordinated approach.

Among the highest profile initiatives introduced to encourage strategy building for innovation and technology transfer were the Commission’s Regional Innovation and Technology Transfer Strategies (RITTS) (DG XIII) and Regional Technology Plans (RTP) (DG XVI), the first of which were launched between 1994 and 1996. The RTP was later superseded by the Regional Innovation Strategy (RIS). RITTS initiatives are undertaken in any region wishing to base its RTDI funding on a specific regional strategy, but RIS initiatives can only be undertaken in regions where the majority of the population is eligible under the Structural Funds, and provide a tool to create a framework on which Structural Fund-supported RTDI measures can be based. The aim of both of these initiatives was to motivate the participation of wide regional partnerships in developing regional technology and innovation frameworks to optimise and co-ordinate existing technology strengths to the benefit of local businesses, especially SMEs. Strategic exercises undertaken during the initial phase included the RTP for Industrial South Wales.

Whether or not they undertook detailed strategic exercises – and most did not - programmes pursing RTDI measures typically proposed a range of complementary policy initiatives including:

��infrastructure, such as advanced research and teaching facilities and technology or science parks;

��technology advice and consultancy support services;

��human resource development, including training SME employees in new technologies, tailoring courses to better suit commercial requirements and introducing science and technology graduates into firms;

��enabling the wider exploitation of existing environmental technologies and/or the development of new eco-friendly products and processes;

��network building between firms and a variety of research facilities;

134

��raising awareness about RTD opportunities and encouraging greater participation by SMEs in national and international projects.

While the potential for innovation and technology transfer existed in many Objective 2 areas, R&D was often not sufficiently applied, and the necessary reflexes, networks and interaction processes between relevant actors were underdeveloped. RTDI funding in 1994-96 was encouraged to focus on the improved dissemination of research results and technology and innovation transfer to business. There was a reduction in levels of funding allocated to RTD-related infrastructure with a shifting in emphasis towards ‘soft’ measures.

In addition, complementary human resource measures emphasised the need to accompany other RTDI measures by equipping the workforce to adapt to changing working methods and environments, thereby increasing the ability of regions to accept and apply new innovations. Training measures frequently appeared in this round of programming in the form of advanced training in technology and innovation, often for those already employed in highly skilled jobs in their field of expertise. This is a reflection of the need for lifelong learning even among the highly skilled. This type of action was evident in many of the UK programmes but also in the Netherlands, Belgium, Germany and Finland, in each case taking various forms. Lorraine, Bretagne and Greater Manchester were some of the regions which provided for training in the application of new technologies. Other concepts included adapting training courses to meet better the requirements of firms (in Liguria) and financial support for graduate training places in firms (in Nordjylland, South Wales and Aquitaine).

Notwithstanding the shift towards softer measures, in some regions there was still scope for infrastructural improvement. Often, this concerned the extension, updating and upgrading of existing research facilities to increase their capacity for applied research, for the direct benefit of SMEs. Regions where infrastructure measures featured significantly included many in Spain, but also Oberösterreich, Bayern and Zuid Limburg. In Madrid the focus was on developing and improving technology parks and in La Rioja on improving university and technology centre facilities.

Networking measures were also supported in a number of programmes including the Auvergne, Centre and Toscana as well as in Niedersachsen and Thanet. These actions were typically concerned with increasing technology and innovation transfer in regions by encouraging interactions between technology providers and local businesses. Lolland adapted this strategy to its own context: with limited indigenous RTDI institutions, it promoted linkages between local firms and research institutes outside the region.

As previously mentioned, RTDI type actions were often included under business support measures and frequently included finance for developing new products and processes or for new innovation networks. Measures supporting advice and consultancy actions were included in Ångermanlandskusten, Nordrhein-Westfalen, Centre and Alsace. Environmental actions received increased attention in the 1994-96 round, a frequent theme linked to RTDI being the development or application of

135

improved environmental technologies. Activities of this nature were evident in around one in four programmes during the 1994-1996 programming period - a financial allocation of around 22 MECU per year, typically focused on the wider use of existing clean technologies and the development of environmentally friendly techniques and products. A region with well-known activities in this area was Groningen-Drenthe, where Objective 2 funding was used to develop an existing cluster of firms active in the environment, waste, energy and recycling sectors using new technologies.

Best-practice approaches identified in the 1989-93 programming period were again evident in 1994-96, among them policies building on existing assets. In this programming period, North East England developed three ‘centres of excellence’ capitalising on the existing regional educational and research strengths of three universities (Teeside, Sunderland and Newcastle), each of which specialised in a specific service or area of expertise related to the character of local firms seeking support or advice. Teeside University specialised in services for the processing industries (inspired by ICI and other chemical firms locally); Sunderland University built on Nissan’s presence by providing expertise for high volume industries; and Newcastle University, inspired by its own local company base, specialised in low volume, built-to-order production. Smaller service providers were able to link with these three centres, which provided a framework for firms seeking support or advice. Other good examples of building on strengths were in France, where the national RTD policy framework helped reinforce the strengths of individual regions in specific fields. In Rhône-Alpes, for instance, the development of a medical fabrics testing and certification laboratory by the Institut Textile de France helped strengthen an existing advanced medical textiles cluster in the area and raise the region’s national and international profile. In Finland, South Karelia and Central Finland also built on existing potential with measures to support their ‘national centre of expertise’ status in specific fields. In Päijät-Häme, the strategic work undertaken in an unsuccessful bid to become a centre of expertise was exploited by the Objective 2 programme, with the inclusion of support for the development of environmental technologies.

While many programmes focused on increasing the supply of technological expertise, facilities and advice, there was a parallel requirement to increase business demand for forms of RTDI support. Actions included awareness raising initiatives, such as those in Rheinland-Pfalz, Saarland and Emilia Romagna, which targeted firms able to benefit from technology policies but which were either unaware or ignorant of the benefits of involvement. Other measures aimed to encourage SME participation in national and international RTD projects (eg. through themed workshops in Ångermanlandskusten in Sweden). Another strand of RTDI support was ensuring improved access to wider, excellence-driven RTDI initiatives such as the Fourth Framework Programme: Liège created centres of excellence to help SMEs become active participants in EC R&D programmes.

In the absence of a comprehensive ex post evaluation of the 1994-96 programmes, a thematic evaluation of RTDI in Objective 2 programmes

136

commissioned by DG XVI177 summarises the types of outputs and impacts anticipated by selected Member States as a result of the 1994-96 interventions (Table 6.3). These provide further insights into the strategic direction and intended consequences of RTDI support during this period. The output measures reflect the wide range of direct, material products or activities co-financed by the programmes, including consultancy projects, innovation centres created or expanded, and academic placements in industry. Impact indicators describe the final economic outcomes of projects; these were primarily jobs created, but also - less frequently - measures such as business growth and investment levels. The list reflects a focus on measuring outputs. It also appears to show that impact indicators may be limited in their sophistication, ambitious in their reliance on direct employment creation (which tends to be modest), and fail to encompass indicators which can successfully describe the degree to which ‘innovative milieux’ will have been promoted. This problem is not confined to Objective 2 areas: an OECD study found that understanding of the knowledge economy generally was constrained ‘by the extent and quality of the available knowledge related indicators’.178

Table 6.3: Impacts and outputs for the 1994-96 programmes, classified by type Countries Indicative Types of Output and Impact for the 1994-96 Programmes Austria Impacts: Jobs maintained and created

Investment generated New firm creation and business growth Outputs: Number of R&D projects, including co-operative Amount of consultation and advice Creation or expansion of technology and/or innovation centres Training for R&D

Finland Impacts: Increased competitiveness of companies Diversification of regional economic structure Employment growth Outputs: Participation in skill-enhancing training Number of project proposals especially for infrastructure projects

Germany Impacts: Associated job creation Patents Outputs: Number of seminars on technological co-operation Number of advice and consultancy projects Counselling sessions and projects and jobs saved through these actions Collaborative and networking projects between SMEs and R&D

institutions Technology transfer projects for the environment Media technology and environmental technology projects Consultancy for manufacturing-oriented environmental production, energy

reduction Projects for modernisation of training centres, training places saved Number of participants of vocational training Construction or support to innovation centres, technology centres and

technology transfer institutes, and associated job creation Projects offering service facilities to SMEs and associated job creation Number of projects on support for innovative products Number of supported innovative enterprises

177 ADE, ZENIT, Enterprise (1999) op cit. 178 OECD (1996) The Knowledge Based Economy, OECD, Paris.

137

Spain Impacts: Job creation (including number of researchers hired) Induced investment Patents obtained Outputs: Participation in fairs and commercial missions Number of enterprises receiving advice Number of strategic plans and innovation plans Entrepreneurial co-operation Technical assistance, management development, management courses Re-settlement of firms in technology parks Number of meetings, R&D workshops and scientific meetings and number

of participants Technological parks, technology centres and laboratories created Equipment purchased, R&D units equipped Number of research projects, number of researchers involved, PhD grants Postgraduate students involved in training

Source: Adapted from ADE, ZENIT, Enterprise (1999) op. cit.

6.6 The 1997-99 Programming Period For the 1997-99 programming period, EC guidance notes179 emphasised more strongly than ever the importance of RTDI in economic competitiveness and its potential in helping Objective 2 areas to substitute or revive their declining industries. The notes linked competitive advantage to ‘being at the forefront of innovation in products, production techniques, clean technologies, services and service delivery mechanisms’ and urged those responsible for programming documents in this period to seek synergy from RTDI actions through the use of the regional technology strategies which had begun to prove their worth in the previous programming period.

Overall, Commission figures estimate that expenditure on RTDI measures increased from 10 percent of overall allocations in 1994-96 to 13 percent in 1997-99 (rising from 691 to 1,226 MECU). Measured by the proportion of funding involved, the major investors in this area were Finland, Denmark, Luxembourg and Belgium (26, 22, 20 and 18 percent respectively) (Table 6.4). In absolute terms, the UK and France are spending the most, in part due to the number and size of their programmes. They dedicate an average of 12 and 14 percent of their Objective 2 funding to this field, which translates as 327 and 322 MECU.

Following Commission guidance, the current programming period has been marked by an explicit recognition of RTDI as an essential element supporting diversification away from traditional industries and raising indigenous potential. As in the previous round, programmes were encouraged to build on their existing RTDI infrastructure and areas of expertise to build strategic programmes for technology and innovation. ‘Soft’, applied measures have been the main focus, aiming to increase demand for RTDI services among businesses, and to establish behaviours typical of a successful innovative milieu. Process issues were emphasised in the Commission note, which stated that ‘technical advances (especially in R&D) are not themselves sufficient to ensure successful innovation’. Human resource measures related to technology transfer and innovation have also been also encouraged to equip

179 CEC (1996) Note for Guidance concerning operations in the declining industrial areas (Objective 2) for the second programming period 1997-1999, Commission of the European Communities, Brussels.

138

workforces to facilitate and take advantage of the new opportunities created by raising technology and innovation levels, and to introduce technicians and scientists into firms, often for the first time.

Table 6.4: Contribution of the Structural Funds to Objective 2 RTDI Expenditure for the Period 1997-99

Country* Total MECU % Belgium 38 18 Denmark 15 22 Finland 35 26 Germany 134 15 Spain 258 17 France 322 14 Italy 61 6 Luxembourg 2 20 Netherlands 34 8 UK 327 12 EU Average 1226 13

*Figures were not provided for Austria and Sweden, whose programmes covered 1995-99, but these are contained in Table 6.2. Source: CEC (1997)180

6.6.1 Programme Objectives The extent to which regions have prioritised RTDI as a means to accelerate structural change and to increase business competitiveness is reflected in their regional profile, overall strategy statement and the position these policies have among the programme priorities.

At the level of the regional profile, the degree of emphasis placed on a detailed diagnosis of the innovation and technology situation remains disappointing in many cases, with passing references made to main weaknesses and assets. Among the exceptions are: West Berlin, whose SPD covers RTDI issues in a number of sections; Bayern, where the theme has a separate heading in the socio-economic analysis; Niederösterreich, where a sub-chapter is devoted to industry and technology; and Midi-Pyrénées, which provides a five-page analysis. Some programmes do not include detailed analysis in the SPD, but state that they have drawn on wider frameworks and strategic initiatives in defining their measures (eg. RTPs in Groningen-Drenthe and Industrial South Wales, RITTS in Ångermanlandskusten, Norra Norrlandskusten, Aquitaine and Western Scotland, and the Contrat de Plan and national R&D priority frameworks in Poitou-Charentes). An increasingly wide body of opinion is convinced of the value of wider strategic exercises in helping to target the Objective 2 resources allocated to RTDI and to ensure durable, co-ordinated and sustainable outcomes. In many cases, however, integrated frameworks and sufficient depth of prior analysis are still absent.

180 CEC (1997) The new regional programmes 1997-1999 under Objective 2 of the Community’s Structural policies - focusing on job-creation, Commission of the European Communities, COM(97)524 final, 14.11.97, Brussels.

139

At the level of the overall strategy statement, over half of the current Objective 2 programmes explicitly mention RTDI. It is a prominent theme in the new Member State programmes, appearing in every Austrian and Swedish strategy statement (for 1995-99) and in the national SPD for Finland. In other countries, notably Germany and the UK, it forms an important element in the large majority of Objective 2 programmes. In France, the picture at the level of the strategy statement is mixed. Aquitaine, Bretagne and Poitou-Charentes are among the regions which place strong emphasis on RTDI in their strategy statement, in comparison with Bourgogne, where it receives less importance, in part because of the perceived lack of technological potential in the eligible area.181

The RTDI component of Objective 2 strategy statements remains vague in most programmes, typically stating the broad aim of increasing regional technology capacity, innovation or technology transfer activities, and in many cases supplementing this with the range of RTDI activities which will be supported. On a positive note, the emphasis is predominantly on process issues and/or applied RTD. For example, the West Midlands aims to encourage SMEs to take advantage of technology transfer, while Norra-Norrlandskusten states that they will improve links between regional companies and knowledge centres. While infrastructure is mentioned rarely in strategy statements, it still forms part of the policy portfolio in some regions, including Schleswig-Holstein (expansion of research centres) and Oberösterreich (creation and expansion of technology infrastructure).

6.6.2 Policy Presentation and Definition Where programmes include RTDI policies, they are presented in four main ways. The first is under dedicated priorities, which often combine ERDF and ESF measures eg. Auvergne where Priority 2 (Develop training and research to benefit firms) consists of two measures, proposing the development of applied research centres, and awareness raising and training in sectors relating to research, science and technology. Other programmes include RTDI at measure level, usually under priority headings relating to indigenous business development. For example, in Lolland under Priority 1 (Business development building on indigenous potential), product development and R&D are among the issues addressed. Some programmes include both whole priorities dedicated to RTDI and further measures under related Priorities (eg. Cataluña – see box). A further group of regions addresses RTDI as a ‘horizontal issue’, integrating it as a sub-theme in a wide variety of programme measures (eg. Picardie). In terms of the quality of presentation, however, programmes rarely explicitly identify or explore the synergies between potentially related RTDI policies.

181 The high priority on RTDI in the current Aquitaine programme has been facilitated by the fact that the universities of Bordeaux are within the eligible areas; were they outside the area, then the capacity to absorb resources on even applied projects would be significantly reduced.

140

RTDI PRIORITIES AND MEASURES - CATALUÑA

Priority 1: Support for employment, competitiveness and internationalisation Measure 1.01 Technology centres for industry Measure 1.07 Central government support for technological development of firms Measure 1.10 Technological development projects

Priority 3: Support for research, technology and innovation Measure 3.01 Building and equipping university and research technical centres Measure 3.02 Support for technological research and innovation Measure 3.03 Central government R&D projects Measure 3.04 R&D infrastructure for central government universities Measure 3.05 Infrastructure for the High Council for Scientific Research

For programmes placing emphasis on RTDI as a significant element of their development plan, there are clear benefits from flagging it prominently – for example by dedicating an entire priority to RTDI. There may also be benefits from including RTDI as an integral element of wider business development priorities and measures, to bring this aspect to the attention of relevant firms and to emphasise its position alongside other business development themes.

At this stage, the RTDI policies contained in Objective 2 programmes reflect a variety of influences including:

��the specific characteristics of the eligible area in terms of technology and innovation capacities;

��current understanding about the nature of the innovation process in specific contexts;

��the policy preferences of the EC; and

��existing RTDI policies and organisations operating in the eligible area.

Key trends in the overall policy balance reinforce themes already mentioned under the account of 1989-93 and 1994-96 programmes, notably a decrease in expenditure dedicated to infrastructure, and a concomitant increase in other types ‘soft’ measures focusing primarily on innovation and technology transfer. In Turnhout, for example, the 1997-99 SPD saw a shift towards technology transfer activity and the promotion of clustering, in comparison with a heavy focus on infrastructure in the 1989-93 programme associated with the creation of a Business and Innovation centre. The creation, equipping or improvement of technology infrastructure continues to form part of the policy mix, although typically as a supporting activity at the measure, not priority level, and combined with ‘soft’ measures. In Oberösterreich, an infrastructure measure for the creation and expansion of technology-related infrastructure is accompanied by a further measure addressing product and production process innovation.

An increasingly frequent theme in technology transfer is the diffusion of established technologies to lagging firms – an important dimension in old industrial economies. In addition, RTDI measures in Objective 2 programmes

141

are also becoming more demand, rather than supply focused.182 Two issues are being addressed: the identification of current technology demands among regional firms and institutions; and the stimulation of demand among these organisations. The focus on improving business success is likely to be continued.

THE WEST MIDLANDS - RTDI MEASURES ENCOURAGING SME DEMAND FOR TECHNOLOGY SERVICES

In the West Midlands (UK), expenditure of MECU 85.6 (19 percent of the total Objective 2 allocation) is financing a Priority for Innovation, Technology and R&D development designed to encourage SMEs to become more involved in the development and take-up of new technologies. The priority is being implemented in the context of a RIS through three measures:

• = helping to create and sustain innovative organisations - including training actions for innovation and R&D, actions aimed at improving awareness of the business potential of telematics, and increasing ‘benchmarking’ expertise to facilitate development of best practices;

• = market and technology vision - to encourage SMEs to make long-term investment in innovation, R&D and technology transfer through a process of strategy and vision development; and

• = exploitation and development of regional intellectual capital - to provide the infrastructure, information and support base necessary to engage SMEs in innovation, R&D and technology transfer.

Training projects are often a complementary element of other technology-related ERDF projects. There are specific measures in over half of SPDs proposing high level technology or innovation-related training in support of other RTDI measures. These training measures form almost 30 percent of the total Objective 2 ESF allocation. The ESF has a key supporting role, but actual take-up of such opportunities is sometimes limited by the preconception of the ESF as a ‘bottom rung’ fund, only relevant to pre-vocational and basic vocational training.

The development or diffusion of environmental technologies has become established across many programmes, particularly in the UK, following experimental inclusion in the 1994-96 programmes, exploring their potential to increase competitiveness by improving processes and developing products for niche markets. West Cumbria is supporting both the development and application of new environmental technologies and their commercial exploitation, while Nord Pas de Calais is supporting clean and lean technologies under the priority heading of ‘Support for and relaunch of economic activities’.

Support for the information society is a further theme receiving increased attention, reflecting its growing importance and the potential it presents to create jobs and reduce the geographical disadvantage faced by some (especially peripheral) economies.183 Such policies are already frequent

182 CEC (1997) The Structural Funds in 1996: Eighth Annual Report, OOPEC, Luxembourg. 183 CEC (1996) Green Paper on Living and Working in the Information Society: People First, COM (96) 389, Commission of the European Communities, Brussels. CEC (1998) Job Opportunities in the

142

features of Italian and UK programmes. Bretagne, under a general economic development priority, aims to provide new technologies for information and communications, aspiring to use resource centres to diffuse information technology skills, as well as focusing on the relatively new and growing area of ‘telebusiness’. Telecommunications is part of the Nordrhein-Westfalen programme, with a specific focus on the media and telecommunications sectors. Overall, however, Objective 2 support for information and communications technologies is uncommon - surprisingly so in view of the extensive publicity given to this field as a potential source of increased competitiveness across all sectors and of job creation.

6.7 Objective 2 RTDI Policies: Approaches to Implementation184 The implementation of different RTDI interventions varies. Some are relatively straightforward, notably measures for the development or enhancement of RTD infrastructure, which usually involve a small number of large projects which are often in their formative stages even before the SPD is approved. In turn, the addition of co-financing to existing national or regional RTD schemes (such as Puma and Logic in France and Smart and Spur in the UK) to increase the number or size of awards, is often straightforward: they are well known to potential applicants and economic developers, well tested and have dedicated channels to market them and appraise the resulting applications. Other policies present more difficulties, among them more innovative or speculative policies, those aimed at increasing the technology and innovation levels of less responsive firms and those promoting behavioural changes through intangible activities such as networking. The difficulties involved in this policy area combined with its importance oblige more active and resourceful methods of implementation than most other fields of intervention. The following sections draw out some principles of programme management and implementation which can help to support the success of this dimension of programming.

6.7.1 Wider Strategic Context The first and most important observation relating to RTDI elements of Objective 2 programmes is that additional effort at the policy development stage pays off at the stage of implementation, facilitating delivery by guaranteeing the appropriateness, demand and coherence of specific initiatives and commitment to them by relevant economic development organisations and other key RTDI actors.185 Undertaking wide, region-specific and partnership-based exercises in strategic thinking - such as, but not exclusively, those promoted by RIS and RITTS initiatives - can only be beneficial, and indeed is now widely accepted as an essential precondition to effective targeting.186

Information Society: Exploiting the Potential of the Information Revolution, COM (1998) 590 final, Report to the European Council by the Commission of the European Communities, Brussels. 184 For further detail on the issues discussed in this section, see: Taylor S (1996) RTD/Innovation Policies in Objective 2 Programmes, IQ-Net Thematic Paper, 1(2), EPRC, University of Strathclyde, Glasgow. 185 Downes R and Rooney M (1998) Thinking Strategically – RTD and Objective 2 Programmes, IQ-Net Thematic Paper, 4(3), EPRC, University of Strathclyde, Glasgow. 186 ADE, ZENIT, Enterprise (1999) op. cit.

143

Such exercises not only improve policies but also build the partnerships essential for their delivery.

STRATEGIC THINKING FOR RTDI – BUT OUTWITH THE RIS/RITTS FRAMEWORK

Nordjylland, one of the regions which has focused on technology and innovation in its Structural Fund programmes since 1989, has actively used a process of dialogue and consultation at regional level between politicians, social partners, the regionally-based university and other interested parties to develop its programme. It has also been kept up to date by the contribution to this debate of newly created organisations, often those which were established with the help of European funding.

A number of factors contribute to the on-going and incremental approach which has been taken to strategic thinking about RTDI in Nordjylland. The wider context and environment for innovation and technology development are positive: technology development and innovation are an important national focus in Denmark. European programmes, and STRIDE in particular, have encouraged a regional culture of innovation by supporting networks between businesses and research organisations. As this process encourages higher levels of technological capacity in the region, new initiatives are able to use this basis as their starting point.

6.7.2 Project Generation and Selection

a. Facilitating the timely emergence of projects The short life of current Objective 2 programmes, combined with the relative complexity of many RTDI-related projects (including the frequent need to demonstrate demand for proposed initiatives and to develop project partnerships prior to approval) imposes a degree of urgency on the implementation of relevant measures. Inclusive strategy-building exercises provide a head start in getting this process underway, as does a pro-active approach to marketing programme opportunities - especially to potential new applicants. A further beneficial element is to ensure continuity between programming periods, so that projects begun under one programme can come to fruition under the next. A specific example of where this has been invaluable is a large-scale project in Poitou-Charentes to develop and commercialise know-how in electric vehicle technologies. Complex projects can also be assisted by making funds available for the development of operating concepts, as well as for the subsequent implementation of projects.

Even with vigorous marketing, RTDI measures are among those most likely to fail to attract sufficient numbers of good applications, due in part to underdeveloped dynamics of innovation and technology transfer. It is therefore necessary to monitor applications closely and to take prompt action to understand and resolve any problems. The diverse reasons preventing applications from being submitted range from a simple lack of awareness among target groups to the availability of similar schemes elsewhere with higher award rates or the perception that measures lack relevance, perhaps being too complex or ambitious.

b. Selection criteria guiding project development and selection An important means to guide applicants and shape project selection is the use of selection criteria in measures. A review of programming documents

144

demonstrates that the range and sophistication of the selection criteria used to assess RTDI-related projects has increased over time, as the range of potential projects has increased, policy objectives have been clarified and greater understanding has been gained about which project features are the most important. Table 6.5 summarises some of the main criteria which are set out in the RTDI measures of 1997-99 SPDs.

The range of criteria identified in a survey of the whole population of Objective 2 SPDs is wide, but it should be noted that these criteria do not typify the choices of most individual programmes, which are much less well developed. Few criteria are used in most cases, and they often tend to be very general. In addition, the picture they give is partial, in that many SPDs do not explicitly state the criteria which are actually used in selection (for example, because there was not time to develop these criteria at the programme development stage or because the programme wishes to retain flexibility of selection). An examination of the range of criteria across different programmes does, however, enable a discussion about their rationale and the potential value of exploiting a wider range.

While RTDI is central to competitiveness, RTDI projects tend to bring only modest immediate and medium-term employment impacts. This is reflected in the selection criteria, where job creation frequently appears, but is nuanced by an explicit consideration of the quality of jobs created, and supplemented by a wide range of other selection criteria relating to wider potential impacts on competitiveness, demand for technological innovations and the ‘innovative milieu’ more generally.

The significance of projects to the wider economy is a popular criterion, relating projects to their ability to reinforce existing clusters or regional specialisms. Some programmes include ‘scientific significance’ as a selection criterion for projects, particularly infrastructure investments and the establishment of centres of expertise. This emphasis on absolute (and not relative) excellence is a useful reminder that regions are trying to compete in the wider economy and that some installations are only worthwhile if they are at the leading edge.

A further group of selection criteria comprises those relating to the potential of projects to change behaviour among technology users and suppliers, raising awareness of, and increasing demand for, the services of technology providers, facilitating contact between suppliers and providers, encouraging providers to become more applied in their focus and using large firms to raise the technology levels of their sub-contractors. It is these processes of interaction which must become self-sustaining to enable regional economies to get and then stay ahead.

145

Table 6.5: Selection criteria for RTDI projects in 1997-99 Objective 2 programmes Employment criteria • = Quantity & quality of jobs Significance of project to the regional economy • = ‘Strategic’ projects, promising a material change in the regional economy • = Furthering the development of the Information Society • = RTDI support to firms in specific sectors (eg. software, technology for the information society), often

associated with reinforcement of cluster development • = Projects reinforcing regional sectoral technology specialisms and infrastructure, coherence with wider

policy framework • = Products – new or improved, innovations and new know-how • = Projects of a model, influential or replicable character Scientific importance of the projects supported • = Excellence measures (eg. national recognition of research teams, scientific value of initiatives

proposed) Increased interactions between technology and innovation providers and users and networking behaviours • = Networking (businesses, training organisations, research organisations), use of information providers,

integration with existing networks • = Improved SME knowledge about services available, and increased use of these services • = Participation of SMEs, level of co-operation between SMEs • = Participation of large firms where they support the development of SMEs, including their suppliers • = Use of information technology to facilitate networking, increased use of information networks • = Projects making research institutes more applied • = For technology information services, intermediaries and projects increasing the applied outputs of

research centres: evidence of prospect of becoming self-financing • = Encouraging researchers/scientists/technicians to be employed by SMEs Material changes to regional and local firms • = Technology and innovation levels of SMEs • = Positive impacts on businesses, especially SMEs, including existence of a technological ‘leap’ • = Projects demonstrating relevance to business needs (eg. through prior study) • = Part of an integrated business strategy addressing technical issues alongside commercial issues and

training Other criteria • = Reduction in harm to the environment • = Environmental impacts, eg. potential to recycle waste resources, use of renewable energies and low

energy technologies • = Projects reinforced by training elements • = Training filling gaps in technology provision, including skills and technologies not already available

in existing centres • = Participation in European projects

Businesses are the weak link in many regions, and a further sub-set of selection criteria sets out expectations in terms of business impacts. Among the criteria under this heading is one stipulating the existence of a ‘technological leap’ for firms. This is a flexible measure which recognises that for many more backward firms the adoption of established technologies represents a significant modernisation. It has been used to good effect in Nordjylland where projects are prioritised which offer a qualitative leap forwards/upwards for a firm, relative to the present level of technological development. A further useful selection criterion is to stipulate that demand for technology initiatives be established prior to approval. This requires a more applied outlook and prior interaction with target groups.

146

c. RTDI project appraisal and selection In spite of the selection criteria, there may be a tendency for programmes to accept any reasonable application in order to consume available funds, since this is a field where it is sometimes difficult to generate projects. However, imposing selectivity and focus may deliver greater rewards in the medium to long term, enabling the generation of more coherent and relevant projects and more robust appraisal. The Western Scotland advisory committees were reactive at first in their implementation of the RTD aspects of their SPD, assessing individual applications on their merits without the benefit of a strong strategic framework. Thanks in particular to a RIS, they have since become more critical, only supporting those projects which really fit their strategic direction and aspirations. It is notable that this RIS also helped to increase both the number of applications and their quality.

The content of RTDI project proposals is frequently specialised, and there has been a range of responses to assessing them. At a general level, the fact that applicants are committed to co-financing projects can be seen as some guarantee of their value.187 Other programmes have gone further, concluding that involving experts - often with independent status - to appraise such projects is necessary to make rounded judgements. Relevant specialists need to be able to identify with certainty whether a proposal duplicates existing initiatives or addresses technological challenges which have already been resolved. The use of Advisory Groups to assess applications has been one means to introduce this safeguard (eg. in Scotland and Sweden). In addition, the use of scoring systems or at least structured appraisal grids can help to deepen and systematise the appraisal process.188

d. Implementation and monitoring At the implementation stage, the complexity or pilot nature of many RTDI projects can mean that they benefit from being monitored more closely and in a more rounded way than other types of project. They may meet unexpected problems, and could also present opportunities for additional, related, support initiatives to maximise their effectiveness eg. in training, marketing or network development. Investment in some form of project ‘mentoring’ service could be worthwhile.

6.8 RTDI in Objective 2: successful projects and approaches Objective 2 programme managers and partners have taken a cumulative learning approach to improving policy delivery. There is also much to be gained from such regions exchanging best-practice ideas related to the projects which are supported, especially where these have been successful in addressing some of the most intractable barriers to innovation and helped to animate or better articulate the regional RTDI environment. By way of example, this section highlights a small number of successful projects and approaches which have been undertaken with Objective 2 support.

187 ADE, ZENIT, Enterprise (1999) op. cit. 188 Ibid.

147

6.8.1 Objective 2 Projects addressing Barriers to Innovation In addition to honing the processes by which RTDI priorities and measures are implemented, one of the most difficult issues in Objective 2 areas is successfully to change the behaviour and outlook of firms and technology providers. First, encouraging ‘traditionally’ structured and managed SMEs in mature sectors to be more receptive to innovation presents significant difficulties. Many SMEs dismiss the relevance of technology suppliers, are unaware what services are available or believe that they are not the type of firm able to benefit from such services. At the same time, their useage would involve complex demands: they would benefit more from the support of a range of applied R&D organisations rather than a closed relationship with just one. Second, technology and innovation generators may have the potential to bring applied solutions to business, but they tend not to be appropriately structured or oriented to fulfil this objective. Third, it can be difficult for generalist business development organisations to develop and maintain sufficient levels of expertise to advise on core business functions as well as to diagnose business problems relating to firms’ use (or underuse) of technology.

Among the effective projects coming forward to provide the bridge between firms and relevant RTDI services are the establishment of technology and innovation intermediaries, or ‘brokers’. In Limburg, an agency was set up to act as an intermediary between firms interested in engaging RTD services, and relevant providers. The support identified by the agency was free the first time, then a sliding fee scale was applied for subsequent requests. The ‘one door approach’ offered by the agency, and the attractive terms of the fee scale encourage firms to seek support which they might otherwise not have considered. ‘Knowledge House’ in North-East England has also made complex environment more accessible and comprehensible to the small firm. An individual in each university in the eligible area identifies fields of special expertise and the people responsible for them. This information is brought together into a common directory. By approaching a single office, companies can identify easily and quickly whether the expertise they require is available and from whom. A similar intermediary organisation exists at the University of Saarland. The Kontaktstelle für Wissens- und Technologietransfer (Contact point for knowledge and technology transfer), set up in 1985, promotes business awareness of the research areas of the university departments and facilitates joint projects between scientists and firms. At a more comprehensive level, Aquitaine has used Objective 2 funding to establish potentially extremely effective structures to give greater visibility to regional RTDI providers and users and to support flexible interactions between them.

In turn, a new form of technology transfer policy in Nordrhein-Westfalen has been the establishment of interest groups to encourage networking in specified areas. On the basis that concrete institutions are an expensive and insufficiently flexible method of transferring information about new technological developments, the idea is to group actors in similar fields, including higher education representatives, firms, consultants and chambers of commerce, to discuss recent developments through workshops, seminars, bulletins, etc. The groups, such as Initiative Teletech, may spawn joint projects which are then part-financed by the firms involved.

148

CO-ORDINATING CHANGE IN AQUITAINE: USING CLUSTER CO-ORDINATORS TO BETTER MATCH TECHNOLOGY

SUPPLY AND DEMAND

The region of Aquitaine is pursuing a cluster strategy, designed with the support of a RITTS exercise and implemented in significant part through the Objective 2 programme. This focuses on both sectoral foci (health sector, agro-industries/nutrition and forestry, wood and papers) and multi-sectoral foci: (materials and electronics/IT/modernisation). Each cluster or ‘pôle’ has an open network structure, and is co-ordinated by a carefully selected animator, whose task is to structure the cluster’s development and ensure its efficient operation.

The animator plays a pivotal role. It does not offer specific services, but rather provides an interface between firms and technology providers/schemes. Animators explore demand in their sector through direct contacts with firms and business service organisations, and, in parallel, analyse available research facilities, projects and services which meet the identified needs. This enables needs to be creatively matched to supply. The animators also stimulate information loops, for example informing business service organisations about projects which are underway. This function ensures that more actors are better informed and possible networks can be more easily identified.

The insights collected by animators mean that they are in an excellent position to influence research providers and make them more responsive to firms. The autonomy and core functions of research institutions are never questioned, but the animator can, for example, highlight the potential applied relevance of pure research projects. Equally, animators help to change the attitudes of firms through thematic seminars, presentations of potentially relevant new technologies, etc. Where specific projects are identified, animators help with the project development, building project teams and identifying funding sources.

Innovative projects have also come forward to raise the ability of general business development agencies to at least start the dialogue with firms about technology. Such services are particularly valuable given that ‘SMEs plan in overall terms’,189 and benefit more from considering options for their development in an integrated rather than a compartmentalised way. In Rhône-Alpes, highly qualified unemployed people, often from the shrinking defence sector, are trained in business counselling and in the current business development policy environment, and placed in local chambers of commerce (CCIs). From this base, they undertake visits to firms and identify business development opportunities - often of a technological nature. In so doing, they not only use the specific skills they have built up during their working lives for the benefit of other firms, but also increase their own employment prospects. The initiative has provided the CCI in Drôme with the personnel to cater for all firms in the locality. When projects are identified, another skilled unemployed person is placed in the firm in a developmental and implementation role. If the project succeeds, then the person responsible is taken on by the firm; if it fails, the person has at least increased experience and improved employment prospects.

Company behaviours can also be changed through human resource development projects. In Nordjylland, graduates are placed into SMEs to help with specific technology-related projects. These graduates help to build a link between the firm and the research community and so increase the likelihood

189 Hassink R (1996) Technology Transfer Infrastructures; some Experiences in Europe, the USA and Japan compared, Paper presented at the conference “A Changing Europe in a Changing World: Urban and Regional Issues”, Crossmead Conference Centre, University of Exeter, UK, 11-14 April 1996.

149

that the firm will be inclined and equipped to seek further technology support if needed. In Friuli Venezia Giulia, training is being provided to graduates in the transfer of know-how so that they can act as an interface between research centres and SMEs. A further similar example is found in Saarland, where the Institute for Technology Transfer at the School of Technology and Commerce acts as an intermediary co-operating with firms, particularly SMEs, in specialised projects in areas such as architecture, electronics and the environment. In some cases, senior students will be placed in firms for up to a year to work on specific projects, providing a bridge between research and business.

In further examples, skilled employees from within firms are sent on high-level training courses which enable them to enhance their skill levels in using or applying technologies. This also forges direct and personal links between the training establishment and the firm. A further benefit of such placement and training initiatives is that they ensure that firms have the appropriate internal capacities to be able to transform the information available on innovation and technology initiatives into ‘useful knowledge’ which the firm can use. Without being selected, decoded and absorbed, such information may be of little value to the receiving firm.190

6.9 Conclusions and Recommendations This chapter has explored how the RTDI content of Objective 2 programmes has evolved over the last decade, in line with growing awareness about the importance of technology transfer and innovation to regions seeking to regain competitive advantage and deepening understanding of the barriers to establishing these dynamic, interactive processes. Across successive programming periods, rising expenditure has been dedicated to RTDI policies, across an increasingly diverse and sophisticated range of interventions. The policy orientation has changed, with the science and technology focus of early programmes being replaced by a stronger emphasis on applied policies, targeted at raising the ability of regional firms to compete.

Initiatives increasing the demand for technological solutions to business problems have been complemented by policies enabling existing regional scientific and research organisations to respond. In addition, the balance of interventions being supported has changed, with a shift away from the focus on infrastructure which characterised the early programmes (science and technology parks, equipping of research-oriented laboratories), towards softer and more demand-side measures, including support services to assist businesses to identify and implement projects to raise their technology and innovation levels. ERDF policies have also routinely been complemented by human resource development measures – unusual among ESF interventions in that they tend to target those who are already highly qualified in relevant fields - helping employees to raise or adapt their technical skills, or introducing technical or research staff into firms. An element of policy which addresses a particular deficit of Objective 2 areas and is receiving increasing attention is

190 Valentin F (1995) Technology Transfer from Networks and Public Research to Innovative SMEs: the Hidden Costs of Stretching Learning Domains, In: O’Docherty D P (ed.), Globalisation, Networking and Small Firm Formation, Gordon & Trotman, London, Dordrecht, Boston, p 67-86.

150

the diffusion of established technologies to lagging firms. In addition, policies explicitly promoting the adoption and exploitation of new communication and information technologies are seeing strong growth.

While RTDI has become a more prominent element of Objective 2 programmes, the overall positive picture masks considerable variations between programmes in terms of the design, internal coherence, justification and potential for synergies of these policies. Most programmes could increase the depth of diagnostic analysis and strategy development phases, and make further efforts to encourage mutually reinforcing projects with genuine potential to raise regional innovation capacity, focusing efforts on specific clusters or fields of unique regional expertise.

6.9.1 Raising RTDI Levels: Programme Development Recommendations

RTDI is a demanding policy field and the depth of diagnostic effort needs to be improved across more Objective 2 programmes to design and deliver more effective policies.

RTDI-related regional policy is distinctive: it has to accumulate the many components necessary to enable innovation and technological advancement, based around regional specificity and unique knowledge; it also has to ‘breathe life into them’, establishing a self-sustaining dynamic of continual learning, change and exchange among all the actors involved. For policies to achieve such an impact, they require a deeper level of strategic reflection, focus and co-ordination, and greater active participation than most policy types. Detailed regional and sector specific analyses are needed to identify weaknesses in the innovative milieu and design policies able to address them successfully and establish new and durable behaviours and systems.

The Objective 2 programme development timetable is not conducive to the necessary diagnostic and policy-making practices. Therefore, parallel, partnership-based RTDI strategy-building initiatives are highly recommended: the Commission’s RIS and RITTS studies have proved highly effective in many cases.

Objective 2 programmes facilitate the partner interactions which can drive effective policy development, but the compressed timetables and broad scope of Objective 2 programmes can inhibit exercises of sufficient depth, focus and continuity – except where the context is particularly favourable.

RTP, RIS and RITTS projects have facilitated dedicated strategy development and partnership building in innovation and technology transfer in many Objective 2 areas. These projects have shown considerable potential in helping regions to address the complex of issues surrounding low technology and innovation levels. In many of the relevant cases, they have significantly reinforced regional capacities and potential.

151

The momentum gained by some of these strategic exercises, whose outcomes have been operationalised with the help of Objective 2 co-financing, is already such that they are likely to become durable elements of the economic development environment. Given the significance of RTDI to industrial regeneration, every Objective 2 area could benefit from such an exercise.

Alternatives are available to the RIS/RITTS route, among them establishing sub-groups of the Objective 2 partnership to work in an incremental way towards improved RTDI policies (as in Nordjylland). Others have undertaken timely thematic evaluations to help to refine their Structural Fund-specific approaches (Aquitaine). Such studies could be undertaken by programmes which are leading into the next programming phase without a strategic framework, in particular to understand better three issues:

��What are the unmet technology and innovation needs of different types of business in the eligible area (demand-side issue)?

��How could/should these be met (supply-side issue)?

��How could increased business demand for technology and innovation services be stimulated?

Artificially confining studies to eligible areas, where these are small or fragmented, is counterproductive. Instead, any strategic exercise must be ‘regional’, selecting its geographic scope on the basis of clear economic rationales not Structural Fund eligibility maps. It should address an area which is sufficiently large and coherent to provide the arena for a future sustainable dynamic of innovation, and assess within this what is required in the Objective 2 area.

In addition, strategy development should be managed as an inclusive partnership process in order to establish a shared vision, securing consensus around common goals and facilitating successful implementation.

Elements of RTDI policy are relevant to every Objective 2 area, and every programme should explore how they could be used. Policies should be tailored to address the specific characteristics of lagging firms as well as those with greatest innovative potential, and should include parallel human resource development measures for best effect.

All Objective 2 areas could increase their competitiveness by raising their technology and innovation levels. It follows that RTDI-related policies are relevant to all of them, although tailored in each case to local conditions. It is strongly recommended that all Objective 2 partnerships explore how their Objective 2 programme could use RTDI policies to serve their competitiveness objectives.

RTDI policies should exploit opportunity, in the form of local and regional specificities, including existing clusters and fields of special expertise. At the same time, they should address the entrenched deficiencies of Objective 2 areas. Two specific fields of intervention relevant to every old industrial programming area which merit further attention and clearer articulation are:

152

the further dissemination of established technologies, to enable the modernisation of firms relying on outdated technologies; and the greater exploitation of information and communication technologies. Stimulating electronic commerce, new private and public services and innovative applications,191 is not only a source of significant job creation potential,192 but also helps to counter the disadvantages associated with remoteness and isolation from which some Objective 2 areas suffer.

To reinforce their potential RTDI actions supported by the ERDF frequently require parallel or related human resource development actions. In the new programming environment, ensuring that this dimension is addressed, must be a priority. Objective 2 programmes should focus at least some of their dedicated ESF interventions (those which are additional to Objective 3) on supporting actions in the field of RTDI.

This is likely to be the last Objective 2 programme of this scale in most regions, and a priority should be placed on actions able to deliver durable, self-sustaining effects.

Raising the technology and innovation levels of individual firms is an important dimension of the task of Objective 2 programmes. However, such actions only impact modestly on the overall regional capacity to innovate on a long-term basis. Therefore, it remains important to prioritise actions which contribute more widely to new and durable systems and behaviours. Giving more shape and visibility to the environment for innovation, by prioritising the development of intermediary organisations and brokerage services which bring life to networks by providing pro-active links between firms and technology and innovation providers, has been a successful strand of programmes. Experience in Aquitaine demonstrates its value (see section 6.8.1). The establishment of new structures or animating organisations must become self-sustaining, and any projects proposed by such structures must be clearly justified in terms of verifiable demand among firms, ideally involving businesses in their design.

Raising a region’s technology and innovation capacity requires a wide range of issues to be addressed through diverse policy instruments. Relevant policies should be organised and presented prominently in Structural Fund programmes, with a clear expression of policy priorities.

Regional RTDI strategies and policies are frequently not sufficiently visible. Sometimes, RTDI-related elements are a minor element of many measures, making it difficult to identify the overall objectives being pursued or to relate elements to each other. Even where policies are more visible, because they are

191 CEC (1999) The Structural Funds and their Co-ordination with the Coherence Fund – Draft guidance for programmes in the period 2000-06, Working paper of the Commission of the European Communities, Brussels. 192 CEC (1998) op. cit.

153

grouped together, limited emphasis may be placed on how they relate to each other in terms of the basic dynamic of supply and demand. There is no single ‘best’ solution, but several features are recommended.

��Clear information should be provided in the regional profile, including a detailed description of the identified needs of firms, and existing technology and innovation services and infrastructure. The profile should highlight the match between supply and demand, and derive proposals which focus on stimulating and responding to business needs.

��The overriding policy emphasis should be clearly expressed as identifying unmet business demand for innovation and technology, stimulating further demand and meeting both – ie. a rationale driven by demand as defined by commercial priorities, rather than one driven by supply-oriented scientific and technological rationales.

��Introducing an RTDI priority, for example for ‘Innovation and Competitiveness’ is one way to give relevant policies more prominence, although this also risks compartmentalising RTDI policies, a sub-set of policies relevant to specialised economic development agencies and the most advanced and dynamic firms.193 Alternatives may be preferable, for example setting out a clear RTDI strategy, and then specifying RTDI-related policies in priorities, addressing the business community, the business environment and then the wider regional environment.

��The rationale for integrating RTDI policies is particularly strong for SME development. RTDI is frequently a key dimension. SMEs have been shown to benefit more from customised, integrated packages of support rather than narrow stand-alone projects. Also, they may not automatically perceive support for technology and innovation to be among their needs, and it makes sense to retain RTDI as an integral element of suites of SME development policies. Different policy responses are appropriate for different types of firm, and this should also be clearly articulated within each priority.

6.9.2 Raising RTDI Levels: Project Generation and Selection Recommendations

Pro-active and inclusive approaches are required to generate sufficient good quality project proposals in the time available under Objective 2.

While the programming period has been lengthened for 2000-06, there will still be a need to ensure that RTDI opportunities are vigorously promoted to relevant audiences to bring forward sufficient good quality projects. Making programme funding available for the development of operating concepts as well as for project implementation can help to drive forward innovative and promising but complex project ideas.

193 ADE, ZENIT, Enterprise (1999) op. cit.

154

RTDI project proposals are some of the most technical and difficult applications to appraise. Provision should routinely be made for specialist involvement in this task.

To provide a reliable indication of their quality, relevance and potential and to identify existing or potential related initiatives, RTDI projects require expert appraisal. In some cases, programmes have established Advisory Groups to feed into the decision-making process (eg. Scotland and Sweden). Where this is not the case, a register of voluntary expert assessors could be compiled. The involvement of expert assessors has implications for the programme timetable. In the same way that relevant projects undergo environmental and, in gender impact assessments, it could be possible to define a form of ‘Technology and Innovation Impact Assessment’ procedure to gauge the potential of individual projects.

The selection criteria used by most programmes could be improved to become more focused and specific. This could help to emphasise further the priority placed on regionally relevant projects for which there is proven demand and an applied economic rationale.

Selection criteria for RTDI projects are growing in sophistication and relevance, but could still improve in Objective 2 programmes. Reference to the range of criteria already in use and the rationales behind them would provide a useful starting point (see Table 6.5).

Among the overriding criteria should be a demonstration of need or demand for the projects proposed. The use of this criterion provides a strong incentive encouraging deeper reflection at the project design stage.

Few programmes can establish genuine competition between proposals as there is not a situation of overbid. To replace this, programmes could apply clear selection criteria more strictly, rejecting projects not meeting quality levels or contributing sufficiently to core programme objectives, even if they are theoretically eligible. This would fulfil the same role as competition in increasing proposal quality.

6.9.3 Raising RTDI Levels: Monitoring and Evaluation Recommendations

The ability to capture the true effects of RTDI actions, particularly with regard to their impact on regional innovation capacity, is under-developed and requires further work.

The primary value of most RTDI-related projects lies in their ability to reinforce competitiveness – and only in the medium to long term to generate or safeguard employment. At the same time, the overriding rationale for

155

Objective 2 programmes remains employment creation. To ensure good quality assessment and monitoring of the impact of RTDI projects as a distinctive high priority sub-group, monitoring indicators and baselines have to be developed and systematically applied which measure value and success in this specialised field more meaningfully. Considerable theoretical work has already been undertaken in this field, and this could be more systematically exploited. In particular, to ensure more rapid and uniform responses, the Commission could provide guidance on minimum or core RTDI-relevant indicators.

Programmes may benefit from being closely and supportively monitored, given the complex, innovative and unpredictable nature of RTDI projects, the opportunities for spin-off projects, and the intangible or long-term nature of their outputs/impacts.

Project monitoring is usually associated with ‘policing’ projects and ensuring their sound financial management and timely progress. In RTDI projects, this function can be extremely valuable in pro-actively following up projects to ensure that they are going to plan and to gather intelligence on other issues. This would include specific difficulties being experienced with given types of innovative project and ideas for spin-off opportunities or supporting elements. This sort of follow-up also helps to strengthen the personal ties between relevant professionals – important in instilling the new linkages and behaviours required for a self-sustaining innovative milieu. Such project mentoring presents considerable resource implications.

RTDI merits special provisions in programme-wide evaluations.

Programme-wide evaluations span a wide range of economic development interventions, and tend to manage them by focusing on employment impacts as the central unifying indicator. In so doing, they frequently fail to capture the main contributions of RTDI policies and projects. Studies should modify their methodology to take into account the importance and distinctive nature of this sub-group of policies and projects. They could also involve expert evaluators where this would provide a greater insight.

The ability of evaluations to capture technology and innovation outcomes also depends on appropriate and effective monitoring systems being established at an early stage in the programming cycle, and followed through systematically.

Exchange of experience between programmes and with related initiatives outside Objective 2 has considerable potential for the dissemination of good practices.

National and transnational exchange of experience on the design, delivery and evaluation of RTDI policies is already intense between RIS and RITTS

156

participants, and has taken place to some extent between Structural Fund programme partners. Such exchanges should be further prioritised, and compartmentalisation reduced between organisations pursuing related objectives but via discrete instruments.

157

7. ENVIRONMENTAL INTEGRATION IN OBJECTIVE 2 PROGRAMMES

7.1 Introduction In 1987, the World Commission on Environment and Development (WCED)194 called for a global effort to integrate economic development and environmental protection. It urged that major central economic and sectoral agencies of governments should be made directly responsible and fully accountable for ensuring that their policies, programmes and budgets supported ecologically and economically sustainable development.

Acknowledging that environmental problems frequently originate from a diverse range of activities including transport, agriculture, tourism and manufacturing, the WCED went beyond the conventional view of environmental policy. It stressed that it is not simply about environment agencies implementing their own policies, but that other sectoral specialists must recognise the environmental dimension within their work. Emphasis was placed upon interdisciplinarity, and decision-makers were urged not to be isolationist, but rather to think broadly and to inform and influence others constructively.

The idea of environmental integration was taken further at the United Nations Conference on Environment and Development (UNCED) held in Rio de Janeiro in 1992. From a sustainable development perspective, the programme of Agenda 21 addressed the integration of environment and development in decision-making, particularly at the strategic level of policy, planning and management.195 Adopting a long-term perspective and cross-sectoral approach, the programme called upon countries to ensure a three-way (economic, social and environmental) integration into decision-making at all levels in all areas of government. Such a vision assumed transparency of - and accountability for - the environmental implications of economic and sectoral policies, advocating systems such as integrated economic and environmental accounting (IEEA) to encompass additional (non-financial) areas in national accounts.

Integration has also been given substantial attention by the Organisation for Economic Co-operation and Development (OECD), which recognises the integration of economic and environmental decision-making as a key challenge.196 This is based on the fundamental link between economic growth and the environment, which implies that economic and environmental policies cannot be made and implemented in isolation. Instead, instruments such as cost-benefit analysis and scientific risk assessments are identified as support mechanisms for setting environmental priorities; and achieving compatibility between environmental and sectoral economic policies is highlighted as a central objective, while being subject to continued monitoring and evaluation.

194 World Commission on Environment and Development (1987) Our Common Future, Oxford University Press, Oxford. 195 United Nations Conference on Environment and Development (1992) Agenda 21: The United Nations Programme of Action from Rio, UN, New York. 196 Organisation for Economic Co-operation and Development (1996), Integrating Environment and Economy: Progress in the 1990s, OECD, Paris.

158

Within the European Union, this momentum is reflected in a number of impacts, including the EU Environmental Action Programmes, the inclusion of environmental objectives within Structural Funds framework regulations, guidance on environmental integration for regional development programmes, and increasing emphasis on identifying environmental progress through monitoring and evaluation. In the following pages, this chapter examines the background and recent trends relevant to the Structural Funds, assesses existing environmental practice in the current round of Objective 2 programmes, and presents a series of practical recommendations on how best to integrate environment into future programmes.

7.2 The EU and Sustainable Development Six years ago, the European Commission adopted Towards Sustainability, the Fifth Environmental Action Programme for the period 1993-2000.197 Representing a fundamental shift in outlook from earlier programmes, Towards Sustainability takes a holistic view of issues, reflecting the wider aims of sustainable development as well as integrating environmental concerns into the social and economic framework. For practical understanding, the concept of sustainable development may be divided into three components:198

��continuing economic development - creating greater wealth in a manner that reduces burdens on the natural environment;

��social inclusion - meeting people’s material needs as well as their wider aspirations which add to quality of life; and

��environmental protection - recognising that environmental quality is linked to economic development - for example, as a factor in inward investment decisions - and social development, as a component of quality of life.

Accordingly, sustainable development is process-oriented, rather than static, referring to the process of developing in a sustainable manner and of progressing towards a specific goal within that process.199 To support moves towards this goal, four key principles are embodied in the Fifth Action Programme:

��the prevention principle, preventing damage before it occurs, in preference to remediation after the event;

��the precautionary principle, which states that authorities should act to avoid the possibility of environmental damage where scientific evidence is inconclusive, but the damage would be significant and difficult to reverse;

��the polluter pays principle, which states that full environmental costs of pollution should be borne by those who cause damage; and

197 CEC (1993) Towards Sustainability - A European Community Programme of Policy and Action in Relation to the Environment and Sustainable Development, Official Journal No C138, 17 May 1993, Commission of the European Communities, Brussels. 198 Khan A (1995) Sustainable Development: The Key Concepts, Issues and Implications, Sustainable Development Vol.3 No.1. 199 Reid D (1995) Sustainable Development: An Introductory Guide, Earthscan, London.

159

��the proximity principle which states that, as far as possible, regions and countries should be self sufficient, generating their own raw materials and products, and disposing of wastes.

The programme also distinguishes between environmental issues that must be addressed and economic sectors with the potential to impact either positively or negatively on the environment. In summary, the main environmental challenges facing the Community were identified as follows:

��reduction of emissions causing acidification and air pollution, especially heavy metals and sulphur dioxide;

��cutting back greenhouse gas emissions considered to contribute to climate changes and ozone depleting substances;

��preserving biodiversity and stopping the depletion of natural resources;

��management of water resources in view of quantity and quality problems in various regions;

��improving the urban environment;

��combating the deterioration of coastal zones;

��management of environmental risks; and

��reduction of waste and improvement of waste management.

The economic sectors considered worthy of greatest attention because of their disproportional contribution to current environmental problems are manufacturing industry, energy, transport, agriculture and tourism (see box).

For manufacturing industry, the main goals to be achieved include improved resource management, the provision of information to promote a more environmentally-friendly consumer choice, and stricter standards for products and production methods.

For the energy sector, in order to reduce the consumption of non-renewable resources as well as the emissions of gases, improved energy efficiency and the use of less carbon-intensive energy sources are the main objectives.

For transport, a balance is sought between public and private transport, alongside improvements to the European railway network and intermodal co-operation, and the strategy also considers stricter measures such as road pricing or increasing fuel prices.

For the agriculture sector, key elements in promoting coexistence of agriculture and the environment include stricter controls over the use of pesticides, herbicides and fertilisers as well as financial incentives to encourage more environmentally friendly farming practices.

For tourism, the diversification of tourism activities and promotion of environmentally-friendly (‘sustainable’) tourism are seen as remedies for the severe negative environmental impacts associated with those regions that are destinations of mass tourism, especially in sensitive areas.

Mirroring the focus of the Fifth Programme, a further tool for realising the aspirations of sustainable development is the methodological framework of ecological modernisation.200 Through policy integration, ecological

200 Weale A (1993) Ecological Modernisation and the Integration of European Environmental Policy, in J Liefferink, P D Lowe and A Mol (eds.) European Integration and Environmental Policy, John

160

modernisation seeks to provide an alternative to the antagonistic relationship between economic development and environmental protection that has prevailed in developed economies. In essence, greater effectiveness is expected if policy-makers can successfully integrate environmental policy goals into related areas such as economic development, energy and transport, so that environmental consequences are considered in decision-making, and adjustments are made during programme formulation and implementation in recognition of environmental implications.

Over time, these systemic changes should ultimately produce more efficiency and mutually-supportive policies. With environmental issues, this becomes even more pressing, due to the complexity of problems and interlinkages between disciplinary approaches. In practice, ecological modernisation requires strong integration and realignment of objectives to become self-reinforcing, not conflicting.201

7.3 EU Regional Policy and the Environment Over the last twelve years, several factors have contributed to raise the importance of the issue of environment within the context of EU regional policy. Particularly since the mid-1980s, the EU Environmental Action Programmes have introduced the concept of environmental integration.202 Whilst acknowledging sustainable development as an essential component of economic growth, the Third Environmental Action Programme (1982-86) called for a strategy to integrate environmental policy with socio-economic development; the Fourth Environmental Action Programme (1987-92) further developed the theme of integration while advancing the idea of environmental responsibility; and the Fifth Programme considerably broadened the approach, requiring the integration of environmental concerns into all other areas of activity, including the economic development process as supported by EC financial support mechanisms.203

Environmental protection now has a basis in EU regional policy through the Structural Funds Framework Regulations. Whereas a requirement for environmental appraisal of the Funds was first incorporated in 1988, this was upgraded and given more weight in the 1994-1999 Regulations, so that, in preparing Regional Development Plans or Single Programming Documents, Member States are now obliged to meet four obligations:

��to prepare an assessment of the environmental situation of the region concerned;

Wiley & Sons, Chichester. Gouldson A and Murphy J (1996) Ecological Modernisation and the European Union, Geoforum, Vol 27 Part 1. 201 Baldock D, Beaufoy G, Haigh N, Hewett J, Wilkinson D and Wenning M (1992) The Integration of Environmental Protection Requirements into the Definition and Implementation of other EC Policies Institute for European Environmental Policy, London. 202 Johnson S and Corcelle G (1995) The Environmental Policy of the European Communities, Second Edition, Kluwer Law International, The Hague. 203 CEC (1995) Progress Report on Towards Sustainability, COM(95) 624 final, Commission of the European Communities, Brussels.

161

��to evaluate the environmental impact of the strategy in accordance with the principles of sustainable development and in agreement with the provisions of Community law in force;

��to make arrangements to associate the competent environmental authorities designated by the Member State in the preparation and implementation of the operations foreseen in the plan; and

��to ensure compliance with Community policy and legislation concerning the environment.

The development of these regulatory requirements was prompted by a series of reports in previous years that had consistently revealed inadequate consideration of the environment in the Structural Funds. Following critical independent reviews of Structural Funds and the environment,204 the European Court of Auditors published a report in 1992 that found little evidence to support any claims of environmental conformity within the Structural Funds.205

Further insight was provided by the 1994 Interim Review of the Fifth Environmental Action Programme. It noted that while there had been progress on the integration of environmental approaches both within the Community and individual Member States, sustainable development was still seen as the business of those who deal directly with the environment. It observed that:

"strategic assessments of the impact of policy initiatives on the environment have yet to take root in most of the Member States…and…initiatives on the introduction and implementation of appropriate mechanisms to ensure that environment and sustainable development concerns are considered in new policy actions need to be speeded up".206

In addition:

"Integration of environmental concerns needs to be better reflected in the regional development plans and proposals for funding being developed by Member States".

More recently, in 1996, the production of Commission guidance and positive responses by European partnerships progressed the recognition of the environment as a key issue in subsequent revisions of Structural Funds/SPDs. However, this has been more evident in descriptive text and environmental chapters, rather than full integration into the detail of priorities and measures. Furthermore, although projects funded have been entirely in line with sustainable development - examples exist of physical environmental improvements, tourism developments, waste minimisation as part of business

204 Baldock D and Corrie, H (1989) The EC Structural Funds: Environmental Briefing, Institute for European Environmental Policy, London. Baldock D and Wenning M (1990) The EC Structural Funds: Environmental Briefing 2, Institute for European Environmental Policy, London 205 CEC (1992) Court of Auditors Special Report No.3/92 concerning the Environment together with the Commission's replies, Official Journal No C 245, 23 September 1992, commission of the European Communities, Brussels. 206 CEC (1994) Interim Review of Implementation of The European Community Programme of Policy and Action in relation to the environment and sustainable development "Towards Sustainability", COM (94) 453 Final, Commission of the European Communities, Brussels.

162

development and a range of training projects - assessment criteria to ensure that these projects are recognised and rewarded for their contribution are often lacking.

Within the implementation of the Funds, there has been a trend away from physical infrastructure projects towards business development and training. As a result, it is arguably more difficult to assess the environmental impact of the Funds (certainly at ex ante stage) because of the greater potential range of projects which could be supported. Built infrastructure projects, given sufficient information, will have relatively predictable impacts in terms of energy use, building materials employed, transport access, and a range of physical factors associated with construction, such as land use and effects on groundwater.

By contrast, business development measures could have positive impacts, for example through new product development in environmental technology, waste minimisation or energy efficiency in manufacturing businesses; or they might have negative impacts through projects which indirectly encourage businesses to produce more waste through expansion or which increase freight traffic through export promotion.

Clearly, advances in Structural Fund implementation mean that a stage has been reached which brings new opportunities and new problems. To find an appropriate means of operation in this changing context, further reference to the principles of sustainable development can highlight useful ways of viewing and assessing the environmental sustainability of both programmes and projects. First, however, the following sections review Objective 2 programmes in a number of EU Member States, to identify strengths and progress within current practice.

7.4 Evolution of Programming and the Environment

7.4.1 First Programming Period, 1989-93 The first rounds of programmes were agreed by the Commission and Member States in the late-1980s. In these early documents, the general philosophy was to give priority to European cohesion, based on reducing economic disparities likely to undermine the forthcoming Single European Market (SEM). Environmental policy was to be adhered to, but it did not condition the operation of the Funds or the SEM initiative.

Accordingly, few of the programmes integrated the environment in the sense of using its protection or improvement as a development objective. Where this did happen, it generally occurred within infrastructure programmes, and the level of integration achieved varied substantially.207 In essence, these programmes were perceived as economic and social - not environmental - strategies, and consequently job-creation and economic development took precedence over environmental protection. Moreover, the Commission had no legal grounds for requiring programme-level environmental assessment.

207 Woodford J (1991) Conflict or Convergence? Environmental Priorities and the Structural Funds Environmental Policy Discussion Paper No.1, EPRC, University of Strathclyde, Glasgow.

163

Environmental references in these early programmes may be grouped into two main classes: those which imply a measure of integration of environmental factors in the programming and execution process; and those which consider environmental policy as an external element of the process. In other words, these interpretations are either that the development philosophy must include environmental matters or alternatively that it must conform to environmental policy and legislation.

As might be expected, the clearest statement of integration came from those northern Member States where environmental protection issues were already a part of day-to-day development processes. For example, Denmark's Objective 5b programme stated:

“It is a fact that future economic and infrastructure development must be undertaken with regard to the environment and nature protection. Forcing production in the primary and industrial sectors as well as the intensive construction of infrastructure, especially in tourism, can eventually damage landscapes and the environment. All measures linked to production and to infrastructure thus have an environmental dimension.”

The other end of the spectrum can be illustrated with an example from the Portuguese Objective 1 submission:

“It should be noted that actions specifically aimed, in the Portuguese situation, at the preservation of the natural heritage and the exploitation of the full potential of internally-generated development will be undertaken in conformity with existing Community legislation, for which systematic relations have been established with the Ministry for the Environment and Natural Resources.”

The UK position was more ambivalent. The North East England Objective 2 programme contained an environmental paragraph that described the planning system as:

"providing the framework for the provision of land for the implementation of all measures described in the plan, including environmental improvement”.

It subsequently emphasised conformity with the Directive 85/337/EEC concerning environmental impact assessment. The potential uncertainty here is whether the UK planning system is assumed to integrate environmental considerations, or whether transposing EC Directives into UK law can be used to determine conformity with Community legislation.

In its first annual review of the activity of the reformed Funds in 1989, the Commission, when describing how the operations might improve and protect the environment, stated that even though there was increasing environmental awareness in the regions, there were still serious problems:

“On the one hand, there is a severe backlog of problems to be remedied, while on the other there is a risk that development measures financed by the Funds will aggravate the pressure on the environment

164

(creating precisely the kind of problem that other funds are seeking to remedy”).208

7.4.2 The Second Programming Period, 1994-96 The programming documents prepared for the 1994-99 period (respectively 1994-96 for Objective 2 areas) generally show some improvements. Reviews of SPDs prepared for the 'North Sea' Objective 2 regions209 examined the state-of-the-environment, reporting, environmental impact, environmental gain and environmental integration.

a. State of the Environment The requirement to provide an appraisal of existing environmental conditions was generally complied with, albeit superficially. The information provided tended to be descriptive rather than analytical, there was an absence of quantitative indicators, and the depth and quality of information varied considerably between countries and regions. A common approach was to present environmental information separately from the main body of the SPD, often in an appendix, symptomatic of the way in which environmental factors were often treated in isolation within regional development thinking.

In approaching the analysis of a baseline environmental situation systematically, six categories of indicators can be identified (see Table 7.1):

��air quality relates to information on matters such as the presence of carbon monoxide, nitrogen dioxide, sulphur dioxide and carbon dioxide;

��water quality relates to rivers, estaurine and coastal waters, and drinking water;

��land quality comprises the inclusion of data on vacant and derelict land and buildings, sites of archeological heritage, or urban centres;

��biodiversity relates to the identification of agricultural land, greenbelt designation, nature conservation and wildlife sites;

��incidence of pollution relates to the locations or degrees of pollution encountered, including polluting industries and contaminated land; and

��waste treatment and recycling relates to the focus on current methods utilised, divisions between methods, and anticipated demand.

The most evident characteristic of Table 7.1 is the failure by the Danish authorities to provide the baseline environmental information. It was assumed that physical planning authorities would process any applications approved for Structural Fund support, and data would be gathered at that stage. Accordingly, the Danish SPDs made reference to information provided through statutory procedures of planning and development control, implying that environmental considerations are inappropriate for SPDs, essentially to avoid duplication.

208 CEC (1990) Annual Review of the Structural Funds, OOPEC, Luxembourg. 209 Clement K and Bachtler J (1997) Regional Development and Environmental Gain: Strategic Assessment in the EU Structural Funds, European Environment, Vol 7 No.1 pp.7-14.

165

Table 7.1: SPD information on the state of the environment, North Sea regions 1994-96210

Air Quality

Water Quality

Land Quality

Bio-diversity

Incidence of

Polllution

Waste Treatment

& Recycling

Western Scotland Yes Yes Yes Yes Partial Yes East of Scotland Yes Yes Yes Yes Partial Yes North East England Yes Yes Yes Yes Partial Yes Groningen-Drenthe Yes Yes Yes Yes Yes Yes Lolland No Yes No No No No Nordjylland No No No No No No Schleswig Holstein Yes Yes Yes Yes Yes Yes Bremen Yes Yes Yes Yes Yes Yes Niedersachsen Partial Partial Partial Yes Partial Partial

Two points arise here. First, with regard to programme partnerships, local authorities (as the competent environmental authorities) should have been included in the programme preparation team. In this capacity, they would be expected to make significant inputs in terms of environmental data and impact assessment. Rather than duplication, this would represent appropriate consultation with environmental authorities, as specified in the Framework Regulations. The second issue is that relying on the local authority regulatory framework to process applications invariably results in minimum compliance with regulations, rather than a positive exploitation of the scope for environmental improvement.

b. Environmental Impact Within environmental impact, an issue worth highlighting is the timing - and consequently the feasibility - of impact assessment. Some of the SPDs stated that ex ante impact assessments were not feasible, with project completion cited as a more appropriate time, essentially to carry out an ex post impact assessment. While this reaction is understandable, it could be challenged on the basis that potential positive and negative impacts can be identified ex ante, even if only as guidelines within which to pursue some form of environmental gain. At this stage, it would be most realistic to consider the impacts of projects such as tourism infrastructure, transport facilities or manufacturing investments; other factors such as environmental training and knowledge dissemination would be more difficult to anticipate, and they may be more relevant to the ex post phase.

The impact of proposed measures on the baseline environmental situation is clearly a central part of any environmental appraisal. For this level of assessment, environmental impact can initially be analysed under two headings: broad impact assessment and specific impact assessment. Broad impact assessment relates to whether the plans contain broad statements on

210 Bachtler J (ed.) (1994) Ex-ante Appraisal of the Single Programming Documents for Objective 2 North Sea Regions, Report to the European Commission (DG XVI), EPR, University of Strathclyde, Glasgow.

166

anticipated likely impact; and specific impact assessment relates to whether the plans contain estimates of the likely impact of specific measures or priorities.

The most interesting approach to environmental impact was contained within the North East England Objective 2 programme, which utilised both broad and specific environmental impact assessments. Borrowing from a methodology for spatial planning proposed by the UK Department of the Environment,211 the North East England Plan rated each of the programme priorities for its likely degree of impact against a series of environmental themes, grouped under headings of global sustainability, natural resources and local environmental quality. This allowed an assessment of broad impact - in this case summarised as “a largely neutral effect on the environment” - and preliminary insights into the specific impacts expected from individual programme priorities against environmental criteria.

Although undeveloped beyond this listing of impact levels (including the absence of explanations on how weightings were derived), the methodology did form a useful first step in developing a more sophisticated environmental impact matrix for regional economic policy programmes, exemplified by the Department of the Environment’s earlier guidelines on policy appraisal.212

c. Environmental Gain Fundamental to using the Strategic Environmental Assessment (SEA) as a policy tool is the identification of ways in which environmental objectives can be advanced as part of the development process. The term ‘environmental gain’ is particularly apt in this regard: it may be secured through the incorporation of environmental targets, the inclusion of environmental criteria or indicators, and the establishment of monitoring procedures.213

��Environmental targets relate to whether the SPDs set targets for environmental improvement, for example for derelict land reclamation, pollution reduction, improved recycling or environmental training.

��Environmental criteria in project appraisal refers to the methodologies used for assessing and selecting projects for financial support. This may involve a scoring system incorporating environmental criteria in addition to factors such as job creation, ethnic minorities and new company formation.

��Environmental indicators in performance measures would be a further step, used to provide evidence of factors such as environmental training, product improvements, process improvements, the uptake of environmental technologies, improvements in energy efficiency, or the number of environment-oriented projects supported.

��Environmental monitoring refers to identifying procedures to monitor the environmental impact of the plans over time.

211 Department of the Environment (1993) Environmental Appraisal of Development Plans, HMSO, London. 212 Department of the Environment (1991) Policy Appraisal and the Environment, HMSO, London. 213 Clement K (1999) Environmental Gain within Sustainable Economic Development: Strategic Considerations within for Mediterranean Regions, in G Benhayoun (ed.) Economie des Régions Méditeranéennes et Développement Durable, L'Harmattan, Marseilles.

167

All the North Sea SPDs were assessed against these criteria, and the results indicated that none of the SPDs achieved consistency across all categories (see Table 7.2). The best performers were the German SPDs, which incorporated environmental targets, identified environmental criteria for use in project appraisal and made provision for monitoring procedures. However, the advantageous use of environmental indicators in performance measures was restricted to individual UK and Dutch SPDs that had set no targets and made no allowances for monitoring. Clearly, there was still scope for improvement in the relative consistency across categories if environmental gain were to be realised through these programmes.

Table 7.2: SPD Measures for Environmental Gain, 1994-96 Environmental

Targets Environmental

Criteria in Project

Appraisal

Environmental Indicators in Performance

Measures

Monitoring Procedures

West Scotland No Yes Yes No East Scotland No No No No North East England No No No No Groningen-Dernthe No Yes No Yes Lolland No No Yes No Nordjylland No No No Yes Schleswig Holstein Yes Yes No Yes Bremen Yes Yes No Yes Niedersachsen No Yes No No

d. Environmental Integration Beyond regulatory requirements, it is worthwhile considering the degree to which the SPDs achieved environmental integration - for example whether the environment was treated as a separate issue or as an integral part of the SPD. From this perspective, none of the SPDs integrated environment effectively. The inclusion of environmental references throughout the strategy documents did not necessarily signify successful integration. In general, environmental input related principally to individual measures, rather than approaching the issue holistically. The environment was frequently perceived as a purely physical issue, involving the redevelopment of industrial areas for instance, rather than focusing on issues such as environmental training, product improvements, the uptake of environmental technology, improved energy control measures, or the setting of environmental targets as conditions within projects.

Only a few SPDs approached the question of integration more broadly with the inclusion of themes such as the promotion of environmental technology and environmental investment. A common characteristic was the tendency to see environmental protection as one aspect of economic development, an approach which is not only restrictive, but which hinders the incorporation of environmental criteria into measures for the success of all the action priorities.

Commission sources such as the Interim Review of the Fifth Environmental Action Programme subsequently concluded that there was still scope for better

168

integration of environmental concerns in the regional development plans and programmes developed by Member States; and similar results were reached from comparative assessments of Objective 1 regional development plans and Objective 2 Single Programming Documents.214 These analyses also revealed marked differences between the Member States’ progress in areas such as environmental profiling, the inclusion of environmental aspects within strategies and environmental impact assessment.

7.5 The current Programming Period, 1997-99

7.5.1 The Changing Context In response to the low level of effective integration and impact assessment in the second programming period, the Commission issued notes for guidance for Objective 2 regions in the third phase of programming (1997-1999), and these notes specifically listed ‘Environment and Sustainable Development’ amongst the new priorities. Acknowledging that the complementary nature of the environment and regional development is increasingly being recognised, the guidance stated that the horizontal character of the environment needed to be borne in mind in the definition and implementation of other Community policies and especially in the Structural Funds programmes.

Two main themes were advocated for more vigorous incorporation in new Objective 2 programmes. The first relates to the traditional approach of improving the physical environment to increase the attractiveness of the region for business development; the second is more forward-looking and seeks future competitive advantage by exploiting eco-products, environmental services, environmental technologies, energy-saving measures and improved production processes.

On the first theme, any proposal for the development and improvement of industrial sites should be linked to the stimulation of local development activity, for example meeting the needs of inward investors, local SMEs or stimulating growth in new sectoral specialisms. Other aspects are linked to the direct needs of firms, including the development of environmental infrastructure such as waste recycling facilities, removing pollution and protecting or enhancing areas of ecological interest. Co-financing these measures must also be shown to enhance economic development.

With regard to the second theme, additional emphasis should now be given to forward-looking measures as a potential source of future competitive advantage, linking ecological awareness with economic growth opportunities. Examples might include environmental measures for industry, energy-saving projects, advice for industry, awareness raising on environmental legislation, green business development and marketing support. The broad aims are to improve the environmental performance of business generally and to develop specialist environment-related sectors.

214 Bachtler J (1995) Regional Development Planning in Objective 1 Regions, European Urban and Regional Studies, Vol.2 No.4. Bachtler J and Taylor S (1996) Regional Development Strategies in Objective 2 Regions: A Comparative Assessment, Regional Studies, Vol.31 No.8. Clement K and Fitzgerald R (1997) Regional Environmental Integration: Changing Perceptions and Practice in Objective 2 Programmes, IQ-Net Thematic Paper, 2(3), EPRC, University of Strathclyde, Glasgow.

169

Two external studies also contributed to the treatment of environment in the Structural Funds. The first, by ECOTEC, offered a working definition of sustainable development as pursuing three objectives in such a way as to make them mutually compatible for both current and future generations. These comprised:

��sustainable, non-inflationary economic growth;

��social cohesion through access for all to employment and a high quality of life; and

��enhancement and maintenance of the environmental capital on which life depends.215

The guidance aimed to assist programme managers to identify and promote those features of Objective 2 programmes that could significantly change regional development towards sustainable patterns. The advice identified areas for positive action to shift the pattern of regional development towards sustainability through three stages:

��'business as usual', where appropriate environmental standards and regulations are generally met;

��minimisation, where firms can go beyond existing pollution control standards by adopting best available (clean) technologies and production patterns that conserve energy and recycle waste materials (as these firms become more economically competitive and resource-efficient, this corresponds to 'win-win' scenarios that secure employment as well as fulfilling environmental objectives); and

��laying the basis for sustainable development, which includes encouraging the regional economy to restructure towards sectors that use fewer environmental resources, orienting spatial planning or spatial policies to reduce the need to travel, and increasing opportunities for firms to share heat or exchange by-products or for waste exploitation.

Thereafter, the guidance suggests means of 'tracking the region', essentially monitoring and measuring moves towards sustainability through a series of indicators with both 'top-down' and 'bottom up' characteristics. The resultant statistics of positive and negative outputs should be measured against the baseline conditions to identify progress. The report also identifies possible core project selection criteria according to the three dimensions of economic development, cohesion and environment, with additional environmental criteria that could be used to assist project scoring. As a useful appendix, a parallel report from ECOTEC contains summaries of 20 projects exemplifying good practice in sustainable development and employment.

The second major study was carried out by Environmental Resources Management (ERM).216 Aimed at development authorities and environmental authorities within Member States, the report presents an overview of the scope for integration between the Structural Fund programming process and the

215 ECOTEC (1997) Encouraging Sustainable Development through Objective 2 Programmes: Guidance for Programme Managers, ECOTEC Research and Consulting Ltd, Birmingham. 216 Environmental Resources Management (1998) Environmental Appraisal of Regional Development Plans and EU Structural Funds Programmes: A Handbook for Programme Managers, ERM, London.

170

environmental assessment process. In practical terms, it is intended to act as a handbook setting out ways in which environmental issues can be more systematically incorporated into the definition and preparation of regional development plans and programming documents. It does not represent a legal requirement.

While recognising that environmental factors arise at all stages of programming from plan formulation to ex post evaluation, the ERM handbook focuses principally on the ex ante phase in the Structural Funds process. This relates to the preparation of regional plans and the strategic environmental assessment (SEA) of the plan or programme as well as its associated development strategy, priorities and measures. For each stage of the process from the regional development plan through to the community support framework, single programming document or operational programme, the handbook describes the relevance of SEA.

The report follows a common structure from basic definitions and the importance of baseline data, through sustainability criteria for programme objectives and project selection, to performance indicators, impact indicators and monitoring arrangements. In summary, the document provides a practical and well-structured methodological aid for environmental integration in preparing plans for the current and future rounds of Structural Funds programming.

7.5.2 Regulatory Features

a. Environmental profiles In accordance with the Framework Regulations, the majority of current Objective 2 SPDs contain environmental profiles. These profiles vary in length from four to eighteen pages, and they also vary in usefulness - some remain essentially descriptive and focused on environmental protection, while others are more analytical and aim for environmental improvement and tangible gains.

The programmes also vary considerably in terms of environmental content. The descriptive programmes list the environmental characteristics of the region, but do not analyse the data in much depth, whereas the analytical programmes tend to identify and explore environmental strengths and weaknesses in detail. A problem common to both types of programme is that the environmental profile is seldom linked effectively into the programme in a strategic manner ie. the environmental data and discussion generally remain confined to the profile section rather than influencing other programme components.

Within the spectrum of content, the programmes range from the environmental detail of Germany and the UK through to the environmental brevity of Spain and Sweden. All German Objective 2 programmes include environmental profiles, and they consider impacts on a wide range of ecological factors, West Berlin being a particularly good example; the UK programmes similarly always present environmental profiles, in most cases with good environmental detail and in some instances incorporating sophisticated moves towards programme impact assessment and sustainability criteria. In this latter category, Eastern Scotland appears to have made the greatest progress.

171

b. Identification of impacts With regard to environmental impact, there is a trend towards the identification of impacts for each measure. This is a considerable step forward from the previous round of programmes, and it represents a very useful tool at the ex ante stage, as it allows some consideration of cumulative effects in accordance with strategic environmental assessment. However, in individual analyses, there is a tendency to identify only positive effects and, in some cases, to focus primarily on environmental measures. In practice, all measures should be evaluated for impact, especially since non-environmental measures are more likely to have a negative impact.

SAARLAND – SYSTEMATIC MEASURE LEVEL ASSESSMENT OF POTENTIAL ENVIRONMENTAL IMPACTS

The Saarland programme is among those taking a systematic approach to the prior identification of the possible positive, negative and/or neutral environmental implications of each programme measure. These are flagged on every measure sheet, and inform project development and assessment. It is notable from this treatment of measures that few are in fact environment-neutral. The extent to which potential positive outcomes are achieved or negative outcomes mitigated in practice depends on the detail of the projects implemented, and the extent to which the visible preoccupation with the environment is followed through. Some examples of the environmental implications of measures are given below.

Measure 1.1 Science Park: The science park will have a negative environmental impact through construction on a greenfield site. However, it will have wider positive effects through technical developments for the industrial sector enabling more careful use of resources and reductions in pollution.

Measure 1.4 Business related infrastructure: Positive implications are the reuse of industrial brownfield sites, decontamination overcoming ground, water and air pollution, and renovation of the town improving the population’s standard of living. The creation of new industrial and commercial areas nevertheless always has some negative environmental impacts.

Measure 2.1 Renewable energy, careful use of resources: Direct and immediate benefits to the environment will result from the ecological audit programme and the construction of low energy housing.

Measure 2.3 Support for entrepreneur/business activities and investments: Largely environmentally neutral, but it is hoped that the creation of new jobs in the service sector, substituting industrial jobs, will lead to a lasting fall in industrial pollution.

Training measures: Aspects of vocational training to be introduced into the design of each qualification measure according to occupation are: responsible implementation of scarce, finite raw materials, development and production of environmentally friendly products, use of eco-friendly production methods, and reduction of polluting emissions.

4.1 Preparing and accompanying measures, evaluation: Impact indicators for the programme will continue to be improved, including those for the environment, and the environmental dimension of measures will continue to be increased.

Programmes from Germany and France regularly include impact analysis by measure in the SPDs, whereas countries such as Sweden simply state that environmental impact will be considered at the project selection stage. In an intermediate position, the UK has a number of programmes with environmental assessments of programmes or priorities, and several -

172

including Eastern Scotland - with environmental assessment of individual measures.

7.5.3 Strategic Approach How have programmes approached the environment from a strategic perspective, in particular in the aims and objectives? In most cases, the current programmes approach environment from a horizontal perspective (see Table 7.3). Under this system, environmental issues appear within sub-strategies or within priorities and strategic objectives, but they are not given separate status. French and German programmes have generally adopted this approach, both countries acknowledging the direct links between regional environmental image, industrial investment and the potential for job creation in a number of sectors. Notable exceptions include the programmes for Picardie and West Berlin, each of which has devoted a vertical priority to the theme of environment. This results in a very high profile for environmental factors, and Berlin especially encourages a wide range of project-types in support of environmental integration.

In comparison, within the horizontal approach, the scope varies from the Spanish focus on basic physical infrastructure requirements and clarifying regulatory systems to the United Kingdom's interpretation of environment as encompassing clean technologies, waste exchange, energy efficiency, green tourism and developing the employment potential of eco-businesses. Denmark adopts a far-sighted methodology, using horizontal integration as a starting point, but simultaneously seeking to reward investments that go beyond minimum legislative compliance to exceed government environmental standards. Across the Member States, examples from France, Germany, Denmark and the United Kingdom fit well with the Commission's perspective on exploiting positive links between environment and economy.

By contrast, the Swedish approach has adopted a lower profile, relating more to the need to protect the environment from harmful economic development rather than emphasising the potential synergies between the two areas.

A particularly advanced example of using sustainable development as a basis for the programme's strategic approach is contained in the current pilot project being realised in the Eastern Scotland programme (see box).

173

Table 7.3: The environment in programme aims and objectives, examples from the 1997-99 programmes

Member State Explicit references to the environment in programme aims and objectives Austria (1995-99)

Oberösterreich and Niederösterreich place limited emphasis on the environment in their strategy. Vorarlberg aims to improve the environmental performance and awareness of industry In Steiermark, the overall objective includes the improvement of environmental and living conditions

Denmark Nordjylland takes a horizontal approach, considering environmental issues in every priority. More specifically, the programme will support environmentally-friendly development and production. Lolland highlights a narrower objective, giving priority to investment in environmental protection which exceeds government minimum standards.

Finland The national programme states the broad aim of eliminating environmental problems and reducing environmental harm by projects.

France A range of approaches to the environment is taken in France. Some programmes take a narrow and/or traditional view of environmental interventions: Alsace: industrial waste treatment to form a specialisation in physical regeneration Bourgogne: the environment is seen as an inter-related area to be addressed for exogenous and indigenous development Bretagne: overall aim highlights the need to refine and continue existing environment-related actions, with sub-strategies including environmental improvement. Increasingly, programmes are proposing harnessing the environment as a source of growth: Centre: environmental issues are to be addressed to improve image and attractiveness of the area; and environment to be harnessed as a source of growth. Picardie: a whole priority is given over to environment in forward-looking and experimental approach; the intention is to exploit potential for new environmental sectors. Poitou-Charentes: the environment is seen as a source of economic development potential for jobs, products and improved regional image.

Germany Hessen: the modernisation and diversification of the regional economic structure is to be environmentally-friendly. Nordrhein-Westfalen: the economic strategy will take account of sustainable development: priority to be given to economic measures that reduce pollution and use resources carefully. Rheinland-Pfalz: environment forms part of the regional development aim, and all supported activities to take environmental aims into consideration, especially climate protection, nature protection, etc. Schleswig Holstein: priority to be given to ecological structural improvement, eg. through attractive commercial areas for businesses with low environmental impacts or which contribute to environmental improvement through products or services such as environmental technology or recycling. West Berlin: a whole priority is devoted to environment, aiming to implement integrated environmental protection for SMEs, by supporting the business sector with recycling, environmentally-friendly production methods, product innovation, technology transfer, emissions reductions, re-use of waste, personnel training, and co-operation with east European partners in environmental technology.

Spain Environmental protection is one of six priority axes of the CSF (Priority 2). The main emphasis is protection of the physical environment. Under the ERDF, all

174

Spanish programmes include infrastructure projects (mainly waste and sewerage infrastructure, networks for control and surveillance of environmental parameters, and the clean-up and re-use of regenerated land). Under the ESF they include relevant vocational training and employment aids. Other relevant actions are included in CSF Priority 1 (Business Development) and Priority 5 (Urban Development). Most projects are regional and local authority intervention; the national level participates to a very limited extent in environmental projects. Programme examples: Aragón: Aid schemes for SMEs; burying power lines in highly industrialised areas, clean-up of derelict sites, environmental training, grants for researchers, restoration of San Pablo in Zaragoza city. Cataluña: Aid schemes for SMEs; waste and water treatment facilities, burying power lines in industrial areas, restoration of derelict sites, clean-up of soil, infrastructure extending control and surveillance capabilities, training, restoration of decaying urban areas and of buildings of cultural interest for socio-economic purposes, flood protection in industrial zones. La Rioja: aid schemes for SMEs; several infrastructural installations (water purification station, two waste treatment sites), employment aid and training in environmental protection and management. País Vasco: Reinstating and redeveloping former industrial sites, extension of the existing environmental control, monitoring and surveillance network, training.

Sweden Environmental objectives usually relate to the need to exploit the environment in a sustainable manner, rather than addressing the inter-relationship between the environment and business: Ångermanlandskusten: the region's environmental resources are to be safeguarded; environment is seen as a horizontal issue across all measures. Bergslagen: environmental aim of the programme is to protect and improve the natural and cultural environment and ensure the sustainability of the development process. Norra Norrlandskusten: standards of living of the regional population are to be improved, while capitalising on the region's valuable natural and cultural environment.

United Kingdom

Frequently, wide-ranging environmental commitments, varying depending on context: East Midlands: emphasis placed on the introduction of clean technologies, while demonstrating anti-pollution measures, waste exchange and improvement of derelict areas. Eastern Scotland: aims include the clean-up and re-use of contaminated land, improvement of local environment, developing environmental awareness, and improving environmental performance of SMEs. Industrial South Wales: programme will enhance the natural environment, support land reclamation, encourage energy efficiency and clean technologies, provide environmental information for SMEs, promote environmental management, and support green tourism. North East England: aims to improve the environmental performance of businesses generally and to develop the employment potential of eco-businesses, especially SMEs, with environmental training. Western Scotland: programme to encourage positive land use and re-use of vacant and derelict land, with implications for transport; focus on waste, energy management and pollution minimisation, training and promotion of environmental management for industry. East London and the Lee Valley: makes commitments relating to the way in which the programme itself will manage its environmental dimension, promising an environmental audit in the early phase of the programme to establish monitoring systems and baseline data.

175

SUSTAINABLE DEVELOPMENT IN THE EASTERN SCOTLAND PROGRAMME: A STRATEGIC APPROACH217

The first step in the approach was to assess all elements of the programme against the sustainable development guidance for programme managers provided by ECOTEC. This indicated that the Eastern Scotland programme was performing broadly in line with other Objective 2 programmes, and that elements of sustainable development already existed in the current SPD.

The second stage included a series of participatory workshops involving the whole Partnership. These sessions evaluated the potential of sustainable development to enhance the programme, and the result was a reduction of ECOTEC's original 16 areas of action to four criteria. Linking these with economic and social criteria enabled the derivation of 12 core criteria for sustainable development.

Stage three comprised an interim report to the Commission to adjust output indicators and to integrate the emerging criteria for project selection into 12 sustainable development criteria, as follows:

• = net additional jobs created evidence of demand leverage

• = infrastructure impact resource efficiency environmental impact

• = access and opportunity local added value capacity building

• = social inclusion strategic integration durability and feasibility

7.5.4 Management and Implementation This section considers the organisations involved in managing and implementing environmental aspects of programmes, whether they work towards overt environmental targets and whether environmental criteria feature in project selection.

a. Lead organisations Frequently, there is no single lead organisation for environmental issues relating to Structural Fund programmes. Instead, the increasing range of possible environmental interventions is reflected in the active involvement of a growing diversity of relevant organisations. The implementation of environmental aspects of SPDs may involve a combination of national or federal ministries, working alongside their deconcentrated offices in the regions, other dedicated agencies, and/or regional authorities.

The framework within which regional programmes are developed and implemented is shaped by national organisations with specific environmental remits, such as the Federal Ministry for Environmental Protection in Germany, the Ministry for Environment in Finland, the Ministère de l’Environnement and ADEME (Agence de l’Environnnement et de la Maîtrise de l’Energie) in France and the Environment Agencies in the UK. They provide guidance on possible measures and eligible actions, and advise on standard methodologies for environmental impact assessment, etc. In federal states such as Germany, relevant regional authorities also have a lead role, supplemented by others as appropriate. For example, in Saarland, the Ministry for Environment, Energy and Traffic takes the lead, but other organisations are involved, in accordance

217 Eastern Scotland European Partnership (1998) The Sustainable Development Project: Consultative Report, ESEP Programme Executive, Dunfermline.

176

with their legal competencies, including the Land Office for Environmental Protection and the Town Institute for Health & Environment.

At regional level, individual programme partnerships tend to involve a range of organisations relevant to the interventions they aim to support. Some of these are dedicated specialist organisations eg. the regional offices of ADEME and the DIREN (Direction Régionale de l’Environnement) in France, and relevant Land Ministries in Germany (eg. the Ministry for Environment in Nordrhein-Westfalen). However, the involvement of regional environmental organisations may only extend to verifying the compliance of projects with legal requirements (such as the county Departments of the Environment in Denmark). In others, it is much more extensive, involving active input to policy design, the definition of selection criteria, the generation and assessment of projects and monitoring of their progress, including participation on Monitoring and decision-making Committees. The range of organisations involved in part relates to the local institutional environment. For example, in the UK, Scottish National Heritage is involved in Scottish Objective 2 programmes, while North East England’s programme involves English Nature and the Countryside Commission.

b. Targets The inclusion of specific environmental targets in SPDs is uncommon. To some extent, this is a consequence of attempting to integrate environment on a horizontal basis, as programmes tend to set targets for vertical priorities, to which greater importance is normally attached. Whereas the environmental profile will generally include a useful discussion of qualitative environmental impacts, usually prepared by the lead environmental authorities identified above, numerical targets rarely form a part of such descriptions.

Another approximation of environmental targets is the reference to likely environmental outcomes in the sections covering the impact of the programme. In instances where this relates to individual measures, it is feasible to use these estimates as guidelines for action or aspirations to work towards where positive impacts have been identified. Good examples of how estimating impacts of a programme might substitute for environmental target-setting occur in the Western Scotland programme, in which the programme priorities are assessed against five environmental issues (see box).

Further development of this approach could link the identification of types of projects meriting support with the establishment of targets differentiated by individual measure.

177

IMPACT ASSESSMENT IN SUPPORT OF TARGET-SETTING: WESTERN SCOTLAND218

Within the Western Scotland SPD, each of the five programme priorities has been assessed for its likely (positive or negative) impact against five environmental issues: CO2/global sustainability, land use, air quality/energy, waste and training.

The scores range from high impact, principally for new-build projects with scope for energy efficiency and brownfield locations, intermediate impacts through environmental training projects, and low impacts during the life of the programme, for example related to energy and waste research and development.

Whereas this approach does not go as far as precisely measuring impacts, it lists positive outcomes for which to aim e.g. auditing should result in waste reduction and increased recycling, new-build projects should conform to high standards of energy efficiency, locating projects on brownfield sites should reduce transport needs, and traffic calming alongside investment in public transport investment should improve air quality at local level.

c. Environmental selection criteria Most Member States identify environmental criteria for use in project selection, this occurs only in a few programmes in each country. In general, environmental criteria have not been included as standard, and they do not feature strongly where regionalised scoring systems are used to evaluate a project's strength. Instead, a typical approach is to specify that projects should have no negative effects on the environment.

Noteworthy exceptions to the trend are found in the Nordic programmes from Denmark and Finland. The Lolland programme explicitly states the standard condition that investments in environmental protection must go beyond governmental demands or minimum compliance; in Finland, the Objective 2 programme's overall environmental strategy is expected to filter down to project level where environmental criteria are to feature in project selection.

7.5.5 Monitoring and Evaluation With the exception of France, very few countries have actually evaluated the environmental impacts of previous programmes. Evaluations tend to take a predominantly economic perspective with a preference for measurable and quantified results. Previous rounds of programmes have not been framed in a manner suitable for easy evaluation of environmental factors, and this has impacted on subsequent assessments.

In the case of France, critical evaluations showed that programmes for the 1994-1996 period did not all achieve their predicted environmental impacts. Some evaluations were focused more on activity than impact; in the programme for Poitou-Charentes, environmental progress was identified with the creation of a nature reserve, a study for an urban charter, and water treatment projects. There was also a clear identification of changing priorities (see box).

218 Strathclyde European Partnership (1996) Single Programming Document 1997-1999, SEP Programme Executive, Glasgow.

178

POITOU-CHARENTES: BROADENING ENVIRONMENTAL OBJECTIVES

The environmental dimension of the 1994-96 Poitou-Charentes Objective 2 SPD was focused on the rehabilitation of industrial sites and on a range of specific projects enabling environmental progress, notably a nature reserve, a study for the urban charters and water treatment actions. This field, which only gained momentum towards the end of the 1994-96 programming period, was continued into the 1997-99 programme in several measures:

• = Measures containing urban and industrial rehabilitation actions: Measure 1.1 Support for local reconversion, insertion and social development projects, Measure 1.2 Rehabilitation of abandoned and degraded industrial areas, including addressing pollution and undertaking studies, Measure 2.4 Support for the development of tourism activities, including improvements to sea fronts and adapting historic sites, and Measure 2.5 Development of major infrastructures including port improvements.

At the same time, the 1997-99 programme built on the phase of rehabilitation by exploiting the potential of the environment as a source of economic development advantage (creating new jobs, facilitating new product and service development and improving the region’s image). Relevant actions invited comprise:

• = Improvements to the area’s image under Tourism Measure 2.4 (mentioned above).

• = Environmental projects for business competitiveness under Measure 2.6 (Improvement of the environment, treatment and re-use of waste) which will include business support projects for energy management, pollution prevention, awareness, environmental audits, etc.

• = New eco-products under Measure 3.8 (Development of economic infrastructure and joint services), which will include an integrated project to develop an electric car.

• = The dissemination of environmental expertise through Measure 3.9 (Support for employment creation and human resources), which will include training in environmental issues.

7.6 Conclusions and Recommendations

7.6.1 Orders of Integration In response to the EU's increasing orientation towards environment and sustainable development, Structural Fund programmes are progressively incorporating new features that support the aspirations of the Fifth Environmental Action Programme, Towards Sustainability. In parallel, the environmental impetus provided both by the revised Structural Fund Framework Regulations and the practical guidance issued through studies directed by the Commission has resulted in a much clearer focus on means of realising environmental integration.

In reviewing successive programming periods, three different orders of integration219 can be identified (see Table 7.4). First order integration relates to the stage in which programmes treat environment as a horizontal issue, in practice often meaning that environment receives no explicit priority, but rather is assumed to feature as a component of other (economic) priorities. One characteristic of dealing with environment in this manner is that external

219 Clement K (1998) Generating Regional Environmental Competitiveness: Future Scenarios for Baltic Co-operation and Cohesion, in W Toczyski (ed.) Competition and Co-operation in the Baltic Sea Regions of Denmark, Germany and Poland, Centre for Strategic Studies, Sopot, Poland.

179

consultations carried out with relevant departments or agencies may mean that time constraints severely limit the level and usefulness of external input. In recent years, the majority of SPDs have fallen into this category. Some appear to be making considerable progress in overcoming the hurdles to environmental integration.

Table 7.4: Orders of environmental integration in European programmes FIRST ORDER INTEGRATION

SECOND ORDER INTEGRATION

THIRD ORDER INTEGRATION

Horizontal integration

External consultations

Limited influence

Vertical integration

Committee of environmental experts assessing projects

Environment budget

Strategic integration

Internal environmental expertise

Generating regional environmental competitiveness

Second order integration relates to the higher profile treatment of environment as a vertical component of the regional strategy. Only a small number of reviewed programmes have these characteristics. Vertical integration means that the environmental objectives are expressed as a separate priority of the programme, given more status and as a result monitored more effectively. At this level, a committee of environmental experts might work as a programme advisory group making recommendations on project selection. Such approaches are more likely at this stage than the previous one, when the horizontal approach would typically result in an individual ‘environmentalist’ sitting as a member of other project advisory groups. There is also scope for a clear link to a specific environmental budget heading, which can assist implementation and project accountability. However, a disadvantage of this approach is that partners tend to assume that environmental issues relate only to projects within the environment priority: consequently, horizontal integration may be lost.

The scenario of third order integration would comprise the adoption of a strategic approach to the exploitation of environmental issues with competitive purpose. Bringing together both horizontal and vertical integration ensures that environment would be a very significant element of the economic development strategy. This signifies a substantial change in the perception and role of environment, potentially overseen by internal environmental expertise during the three phases of programme design, implementation and evaluation. Employing both qualitative and quantitative targets for environmental improvement, this programme orientation would set environmental excellence as a central objective, subsequently giving rise to environmental competitiveness. In the current round of programmes, this third level of integration is principally aspirational. Although several programmes appear to be moving towards this stage, none has yet progressed sufficiently to generate regional environmental competitiveness.

The following recommendations highlight practical ways in which Objective 2 programmes can more fully integrate the principles of sustainable development into regional development practice.

180

7.6.2 Realising Environmental Gain through the Structural Funds - Principles

To achieve sustainable development, the environmental strand of Structural Fund programmes needs to be viewed as integral to the programme, rather than a detached sub-component.

Sustainable development is expected to become the means by which environmental improvement can be secured in the EU in parallel with economic development to provide durable jobs and raise the levels of skills while also promoting social inclusion. This complicated task is one for which Structural Funds programmes are ideally suited. Accordingly, the Commission should now assist partnerships to recognise existing synergies and encourage them to draw these strands together. This does not work against existing practice, but instead identifies good practice and builds on it.

The change in perception requires that partnerships do not view environment and sustainable development as specialist or detached sub-components of the overall programme, but rather that these themes represent core factors for all stages in programme formulation, implementation and evaluation.

Environmental factors should be followed through consistently and meaningfully at every stage of the programming cycle.

Whereas SPDs outline environmental conditions within a region, it is widely recognised that these environmental profiles are often not well integrated with the main parts of the SPD, and that environmental issues are sometimes not clearly focused at project level. In many instances, there is also a lack of baseline data against which the impact of the programme can be measured. To facilitate environmental integration into the post-1999 SPDs, the table below outlines a sequence for considering environmental factors at the different phases of programme development and delivery. Each of the phases is then considered in turn in the recommendations which follow.

181

PROGRAMMING PHASE

ACTIVITY

Preparation of Environmental Profile

Outline existing conditions in the programme area (trends, present and future environmental issues); Outline the institutional framework for tackling these issues; Identify appropriate environmental indicators to measure changes in conditions.

Formulation of Priorities and Measures

Identify specific environmental targets (quantitative and qualitative); Ensure outputs for all priorities and measures encourage good practice in terms of horizontal environmental performance; Conduct a Strategic Environmental Assessment (SEA).

Development of Project Selection and Scoring System

Include environmental criteria in project selection; Encourage good practice and award points for positive environmental impact; Ensure that project assessments by partners are transparent and robust.

Programme Monitoring Use indicators to ensure that the contribution of projects towards environmental aims of programmes can be measured and reported.

7.6.3 Realising Environmental Gain: Programme Development Recommendations

Environmental profiles should be enhanced to improve their ability to inform and direct programmes, bearing in mind the business development focus of SPDs.

SPDs are required to present an environmental profile which details current conditions and subsequently allows (and potentially directs) an assessment of the likely impact of projects funded through the programme. Environmental profiles should act as a starting point for partners to identify key issues for the region and for projects to be funded, and they should also provide baseline data against which subsequent programme performance can be assessed.

To overcome deficiencies, environmental profiles should do the following.

��Include a description and informed analysis of regional conditions, including trends and pressures upon the natural environment, employment in environmental sectors, and sites of importance, supplemented by maps and quantitative information. Such data is essential for the development of projects. Environmental databases and local and regional authority development plans may be directly useful here. In this section, issues relevant to the area as a whole - for example, traffic congestion and resultant emissions – should be differentiated from key issues for projects to be funded through the programme, such as pollution management in individual industries.

��Make a direct connection between environmental information and the aims of the SPD, framing the analysis in terms of the aims of the programme,

182

for example by highlighting environmental issues linked to social and economic investment. This helps in justifying projects and in relating them to the wider context. The environmental impact of previous programmes would be cited in this section. Wider issues relevant to businesses should also be addressed whenever possible, such as material and energy use, waste generation, and the potential for economic and sustainable development they offer.

��Where linkages are identified, the profile should identify the action required to achieve environmental gain through positive action, as well as minimising negative impacts. If they are to be used constructively, these actions must be measurable at project level.

��Highlight any assistance available through competent environmental authorities and other responsible bodies to help achieve the desired results. Depending on the nature of the project, these agencies might include local authorities, public-private sector initiatives, and the voluntary sector.

��In accordance with the categorisation of environmental problems, identify a range of environmental baselines and indicators to monitor the impact of programme measures. Again, these indicators should correspond with the economic development philosophy of Structural Fund programmes.

Following through from the environmental profile, all priorities and measures (both mainstream and environment-focused) should state their environmental relevance, and set output targets for their value added to the environment.

Environmental targets should encompass qualitative and quantitative dimensions: qualitative in the sense of defining an aspiration to work towards, for example creating a successful and competitive environmental sector; and quantitative in the sense of establishing numerical targets against which both the activity and outcomes of supported projects can be statistically measured. To continue the example, these targets might include the number of postgraduate training courses for environmental managers and the associated number and market share of new SMEs appearing in the environmental services sector. The derivation of appropriate targets should be clearly aligned with the assessment of desired environmental impacts.

In integrating environment as a horizontal issue, priorities and measures must be designed in ways which facilitate environmental integration, so enabling mainstream economic projects to contribute towards wider sustainability goals.

New-build projects should: maximise opportunities for energy efficiency; contribute, through siting, towards goals such as reducing the need for road transport; and maximise public access and habitat protection and creation, wherever possible.

183

As an alternative, some programmes have opted to create a separate environmental measure within one of the priorities. This action can disadvantage the horizontal treatment of environment, as programme managers perceive environment as accommodated within a separate budget, and the link between environment and economic development is weakened. An additional feature of creating a separate environmental measure or priority is that it can reduce the requirement for these projects to compete in terms of economic and social impact. Environmental projects should be valued because they are seen to address core programme issues - such as creating the preconditions for growth and job-creation - as well as improving the physical environment.

Consideration should be given to wider and more systematic use of Strategic Environmental Assessments at the programme development stage to derive useful early estimates of the potential environmental impact of programmes.

A number of existing SPDs incorporate matrices showing possible environmental impact by issue, set against different priorities and measures within programmes. The process of assessing programmes at this level is known as Strategic Environmental Assessment (SEA). The technique aims to assess cumulative impacts of a number of individually less significant actions.220 Ex ante SEAs of Structural Fund programmes are necessarily carried out without detailed knowledge of the projects that will come forward, and impacts are therefore uncertain. Nevertheless, useful estimates of positive and negative impacts can be derived, and this information can serve to profile the specific types of projects to be encouraged.

Whereas built developments have the greatest potential to impact on the physical environment directly, projects supported under business development measures could potentially have far greater total impact, if indirect effects are considered. Of course, this will vary, depending on the types of businesses involved. Such indirect impacts could require consideration of issues such as demand for raw materials, the need for expansion of energy generation facilities, or the choice of transport modes.

Under these circumstances, SEAs can highlight the types of activity that could take place and identify preferable or favoured options. This activity should take account of experience gained from assessing projects funded in previous rounds.

220 Clement K, Schremmer C and Tortto H (1998) Utilisation of SEA in Programme Design and Project Selection, Report for European Commission (DG XVI), Brussels.

184

7.6.4 Realising Environmental Gain: Implementation Recommendations

Project selection and scoring systems should reward projects which meet environmental criteria. This element of project assessment needs to be consistent, understandable and justifiable for applicants.

Whereas existing scoring systems often take some account of environmental issues, it is not always clear what this means in practice or how a numerical score is awarded to projects.

Fundamentally, the number of points allocated to the environment within the scoring system should be large enough to encourage good practice. Scoring systems offering between five and ten percent of total points to environmental factors are unlikely to make much difference, other than at the margin when resources are scarce. In effect, this disadvantages good practice by those applicants going beyond minimum environmental requirements.

While it is usually possible to assess the effect of a physical development project will have on the natural heritage, it is generally more difficult to comment on business development or training projects, where impacts may be secondary or indirect consequences of project activity. Nevertheless, it is clear that most activities have some impact, and that there will frequently be a more sustainable alternative - for example, a project using local raw materials is likely to have more benefit to the local economy than a similar project which imports raw materials from outwith the region. The use of local materials is also environmentally beneficial, in terms of reducing the need for transport.

To facilitate the integration of environment and sustainable development into project selection, the following questions could be used to highlight the range of positive and negative environmental impacts a project could have.221 Reflecting conventional themes in Structural Funds programmes, the questions are grouped into four categories with different environmental impacts and characteristics.

QUESTIONS TO ASSESS THE ENVIRONMENTAL IMPLICATIONS OF PROJECTS

Physical projects • = Does the project have positive or negative direct impacts on the environment through

change of land use or construction works?

• = If building land is required, is the site being re-used (ie. brownfield in preference to greenfield sites)?

• = Has the project been sited to maximise access by public transport, pedestrians and cyclists and to minimise road traffic generation?

• = Is energy efficiency and waste minimisation/re-use of materials considered?

• = Are local materials used?

221 Faulk A and Clement K (1998) Environmental Performance and Environmental Integration in Scottish Objective 5b Programmes, Report for Scottish Natural Heritage, Inverness.

185

• = Does the project contribute towards social cohesion, through the involvement of local people?

Business development/community development projects • = Does the project have positive or negative impacts on the environment through the

processes being used? Is there evidence that more sustainable alternatives to the processes being promoted have been considered and assessed?

• = Does the project contribute to the more efficient use of raw materials, especially where these are non-renewable, renewable but over-exploited, or scarce? If materials with a high environmental impact are involved, is there evidence that alternatives have been considered?

• = Does the project add value to local primary materials that would otherwise be exported from the region for processing?

• = Does the project reduce transport demand through provision of local services or by import substitution of products or services previously brought in from outside the region?

Tourism projects • = Does the project enhance the tourism experience through environmental improvements that

would not otherwise take place?

• = Does the project help businesses to promote the green image of the region by reducing waste?

• = Does the project create links with local businesses, reducing imports to the region and helping to maintain the viability of local services?

Training projects • = Does the project maximise the potential for education or training in environmental

management (waste minimisation and pollution prevention, habitat management, use of information technology to minimise travel, etc)?

• = Does the project add to the local knowledge base in ways that will encourage local development of services in the future?

• = Is the project accessible to beneficiaries without access to private transport?

7.6.5 Realising Environmental Gain: Monitoring and Evaluation Recommendations

Monitoring environmental outcomes presents particular methodological problems. Nonetheless, the collection of even relatively basic indicator information can be extremely valuable – if it is undertaken systematically, and if it is seen as integral to the programming process and not a voluntary, one-off optional exercise.

The indicators chosen to measure the environmental performance of programmes during their operational period will often be proxies for those that would ideally be chosen for the region as a whole. This occurs because most natural heritage indicators must be monitored over decades to provide reliable data on measurable change, as there will inevitably be short-term variations not attributable to any single source. Nonetheless, programmes should be able to support the development of structures that will allow long-term monitoring and reporting and the more rigorous definition of future baselines.

186

Recent work on the selection of environmental indicators222 has met with limited success. This is partly because of macro-micro level incompatibilities - ie. the difficulty of translating overall issues into detailed project assessment criteria - but also because of the problems regularly encountered in traditional methods of collecting information through checklists of environmental questions.

Another problem with the environmental checklist approach - which is designed to promote good practice above the minimum required - is the frequent perception that it operates on a voluntary, one-off and detached basis, rather than being integral to the programme as a whole. However, collection rates improve considerably when project forms without a completed environmental checklist are returned to applicants as ‘incomplete’.

Requests can also be made for retrospective information, for example on issues such as:

��Landscape: How many hectares of vacant or derelict land have been improved (including soft landscaping)?

��Village development: Does the project utilise vacant or under-used premises, or how many buildings have been renovated?

��Biodiversity: How many hectares of semi-natural habitat have been enhanced, safeguarded or damaged?

Although such indicators are relatively basic, they have the advantage of being directly relevant to projects funded, and they can serve as the basis for establishing further subsidiary indicators.

Implementing a ‘Cascade System’ for indicators is recommended for the more effective tracking of environmental outcomes. Here, indicators and targets which are meaningful at the overall programme level are disaggregated into a number of indicators relevant to individual projects which combine to help to achieve the overall targets.

Cascade systems follow the progress of issues through programme implementation (see box). There are two key points in this approach:

��indicators for programme monitoring should be distinct from, but must be linked to, project assessment criteria collected at the application stage; and

��the indicators must be seen as an integrated part of the application process (ie. mandatory) if they are to be taken seriously. Otherwise, applicants making positive contributions will use indicators to help promote their projects, while others will give them only cursory attention, weakening the procedural link.

CASCADE SYSTEM

ISSUE

222 For example, the Highlands & Islands Objective 1 Programme, Western Scotland Objective 2 Programme, and Dumfries & Galloway Objective 5b Programme.

187

• = Pollution reduction/improvements in resource use METHOD

• = Promotion of environmental management in business

INDICATORS

• = number of environmental management audits carried out in SMEs

• = recorded savings from waste minimisation initiatives

• = number or proportion of supported businesses adopting a formal environmental management policy

• = number or proportion of businesses adding value to local primary products

In order to operate such a system successfully, information needs to be collected systematically through the project assessment, scoring and monitoring process. If this method is followed, then by the time an application is completed and submitted, it has the following capacities:

��it is able to demonstrate assessed impacts (both positive and negative), on all relevant aspects of the environment, so directly assisting project scoring; and

��it is able to identify environmental indicators against which the project can be monitored, and which will be included as part of the information submitted as the project progresses.

189

8. GENDER MAINSTREAMING IN EU REGIONAL POLICY

8.1 Introduction In common with the environment, equal opportunities has become a universal horizontal element of Structural Fund programmes. This section of the report summarises its rise as an integral dimension of Objective 2 programming, showing how the speed of introduction of ‘gender mainstreaming’ concepts in particular has left some programmes ill-prepared. To deliver on the requirement for gender mainstreaming, responses have to be integrated throughout programming documents and into every stage of the programming cycle, for the first time actively involving both gender specialists and non-specialists alike. This chapter reviews the variety of equal opportunities and mainstreaming practice which has emerged in terms of plan content and implementation and identifies ways in which future equal opportunities responses could be strengthened.

8.2 Context – the Evolution of Equal Opportunities Policy Policy for equal opportunities has evolved incrementally over the latter half of the 20th century, in response to changes in society and the economy and the demands of the women’s movement. The significance of gender issues has grown over time in the EU, to the extent that the European Council meeting in Essen (December 1994) declared that the promotion of equal opportunities for women and men was a key priority of the EU and the Member States, on a par with the struggle against unemployment. The 1996 Commission report, Equal Opportunities for Women and Men in the European Union, reiterated the Union’s commitment to equal opportunities, this time as ‘a basic principle of democracy and respect for humankind’.

In addition to reflecting changes in the wider environment, the responses of the EU over time have reflected its own learning curve and the preoccupations of Member States. Countries with more advanced positions on equal opportunities have pushed for other members to meet their own standards for economic and social justice reasons, via the conduit of EU resolutions and directives. The EU’s involvement in gender issues began with policies for equal treatment, particularly in an economic context, then extended to the active pursuit of equal opportunity. Most recently, and following trends in the wider policy environment, the idea of applying a systematic filter to all policies by ‘gender mainstreaming’ has become a third flank of the policy agenda.

The starting point for equal opportunities in the European Community – Article 119 of the Treaty of Rome, providing for equal pay for equal work – was included partly in response to French lobbying to ensure that other early members matched its own responses to equal opportunities (which were the most advanced among the Member States) and so eliminate the unfair competitive advantage potentially provided to them by a cheaper female workforce.223 ‘Objective 1’ Member States have tended to be followers in this

223 van Overbeek J (1994) Handbook on equal treatment between men and women in the European Community, Equal Opportunities Unit, DG V, Commission of the European Communities, Brussels.

190

policy area, and so have changed most as a result of the ongoing, indirect policy push of other Member States, mediated through the framework of the EC, and, latterly, as a result of the growing momentum of the EC’s own initiatives. More recently, the accession of Nordic states to the EU, with their more sophisticated, integrated approaches to gender issues, has pushed the issue still further across a wider group of Member States.

Expansion of the EC’s involvement in gender issues has several aspects, which have developed incrementally. First, a set of consistent legal provisions has gradually accumulated (see box). Legal frameworks have been imperative in setting the ground rules, but have not been - and indeed could not be - sufficient to deliver equal outcomes. Overall, their influence has been limited; this includes failing to close the gap between average male and female incomes, which persists largely because of continuing gender differentiation between occupations and the difficulty in demonstrating the ‘equivalence’ of different types of work and thus of their resultant remuneration. Also, despite relevant legislation, inequality cases are rarely brought to trial.

EU DIRECTIVES ON THE EQUAL TREATMENT OF MEN AND WOMEN

European Directives on the equal treatment of men and women adopted since 1975 have established the following rights and principles:

• = equal pay for equal work;

• = equal treatment in terms of recruitment, training, promotion and working conditions (the directive in question underpins measures banning all sex-based discrimination at work);

• = equal treatment in matters of social security;

• = equal treatment under occupational social security schemes;

• = equal treatment for self-employed women in the various sectors, including agriculture;

• = right to paid maternity leave and guarantees concerning health and safety at work (protection of pregnant workers);

• = parental leave (June 1996 Council Directive);

• = COM(97) 202 final, of 14 May 1997 - provides an amended proposal for a Council Directive on the burden of proof in cases of discrimination based on sex.

Second, because gender has persisted as ‘the most significant determinant of occupational lifechances’,224 there has been a need to go beyond passive (legal) provisions to pro-active measures: positive action and positive discrimination. Positive action comprises interventions which seek to make good the deficiencies of an ‘underperforming’ group – helping them to ‘fit in’ better to the labour market by addressing their specific needs eg. the barrier posed by caring responsibilities and different behavioural patterns. Positive discrimination, outlawed in some countries, involves redressing imbalances by unequally benefiting the under-represented group. Since the 1980s, the EU has supplemented the legislative framework with specific programmes to implement equality in practice. These initiatives include the Community’s Action Programmes, launched in 1982, and the NOW element of the

224 Rees T (1998) Mainstreaming Equality in the European Union, Routledge, London.

191

Employment Community Initiative, which began in 1991. Although with limited resources, these initiatives are considered by the Commission to have had ‘a substantial knock-on effect, particularly by stimulating further action in the individual Member States’.225

The Action Programmes have provided a format to develop guidelines for medium-term action at European level and for achieving equal opportunities by changing attitudes and breaking down cultural, economic and social barriers. The First Action Programme was launched in 1982 with subsequent programmes introduced at four-year intervals; the Fourth Action Programme (1996-2000) is currently entering its closing stages. The focus of these programmes has evolved over time. In the context of Structural Fund programming, early positive action was first concentrated on specific, dedicated ESF programmes targeting women, notably NOW. At this stage, positive action projects are taking their place in other general economic development programmes, including those for Objective 2.

Positive action can make an important contribution, either through stand-alone measures or as part of wider programmes, but it does have deficiencies. Projects targeting women frequently aim to assist them to fit in better with the expectations and operating modes of a status quo which is male-oriented. There has been increasing pressure for an approach which, ‘(r)ather than fitting women into existing structures and systems, or seeking to adjust structures to ensure a better fit…involves designing programmes and projects informed by knowledge of the diversity of needs of potential participants’. This can involve broadening the ‘norm’ to encompass a wider range of working practices and life patterns, and challenging structures and organisations which, while they are assumed to be gender neutral, are in fact male-oriented. This in turn can enable more actors to achieve more in the economic sphere.

The need for gender awareness across all institutions and interventions has led to the third aspect of EU action – the movement for ‘gender mainstreaming’. This concept, which is central to the EU’s current policy for equal opportunities, was originally given formal expression at the UN Third World Conference on Women held in Nairobi in 1985, where the objective was to ensure that women’s values were taken into account in development work. It was deemed necessary to target and take account of this group more effectively, on the one hand because they are more frequently burdened by poverty and exclusion and enjoy the least choice in life chances, and on the other hand because they are pivotal actors in the socio-economic fabric, frequently delivering educational and care services which it would otherwise be the task of governments to assume.

The idea was then explicitly endorsed at the UN Fourth World Conference on Women in Beijing in 1995, with the adoption of a Platform for Action. This Platform states that ‘governments and other actors should promote an active and visible policy of mainstreaming a gender perspective in all policies and programmes, so that, before decisions are taken, an analysis is made of the

225 CEC (1996) Equal Opportunities for Women and Men in the European Union, Annual Report 1996, Office For Official Publications of the European Communities, Luxembourg.

192

effects on women and men, respectively’.226 The drive for greater gender awareness in policy-making and delivery in this context is motivated by both human rights objectives, and socio-economic issues. Among the economic principles is that the ‘eradication of poverty based on sustained economic growth, social development, environmental protection and social justice requires the involvement of women in economic and social development’. As such, the Platform aims to ‘promote women’s economic independence, including employment, and eradicate the persistent and increasing burden of poverty on women by addressing the structural causes of poverty through changes in economic structures’. In this, women are viewed not just as beneficiaries, but also as ‘vital development agents’, who can only achieve their potential with equal access to productive resources, opportunities and public services.

THE MAINSTREAMING PRINCIPLE – SOME EXAMPLES

Mainstreaming is a global approach which involves developing a gender perspective for all policies. It is defined by the Commission as ‘the systematic consideration of the differences between the conditions, situations and needs of women and men in all Community policies, at the point of planning, implementing and evaluation, as applied to Europe, the industrialised countries and the developing countries’. Its aim is to ensure that opportunities are genuinely available to the spectrum of potential participants and that initiatives have no negative gender impacts.

An application of the mainstreaming principle in transport policy would involve explicit consideration of the fact that women are more frequent users of public transport and less often own or have access to a car. Women also regularly travel with children, use prams and pushchairs and require predictability in timing their travel, in order to be able to meet often complex domestic and professional obligations. The development of good quality, efficient and appropriate transport systems, which take into account accessibility for the whole range of likely users, contributes to equal opportunities.

In response to the Platform for Action, a sustained effort is being made progressively to integrate gender mainstreaming into policies for both the developing and the developed world. The EU has joined this endeavour, promoted by political support at the highest levels,227 pursuing the principle through its Third and, in particular, Fourth Medium-Term Community Action Programmes for Men and Women. The Fourth Action Programme (1996-2000) is seeking to ensure that an awareness of equal opportunities becomes an integral part of all the Community’s economic and social policies, including the Structural Funds, in order to achieve the maximum number of women in employment, and to upgrade their contribution to economic, social and public life. Supporting this effort, a Communication on mainstreaming adopted in February 1996228 took stock of progress and made suggestions for future priority areas for action. In particular, the Communication underlined

226 UN Fourth World Conference on Women (Beijing, 1995) Global Platform for Action, United Nations, New York. 227 Council of Europe (1998) Gender Mainstreaming: Conceptual Framework, Methodology and Presentation of Good Practices, EG-S-MS (98) 2, Strasbourg. 228 CEC (1996) Incorporating Equal Opportunities for Women and Men into all Community Policies and Activities, Commission of the European Communities, COM (96) 67 final of 21.2.96, Brussels.

193

the considerable further potential for the Structural Funds to pursue equality of opportunity in the economic sphere.

The problem of gender inequality is common to all EU Member States. In spite of progress to date, broad spatial patterns of economic disadvantage are still counter-crossed by pronounced gender inequalities. ‘Flagrant inequalities still exist between women and men in the EU as regards training levels, access to the employment market, levels of activity and of participation in the decision-making process.’229 Unemployment is in many cases higher among women than men (exceptions being in the UK and Scandinavian countries), and, in every Member State, it is higher among women with children. This latter pattern varies depending on the degree to which the state takes an active role in meeting caring responsibilities for the young and elderly.

In the labour market, activity rates for women vary significantly between Member States, but are consistently lower than for men. In employment, women are adversely affected by three types of gender segregation: horizontal (where men and women work in different sectors); vertical (where they achieve different degrees of seniority); and contractual (where there are differentials between temporary and permanent and full- and part-time working, and the associated quality of conditions and remuneration).230 The career progression of women and men also varies, with females disadvantaged by their greater likelihood to take career breaks to attend to caring responsibilities. Rates of entrepreneurship are also lower among women. Almost ten percent of women were self-employed in 1992, compared with 18.9 percent of men. In addition, women form only a quarter of all new entrepreneurs.

The character and perception of inequality, and the policy responses to it vary across the Union, reflecting different socio-economic, cultural and legal structures. Influenced by leading Member States (as above), the Commission’s policies reflect the different regulatory and legal provisions at national level. Among EU Member States, the Nordic countries are widely considered to be those where equal opportunities are at their most developed, with strong integration into legal frameworks, and a vision which aims to facilitate the activities of both women and men and the reconciliation of family and working life. However, even here, the results in terms of Structural Fund programmes, and mainstreaming in particular, have not been as good as their overall commitment to gender equality would imply, reflecting the genuine novelty of this policy approach.

The rationale for taking gender into account in Structural Fund programming in particular lies both in equity arguments (the agenda of social responsibility and human rights) and in efficiency arguments (the potential negative economic impact of the under-utilisation of human resources). The first economic argument for integrating gender awareness into regional development policies is quantitative: the greater the proportion of the

229 CEC (1996a) Women, players in regional development, Commission of the European Communities, Luxembourg. 230 Maruani M (1995) Inequalities and Flexibility, in CEC, Equal Opportunities for Women and Men: Follow-up to the White Paper on Growth, Competitiveness and Employment, Report to the European Commission’s Task Force (Directorate-General V) Brussels, V/5538/95-EN.

194

workforce involved in economic activity, the better for the economy. The second is qualitative: to ensure their competitive edge, regions and the firms within them must ensure that the best among the available workforce – whatever their gender - are employed, and at a level which maximises their contribution.231 Where the role which individuals play in the economy is strongly shaped by gender differentiation (reflected in particular in the uneven distribution of social responsibilities), rather than merit or ability, then it follows that the capacities of one group may be being underutilised while those of another are overutilised, leading in turn to inefficiency and reduced output.232 The clear differences between male and female patterns in the labour market indicate that, in order for economic development policies to maximise their impact, they need first to understand better the causes for inequality and second to be tailored in such a way as to address the specific circumstances of sub-groups. It is only then that the suppressed potential of all individuals as the agents of regional development can be freed up, for the benefit of the economy as a whole.

8.3 Gender Equality and Objective 2

8.3.1 1989-93: Limited Positive Action – and outside Objective 2 In 1988, the Structural Fund regulations made no specific reference to equal opportunities, nor was gender mainstreaming yet an official policy dimension of Structural Fund programming. Dedicated, mainly ESF-financed, programmes did emerge in response to identified issues, with the objective of undertaking positive action to create or facilitate opportunities for women and other disadvantaged groups. However, Objective 2 programmes themselves did not contain a systematic equal opportunities dimension, either in terms of the regional analysis they undertook or their resultant policy and project responses.

During that period, the best-known, specific instrument supported by the Structural Funds to improve women’s situation in the labour market through positive action was New Opportunities for Women (NOW), under the Employment Community Initiative (CI). This was mainly co-financed by the ESF (with a small proportion of ERDF). NOW, which helped to pursue some of the objectives of the Third Action Programme adopted in 1990, was developed at the same time as the other CIs for 1989-93, but was only launched in 1991, midway through the 1989-1993 programming period. It set out to tackle segregation in the labour market and growing social exclusion in Europe, both of which have a considerable impact on women. In its first period, with a budget of 156 MECU for three years, NOW had considerable success, assisting 800 projects, 300 of which were aimed at the creation of small businesses or co-operatives. Other projects set out to raise female awareness of, and access to, labour market opportunities. In a wider Structural Funds context, women also benefited from ESF funding of measures to train and help young people and the long-term unemployed. However, the

231 Beck B (1998) Women and Work Survey - Balancing Act, The Economist, 18 July 1998, pp15. 232 For a broader discussion of this issue see Adnett N (1996) European Labour Markets - Analysis and Policy, Longman.

195

resources earmarked specifically to support women were relatively small - only five percent of ESF appropriations (380 MECU).

8.3.2 1994-99: Positive Action and Mainstreaming across all Structural Fund Programmes For the 1994-99 programming period, the positive action programme NOW was renewed as a separate and specific initiative233 following strong lobbying and a positive appraisal of its achievements. There were, however, some modifications, including a broadening of eligibility rules and an extension of its activities to integrating best practice into mainstream systems of training and employment. At the same time, the Structural Fund regulations approved in 1993 made the pursuit of ‘equal opportunities’ outcomes an explicit goal to which all structural measures should contribute.234 The EC became responsible for ensuring that equal opportunities was developed as an integral part of all Structural Fund programmes. The references were stronger for the ESF, with Article 1 of the ESF's regulation requiring the Member States and the European Commission "to ensure that operations under the different objectives respect the principle of equal treatment for men and women".

In spite of the new regulations, the active integration of the theme into programmes other than NOW after 1994 was slow to develop, with most progress being made in Objectives 1 and 3. In many Member States, equal opportunities was translated into a specific priority under Objective 3 (Austria, Belgium, Spain, France, Italy, Luxembourg, Germany, United Kingdom) and into the allocation of human resources appropriations under Objective 1 (Ireland, Portugal, Italy, Greece, United Kingdom, Germany). At the same time, equal opportunities was a very limited element of the Objective 2, 5b and 4 programmes drawn up in 1993.

Under Objective 2, gender-oriented equal opportunities elements of the 1994-96 SPDs largely consisted of standard clauses concerning respect for relevant legal frameworks. A similar situation prevailed in the new Member State programmes launched in 1995, although with some change in emphasis in Sweden and Finland particularly, reflecting the greater status given to equal opportunities issues by these Member States.

There is an interesting contradiction here. While few programmes had clear proposals for gender awareness in programme implementation, most SPD economic analyses identified a need for specific equality and inclusion-driven interventions and for wider awareness of the distributional impacts of interventions, highlighting the presence of individuals whose employment prospects were worse than those of the rest of the population. Among the disadvantaged groups highlighted were not just women, but also ethnic minorities (East London and the Lee Valley, East Midlands), young people (French and Belgian regions), the long-term unemployed, the disabled (Nordrhein-Westfalen, Schleswig-Holstein and Cataluña) and those nearing

233 CEC (1994) The Future of Community Initiatives under the Structural Funds, COM(94)46 final, Brussels, 16/3/94. 234 Regulation 2081/93 (OJ No L 193, 31.7.93, p5), Regulation 2082/93 (OJ No L 193, 31.7.93, p20), Regulation 2083/93 (OJ No L 193, 31.7.93, p34), Regulation 2084/93 (OJ No L 193, 31.7.93, p39) and Regulation 2085/93 (OJ No L 193, 31.7.93 p44).

196

retirement (Twente, Oberösterreich and Vorarlberg).235 The gender dimension itself was multi-faceted: Nordrhein-Westfalen had low participation rates among women; Niederösterreich highlighted low quality jobs for women; while regions like Fyrstad, Kokkola and the East Midlands were addressing rising female unemployment rates as a result of rationalised employment opportunities in the public sector or labour-intensive sectors such as textiles. In some cases, including Auvergne, Saarland and Twente, the decline in male-oriented traditional industries meant that there was greater concern expressed over male unemployment rates. Among the few 1994-96 Objective 2 programmes with specific policy responses to the gender dimension of inequality was East London and the Lee Valley, with measures helping women, including from ethnic minorities, to enter the mainstream labour market through child care provision and training, and enabling the integration of homeworkers (usually women) into the formal economy.

The main step forward in integrating responses to unequal opportunities took place with the drafting of the 1997-99 Objective 2 programmes. The Commission provided a much stronger steer in its guidance for these, drawing attention to the Structural Fund-related advice contained in the 1996 Council resolution on mainstreaming.236 The role of the Structural Funds in equal opportunities policy was to ‘reduce inequality of opportunities between men and women with regard to the rate of employment, the level of training, access to the labour market and involvement in the decision-making process’.237 The resolution encouraged, inter alia, a deeper diagnostic process and the support of operations able positively to contribute to equal opportunities. The Commission’s guidance note on reprogramming priorities was supplemented or reinforced by specific input from designated desk officers during the programme preparation phase. In some cases (eg. France), further formal advice was issued by national authorities on common approaches to interpreting this aspect of the EC guidance.

The greater equal opportunities content of the 1997-99 Objective 2 programmes, which resulted largely from the EC priorities, is reviewed in detail below. In summary, while the Commission’s requirements resulted in many programmes placing more explicit emphasis on this theme, responses overall were limited, with only 16 out of 73 programmes taking full account of the DG XVI guidelines according to Commission analysis,238 and many responses appearing superficial or ‘tokenistic’. The main responses comprised: statements of general compliance with equal opportunities legislation (as previously); the inclusion of more gender-disaggregated statistics and (less often) analysis in the regional profile; and some programme objectives relating to gender outcomes or equality of access to opportunity. In addition, many programmes invited ‘positive action’ projects to address identified imbalances,

235 Bachtler J, Taylor S with Kearney C (1996) Extended Synthesis of Agreed Single Programming Documents in Objective 2 Areas, 1994-96, Report to European Commission (DG XVI), EPRC, University of Strathclyde, Glasgow. 236 CEC (1996a) op. cit. pp1. 237 CEC (1996b) Resolution on mainstreaming equal opportunities of men and women into the European Structural Funds, Official Journal C 386, 2.12.1996, pp1. 238 Wulf-Mathies M (1998) Information Note from Mrs Wulf-Mathies to the Members of the Group of Commissioners on Equal Opportunities for Men and Women and Women’s Rights, DG XVI, Brussels, 24.2.98.

197

targeted at women, but forming integral parts of general measures. Arguably, the Commission’s hope (but which was not realised) was gender mainstreaming: the application of a ‘gender filter’ to all policies and projects, enabling the identification of their potential gender bias and adjusting them accordingly.

The relatively limited responses of Objective 2 programmes were disappointing, but less so given the magnitude of the demands being made of them. While equal opportunities policy had evolved incrementally over several decades, through the accumulation of legislation, increasing positive action initiatives and, most recently, exploration of a mainstreaming approach, most Objective 2 programmes were being asked to move, with very limited preparation, directly from taking no specific action in the field of equal opportunities to integrated ‘mainstreamed’ responses spanning all policies, and with elements of positive action where appropriate. The programmes were also asked to mainstream gender awareness before most of the domestic policy environments with which they mesh – even in Nordic states - had time to elaborate their own responses. Programme responses were also expected to span both the ESF and the ERDF. The ESF poses relatively few problems, with practice being established outside Objective 2 programmes (for example in NOW) and a clear link to individuals – ie. quantifiable men and women as beneficiaries and participants. There was less experience under the ERDF, however, and more complications in linking these interventions to their gender outcomes.

The achievement of gender mainstreaming has to be regarded as a long-term objective, requiring ongoing work. It is positive, therefore, that in addition to maintaining equal opportunities as the theme of a CI in the 2000-06 programming period, continuing emphasis will be placed on all Objective 1, 2 and 3 programmes moving towards gender mainstreaming, and accommodating positive action as one of the dimensions of this activity.

8.4 Integrating Equal Opportunities: the 1997-99 Programmes ‘Mainstreaming’ is a genuinely new policy approach to equal opportunities, and not only in Structural Fund programming. Conscious of the difficulties involved in responding to this agenda since 1994, often ahead of responses being elaborated for domestic economic development policy, a series of thematic studies has already been undertaken at European, Member State and programme levels. Overall, significantly more attention has been paid to disseminating examples of ‘positive action’ than providing workable solutions on how to gender mainstream, but, as experience with mainstreaming accumulates, studies of increasing utility are being produced. The EC requested several country-specific reports on applying the principle of equal opportunities in Structural Fund programmes (eg. for France, Finland and Ireland),239 and has organised a series of international events to identify and exchange best practice, notably in Portugal (1998) and Germany (1999). The

239 Studies Commissioned by DG XVI include: Ferres M (1997) États des lieux des bonnes pratiques relatives à l’application du principe de l’égalité des chances dans les interventions du FEDER en Espagne; Horelli L and Roininen J (1997) Gender Aspects in the Application of ERDF Interventions in the Finnish Context in 1995-96; and Murphy C (1997) Structural Funds and the Application of the Principle of Equal Opportunities in Ireland.

198

latter event, which focused specifically on mainstreaming, was linked with a comparative study of best practice in France, Germany and the UK. A further valuable international initiative has been a review of practice in gender mainstreaming undertaken for the Council of Europe. In France, a comparative study into regional strategies for equal opportunities was commissioned with funding from the national Objective 2 technical assistance programme. In addition, work is ongoing at EC level and in selected Member States to develop new analytical systems, including evaluation tools, with ‘gendered methodologies’ for use with Structural Fund programmes.240 At the programme development stage in 1996, the involvement of equal opportunities experts in consultation and development processes offered the best possibility to orient programmes around the new gender issue in a short timescale. In the longer term, effective responses to gender mainstreaming require the active involvement of non-specialists, integrating gender awareness mechanisms into their ‘mainstream’ activities – although still supported, for the more technical aspects, by specialists in the field. This means at least a basic level of expertise has to be disseminated widely. There is still an urgent need for more programme actors to appreciate with more clarity what gender mainstreaming is, and how it can practically be integrated into specific contexts.

Few Objective 2 SPDs provided adequate descriptions of their consultation processes in preparing the 1997-99 programmes, but it appears that many did consult relevant equal opportunities specialists. Where this took place, including in many French and UK areas, the results varied according to the experience, knowledge, approach and abilities of the regional equal opportunities or women’s officers. Those with some previous experience of economic development (and preferably Structural Fund) programmes were the best placed to provide a constructive input. In most cases, however, two learning curves were involved – in gender issues for the Structural Fund partners, and in Structural Fund programming for equal opportunities specialists.

It is difficult to judge the number of specialist organisations each programme mobilised (see Table 8.1). In Industrial South Wales, the Equal Opportunities Commission for Wales, the Chwarae Teg (Fair Play) organisation and the Welsh European Programme Executive (WEPE) were all involved, while in Western Scotland, the Equal Opportunities Commission was supplemented by Glasgow City Council (which has developed women’s business initiatives) and the Scottish Office. In Aquitaine, the input of the regional women’s rights officer was supported by regional labour market analysis undertaken by the regional training co-ordinator, the DRTEFP. In Picardie, the French statistical office ‘INSEE’ provided quantitative information. In some other cases, consultants were hired to provide input.

240 Horelli L (1997) The Quest for gendered Methodology of Evaluation in the Context of Structural Fund Interventions, Paper presented to the Conference of the European Evaluation Society, 508 March 1997, Stockholm, Sweden.

199

Table 8.1: The identity and role of representatives of equal opportunities issues Member State

Agencies/agents responsible for equal opportunities issues at programme level and their role

Austria Not specified in the SPDs. There is a Federal Minister for Women’s Affairs and a federal commission for equal opportunities. At Land level, there is a Frauenreferentin, a consultant working to promote women’s issues, sometimes working with the back-up of a small office, the Frauenreferat. Commissions for equal opportunities are also being established at Land level.

Belgium Not specified in the SPDs. Denmark No specific specialist agencies appear to have been involved at the regional level.

Instead, responses appear to have been internalised to programme management teams, in consultation with the partnership.

Finland Not specified in the SPDs. There are no specific equal opportunities organisations at the regional level. Instead, equal opportunities work is undertaken within programme management teams in the regional councils. No external experts, agencies or academics are involved.

France In each region, the authority responsible for taking the lead on equal opportunities issues is the Délégué Régional aux Droits des Femmes. In most cases, they first became involved in Objective 2 in the lead-up to the 1997-99 programmes. The agency is mentioned explicitly in a few SPDs (eg. Aquitaine, Basse-Normandie). Additional agencies involved can include the state’s training agency in the regions, the DRTEFP (which undertook gender-disaggregated labour market analysis for the Aquitaine SPD). The regional Délégués could play roles at several stages of programming: providing statistical analysis for programmes, being consulted on programme content, and advising on some issues of project selection. They are now invited to attend Monitoring Committees.

Germany While it is specified in few SPDs, the representative of equal opportunties issues would be expected to be a representative of each Land’s relevant ministry. In Nordrhein-Westfalen it is the Ministerium für die Gleichstellung von Frau und Männer des Landes NRW. In Saarland, it is the Ministerium für Frauen, Arbeit, Gesundheit und Soziales des Saarlandes.

Luxembourg Not specified in the SPD. Spain Not specified in the OPs. Sweden The Regional Expert for Equality Affairs, within each County Administrative Board

plays the lead role. CAB equality experts become involved in Objective 2 programmes through their responsibility to promote the integration of a gender perspective into regional policy fields.

United Kingdom

Of the SPDs which state who will be responsible for equal opportunities, most state that elements of the wider partnership have a legitimate role in raising and pursuing relevant issues. North East England additionally highlights the Equal Opportunities Commission. The Industrial South Wales programme is also specific, highlighting the Equal Opportunities Commission for Wales, the Chwarae Teg (fair play) organisation and the Welsh European Programmes Executive (WEPE). Western Scotland highlights Glasgow City Council, the Equal Opportunities Commission, and the Scottish Office.

Where limited or no consultation appears to have taken place, this was sometimes because there were no local specialist agencies (eg. in Denmark and Finland, where the issues are internalised to programme actors), or because equal opportunities issues were already considered to be well understood and highly integrated into national and regional government policies (eg. Sweden). There is a weakness in the latter case in that gender

200

mainstreaming is new, and even Member States with strong equal opportunities legislation and a culture which accepts the concepts involved have to adapt to it (as confirmed by the supranational gender mainstreaming policy project launched in 1997 by the Nordic Council of Ministers).

The involvement of experts is crucial: ‘they can provide a dynamic force for change, putting pressure on other official bodies and providing relevant know-how and information’.241 Where they are in place, it is strongly recommended for them to be involved. However, their input must be carefully managed, involving them as early as possible and assisting them in building up a familiarity with Structural Fund programming. Their involvement should be continued through into the implementation phase and should be targeted at points which enable them to influence programme outcomes effectively.

Where mainstreaming is the objective, a total reliance on the input of experts is insufficient. Equal opportunities can only really be ‘mainstreamed’ by disseminating the concepts and approaches proposed to the so-called ‘agents for change’ – including applicants, and ensuring they can apply them. In addition to equal opportunities experts applying their expertise to programmes, they should actively be involved in transferring relevant skills and concepts to mainstream programme players. Recent research funded by DG XVI highlights some best practice examples in which programmes have undertaken initiatives to ensure the active dissemination of concepts and skills.242 Among the examples are the ‘Trans-Faire’ project in Rhône-Alpes. Supported by the Fourth Action Programme on Equal Opportunities (1996-2000), this project enabled significant progress to be made in communicating the mainstreaming approach to partners and developing the necessary instruments to operationalise the concepts across the Objective 2 and 3 programmes. In Industrial South Wales, similar work was undertaken under the Objective 2 technical assistance budget, training programme and project managers, disseminating base-line data and establishing suitable project selection criteria. These initiatives could be replicated cost-effectively across other programmes, capitalising on the materials and methods they have developed.

8.5 Equal Opportunities as a Dimension of Programme Development In response to wider policy trends in the EU, the DG XVI guidance to Objective 2 programming authorities for developing their 1997-99 programmes required for the first time that the new programmes take gender issues into account. The addition of a new and complex issue to the programme development agenda was a significant challenge, especially when responses were impeded by a range of problems, including:

��the lack of time available to gain familiarity with the issues (a stronger disadvantage outside the Nordic states);

241 Braithwaite M, Fries R and Fitzgerald R (1998) Mainstreaming Equal Opportunities for Women and Men in the Structural Funds: Findings of a survey of current practice in Germany, France and the UK, Study undertaken for DG XVI in connection with the DG XVI seminar on mainstreaming, Gelsenkirchen, 21-22 January, 1998, p5. 242 Ibid.

201

��the lack of precedents on which to draw, gender mainstreaming being a new concept, compounded by the perceived lack of timely practical advice and support from the Commission;

��confusion about what was being requested. Different actors in the same partnerships interpreted the instructions and concepts differently, depending on their background and experience. In addition, conflicting information sometimes came from the Commission; and

��‘issue overload’, with improved employment outcomes, indicators, quantification and responses to RTDI promotion and the environment also expected.

The importance ultimately placed on the equal opportunities dimension varied significantly between programmes. This was the result of the interplay between a number of factors, notably:

��the degree of importance accorded to the theme at national, regional and programme level (including by programme managers and different members of the partnership - reflecting both their institutional and personal positions);

��the availability of accurate and up-to-date information on gender inequalities and their causes;

��the intensity and nature of the known inequalities;

��the availability of practical, specialist support for programme and project development; and

��the amount of prior experience in the field.

The equal opportunities outcomes of the programme development process for current SPDs are described in the sections which follow. Ideally, a diagnostic phase would be undertaken first, identifying and interpreting patterns of gender inequality in the eligible area and the gender-disaggregated outcomes of previous economic development interventions. This would then inform the details of policy development, leading both to proposals for positive action where appropriate and, ideally, the integration of gender sensitivity into other policies. In practice, given the lack of time and expertise, a common final response by programmes was for them to set out the main, known equal opportunities issues, to state adherence to the broad concept of reducing inequality, and to invite programme applicants - the actors at the end of the chain - to translate this commitment into their projects.

8.5.1 Diagnosis: Equal Opportunities Analysis in the 1997-99 Programmes The identification of patterns of gender inequality is an essential first step to developing targeted gender-oriented policies, introducing gender sensitivity into other policies and establishing baselines from which to monitor the gender-differentiated impact of programmes. The inclusion of gender disaggregation within regional analyses was limited, especially in Spain, Finland and selected UK, French and German programmes (Table 8.2). Gender-disaggregated data is usually part of the labour market analysis, although a small number of (French) SPDs dedicate a separate section to the

202

theme, setting out the issues alongside the proposed responses (eg. Bourgogne, Basse-Normandie, Centre).

Table 8.2: Equal opportunities analysis in the 1997-99 programmes243 Member State Equal opportunities assessment in the regional profile Austria In all four programmes, the description of the eligible area labour market includes

statistics disaggregated by gender. Belgium There is limited or no data or analysis in the Walloon programmes, in spite of ex ante

criticism. In the case of Aubange, the limited range of activities being undertaken, and the lack of an ESF flank explain the deficit.

Denmark Detailed information is provided for Nordjylland in a separate five-page section, giving extensive indicators and providing contextual information.

Finland Limited assessment is contained in the national SPD for the Finnish Objective 2 areas. First, equal opportunities issues are considered already to be well understood and addressed, and second, the national programme was designed to be elaborated with regional-specific information for individual eligible areas.

France The approach taken varies between programmes. A minority contain little or no information (eg. Bretagne). Among those providing an equal opportunities assessment, some integrate gender disaggregated statistics into the body of the regional analysis (Alsace, Aquitaine, Pays de la Loire, Picardie), while others present it separately (Basse-Normandie, Centre, Champagne) or as an annex (Bourgogne).

Germany Most German programmes contain some analysis of the gender situation, generally integrated into the labour market analysis. One programme (Bayern) has no analysis.

Spain No gender-disaggregated assessment in any programme’s regional profile. Sweden The Swedish programmes set out a brief overview of gender differences in the labour

market. United Kingdom Various responses are found in the UK programmes: almost half contain no gender

disaggregated assessment (eg. East Midlands, Gibraltar, Plymouth, West Cumbria and Furness). Of those which do, most are thorough (eg. Eastern Scotland, Industrial South Wales, Western Scotland and North East England, where elements are derived from the work of the Labour Strategy Information Group).

The quality and quantity of data provided does not necessarily relate to the degree to which equal opportunities issues are integrated into economic development practice. For example, Nordjylland provides a detailed profile of the equal opportunities situation because this data is freely available and the policy area is already well embedded. Finland cites the same reasons for providing very little data within its SPD. In some cases, analytical deficits are due to the lack of available statistics. Swedish and some French programmes highlighted problems with obtaining detailed labour market statistics for the eligible areas, which impeded analysis. In some SPDs, a specific data collection exercise appears simply not to have been undertaken.

The quality, completeness and relevance of the data provided varies significantly. A wide range of indicators is needed, not only describing employment and unemployment rates, but also differences in the quality of

243 1995-99 for Austria and Sweden.

203

employment experience. Braithwaite244 lists the following requirements for gender-disaggregated data in any regional profile:

��rates and experiences of unemployment and underemployment;

��participation in the formal labour market (including horizontal and vertical patterns);

��participation in informal labour and domestic work; and

��education and training (including skills and qualification levels).

Braithwaite further suggests that gender-disaggregated statistics should be gathered on participation in economic development support measures (eg. self-employment, training, reintegration) and on the availability of services important to equal opportunities, including care services for dependents and public transport. It is rarely possible to identify single causes and, therefore, simple solutions to gender imbalances in the labour market. Causes relate to a complex of economic and socio-cultural factors. As such, a more detailed overview is required than most programmes currently provide.

The quality and completeness of information available in SPDs limit their utility for drawing up an overview of the gender situation in Objective 2 areas. In addition to the data deficiencies, not all programmes analyse the same issues in the same way. Some compare the situation of men with that of women within a given eligible area, while others compare female and/or male data in the eligible area with other areas, the wider region or national averages.

Notwithstanding the limitations of information provided, a synthesis of SPD information produces several conditions. First, women face a difficult employment situation. Women variously form a minority of the working population (Basse Normandie), a majority of the unemployed (Basse Normandie, Poitou-Charentes), and a majority of the young unemployed (Basse Normandie). In some cases (UK), female employment, and the female share of the working population are growing, but mainly entail women taking up part-time, low-skilled and low-status jobs.

Second, there are distinct sectoral employment differences between women and men. In Sweden and Finland, the clearest pattern is female dominance of the public sector and male dominance of industry, both of which have contracted. In other regions, female employment tends to be concentrated in low-order services and specific industrial sectors, while males are concentrated in higher-order service and industrial occupations (eg. Vorarlberg and Basse Normandie). In Aquitaine and Champagne-Ardenne, gender concentration means that female unemployment is rising as a result of decline in the leather and/or textiles sectors. Frequent patterns are the greater concentration of women in lower paid, lower skilled or more precarious jobs (West Midlands, Industrial South Wales, North East England), and their under-representation in management (Industrial South Wales), academia (Niederösterreich), the professions (Poitou-Charentes) and among the self-employed (Finland). Shorter contract lengths and higher turnover are a feature

244 Braithwaite M (1996) The Promotion of Equal Opportunities between Women and Men in Industrial and Urban Regions within the Framework of Objective 2 1997-99, DG XVI Training Seminar Discussion Papers, October 1996, Commission of the European Communities, Brussels.

204

of female employment in selected regions (variously, in Aquitaine, Basse Normandie, Pays de la Loire), such quality issues also leading to women’s careers being less structured, with consequences for average remuneration.

Third, there is a mismatch between employment vacancies and the profile of the men and women seeking work. While skills grow in importance for economic competitiveness, women remain less skilled in many cases than their male counterparts (eg. in Bourgogne). Fewer women are highly qualified (Niederösterrreich, Oberösterreich, Poitou-Charentes), and fewer undergo on-the-job training – partly because of the nature of the jobs they hold and/or the sectors in which they are employed (eg. Champagne-Ardenne).

Fourth, the availability of women for conventionally organised work is also affected by caring responsibilities for children and ageing family members, which still fall largely to women. Career prospects are limited for those who have ongoing caring responsibilities or who have taken career breaks for this reason. For those with ongoing responsibilities, access to work or training may be affected by limited mobility - especially in rural areas and for those in lower income groups. They may also be affected by inappropriate childcare provision. Childcare facilities may be lacking entirely, they may be too expensive relative to earnings or they may be incompatible with actual needs (eg. due to their opening hours or the ages accommodated). In other cases, regulatory issues may affect the employment prospects of women, for example the legal preclusion of women undertaking night shifts in Austria.

Lastly, some patterns of disadvantage affect men rather than women, but there are very few examples where such patterns are highlighted. Exceptions are descriptions of rising unemployment for women and men in North East England and patterns of higher female employment in Swedish Objective 2 areas.

8.5.2 Strategy Definition At the level of the strategy statement, the visibility of the equal opportunities dimension varies (see box). The theme is not explicit in the strategy statements of programmes in Wallonia (Belgium), the Netherlands or Spain, except in their commitment to respect relevant legislation in this area. The picture is mixed in France, Germany, the UK and Austria. In France, programmes not mentioning equal opportunities explicitly include: Alsace, Bretagne, Centre, and Picardie (in this case, reportedly because no specific gender issues became apparent in the analysis). In most German programmes, there is a clear Fund-related divide: the theme is almost absent among ERDF measures, but more prominent in the aims of ESF policies. In some individual cases, commitments are only implicit, for example in aims such as ‘increasing the well-being of the population’ (Luxembourg), ‘socially responsible development’ (Twente), and the ‘integration of disadvantaged groups’ (Vorarlberg).

205

EQUAL OPPORTUNITIES AIMS IN OBJECTIVE 2 PROGRAMMES

• = To respect relevant legislation or to observe the ‘principle’ of equal opportunities in the actions undertaken, including job creation (Schleswig-Holstein, Poitou-Charentes).

• = To redress specific gender imbalances, including through strengthened support for female entrepreneurs (West Berlin and Finland) and focused support for women in technology growth sectors (West Berlin).

• = To ensure equal access of women and men - including widening employment opportunities for men and women (Bergslagen) and quantifying the job creation outcomes for women as a percentage of overall outcomes (Norra Norrlandskusten and Blekinge).

• = To better reconcile family and working life, including by flexibilisation of work (eg. tele-service centre development in Rheinland-Pfalz).

Among the programmes promoting equal opportunities as a ‘horizontal principle’ (Table 8.3) are Nordjylland, Pays de la Loire, Provence Alpes Côtes d’Azur, Poitou-Charentes, North East England, Industrial South Wales and the Finnish programmes. Such programmes, however, do not fully embrace the ‘gender mainstreaming’ philosophy, or mention the term explicitly. In particular, they do not highlight how most types of project can have possible negative gender outcomes, depending on the details of their design and delivery. Among the programmes which go furthest is West Berlin, whose strategy states that all economic support programmes are examined to see how far women-specific aspects are taken into account, (the Senatsverwaltung für Wirtschaft und Betriebe having initiated a working group ‘Women and Economy’ in 1994 for this purpose). Aquitaine and Pays de la Loire, in addition, state that projects will be required take into account the specificities of the local labour market, including gender differences. Another programme with a useful response is Industrial South Wales, which discusses the issue at overall programme level and under each priority, drawing out the main areas of clear and direct gender relevance (eg. ensuring that the strategic objectives involving new technologies include raising the skills of women and men to compete for new jobs). Provence Alpes Côte d’Azur, too, focuses on specific issues where the Objective 2 programme can make a difference (in this case contributing to improved access to jobs for women, a better match between family and professional life, new care services, transport, flexible and distance working, information on rights and employer awareness). Other programmes state a programme-wide concern to ensure equal access to the measures funded – but without following this up in concrete terms with what this means or how it will be achieved. The most frequent, ‘horizontal’ responses seek more opportunities for positive-action projects to respond to identified gender imbalances. This is a positive first step, and provides a base from which to build a mainstreaming approach.

One risk of expressing equal opportunities as a broad transversal dimension without embedding this in a true mainstreaming approach is that nothing concrete actually takes place to follow this up. Some programmes fail to explain what equal opportunities means as a horizontal issue, beyond respect for relevant legislation and the overall principle, and it is these programmes where a lack of follow-up is most likely. Where the Commission encouraged

206

programmes to replace positive action proposals with a commitment to a ‘horizontal dimension’, this may ironically have reduced the visibility of equal opportunities and programmes’ abilities to have an impact (eg. Nordjylland).

Table 8.3: Examples of the equal opportunities dimension in 1997-99 programme aims and objectives245 Member State Explicit references to equal opportunities in programme aims and objectives Austria Explicit: Oberösterreich - part of the human resources strategic objective is support for

employment and related education and training of employed women to diversify the economic structure. Targets are set for the improved qualification of unskilled women and women employed in key positions. Steiermark - support for equal opportunities is one of six sub-objectives – combined with easier access into the labour market for the unemployed and region-specific problem groups. Implicit: Vorarlberg – integration of disadvantaged groups into the labour market is one of six sub-objectives. Support to women may be implicit in this. Absent: Niederösterreich.

Belgium Absent: Liège, Aubange Denmark Explicit: Nordjylland – the programme lists equal opportunities among the horizontal

development principles which will be pursued (alongside support for economically healthy activities and activities promoting the environment).

Finland Explicit: The strategy section of the national SPD states the broad intention to ‘carry out several projects to promote equality under various measures'. In addition, the encouragement of female entrepreneurs through advisory and training services is mentioned.

France Explicit: Aquitaine – equal opportunities figures as a horizontal strategic element. The specificities of the Aquitaine labour market (including gender differences) will be taken into account in projects. PACA, Pays de la Loire and Poitou-Charentes - a section on how the EC’s horizontal priorities have been taken into account addresses equal opportunities. In Pays de la Loire, it will be a horizontal dimension, with relevant actions under existing measures. In Provence Alpes Côte d’Azur, a dedicated section specifies issues to address: eg. improved access to jobs for women, better matching family and professional life, new care services, transport, flexible and distance working, information on rights and employer awareness. In Poitou-Charentes the theme is not followed through strongly into the programme itself. Explicit in the strategy, but presented separately: Basse-Normandie – equal opportunities aims are presented separately, in a dedicated section. The programme will work horizontally to promote equal opportunities, but will also pursue vertical actions, principally training for women in male-dominated areas, and the orientation of the female unemployed towards more sectors. Absent at strategy level but present in measures: In Bourgogne, Champagne-Ardennes and Languedoc-Roussillon. Absent: Alsace, Bretagne, Centre, and Picardie.

Germany Responses are mixed. Explicit and well-developed: Nordrhein-Westfalen – the section on aims mentions that the Objective 2 programme is in general oriented towards equal opportunities. The theme is highlighted more explicitly in the most relevant measures. Explicit: West Berlin - the SPD states that all economic support programmes are examined to see how far women-specific aspects are taken into account. The Senatsverwaltung für Wirtschaft und Betriebe initiated a working group ‘Women and Economy’ in 1994 for this purpose. Targets of SPD action include: support for women in the framework of regional economic support; strengthened support for women entrepreneurs and focused support in technology growth sectors.

245 1995-99 for Austria and Sweden.

207

Explicit: Rheinland-Pfalz - the strategy section states that the technological development measures will contribute to equal opportunities eg. through tele-service centre development. It also highlights the potential benefits of this priority for the handicapped. Explicit but under-developed: Schleswig Holstein states all measures will observe the principle of equal opportunities for men and women. Explicit as an ESF objective: In Bremen, integration opportunities for men and women is one of the two main ESF objectives. In Saarland, the ESF part of the SPD states that equal opportunities will be a horizontal priority. In Hessen, the importance of vocational qualifications in implementing equal opportunities objectives is mentioned in the strategy. Not explicit: Bayern.

Italy Not explicit: The theme does not appear to be explicit at an overall level in the Italian programmes, which are generally pursuing the same objectives as the 1994-96 round.

Luxembourg Implicit: Increasing the well-being of the population is mentioned, which could be an allusion to equal opportunities.

Netherlands Implicit: The theme is not prominent in most Dutch strategy statements. However, the commitment to ‘socially responsible’ development in Twente could imply the integration of equality issues.

Spain Absent: The theme is not explicit in the strategy statements of the Spanish programmes, except in their commitment to respect relevant legislation in this area.

Sweden Explicit: The equal opportunities theme is highlighted in all Swedish SPDs as a horizontal dimension. A clear practical commitment in most programmes is that the situation of women in the labour market will be taken into consideration at the project selection stage. Bergslagen and Norra Norrlandskusten flag the issue in their overall strategy statements. Bergslagen will widen employment opportunities for men and women, while Norra Norrlandskusten will seek wider equality of opportunity between the sexes. Blekinge and Norra Norrlandskusten also give strategic equal opportunities targets in terms of job creation outcomes for women as a percentage of total outcomes.

United Kingdom

Prominent and explicit: Industrial South Wales and North East England. The theme is explicit in the strategy statement, sub-objectives and priorities of ISW. At strategic level, the global objective of job creation includes equal access for women and men. The new technology priority includes raising the skills of women and men to compete for new jobs. In North East England, a strategic aim is ensuring that all members of the active population, can participate and benefit on an equal basis from measures and actions supported by the programme. Explicit: Eastern Scotland, East Midlands, East London and the Lee Valley and Gibraltar. In Eastern Scotland and the East Midlands, it is a horizontal dimension, with interventions planned in community economic development, entrepreneurship, flexible working and equity in the workplace. In East London and the Lee Valley, equal opportunities is prominent in the strategy – but more focused on ethnic rather than gender opportunity. In Gibraltar, a specific mention is made of the need to develop women’s skills to enable them to participate in non-traditional sectors. Implicit: Plymouth, West Cumbria and Furness, Yorkshire and Humberside and Western Scotland. In Western Scotland, equal opportunities is implicit in the sub-objectives of improving people’s job prospects and quality of life, enhancing economic and social cohesion and combating social exclusion. Rather than setting out a strategy for equal opportunities, the SPD states a focus group will be established to identify the issues and the best way to tackle them. In West Cumbria and Furness and Yorkshire and Humberside, the theme could be implicit in commitments to target economic and social exclusion. The Yorkshire and Humberside SPD states the theme will be treated as a horizontal priority, and highlights specific areas for action including support for the provision of childcare to allow both women and men returning to the labour market to access work and training. Largely absent: West Midlands.

208

The strategies of several (principally UK) programmes concern themselves with multiple types of inequality. These include: East London and the Lee Valley, where ethnic minorities are highlighted; North East England, where all members of the economically active population, whatever their disability, race or gender should benefit equally from measures and actions supported by the programme; and Rheinland-Pfalz where the disabled are highlighted.

8.5.3 Policy Definition: How is Equal Opportunity being promoted? Once highlighted in the strategy section, a key question is whether the theme of gender mainstreaming and equal opportunities is followed through into the programme priorities and measures. In a small number of programmes, the theme is consistently given prominence throughout. Most regions are less consistent, placing more emphasis on equal opportunities in the strategy section or in the measures, or addressing it more seriously with one or the other Fund (eg. in many German programmes the ESF commitment is more visible).

The most frequent explicit policy response is ‘positive action’ projects. Affirmative action usually forms part of wider measures, although four programmes do appear to dedicate the whole or most of a measure to this (two of these are in Austria, both ESF, and neither has actually been implemented as women-only):

��Niederösterreich: skills development for women in problem areas;

��Oberösterreich: support for employment related education and training - a measure referring only to female workers affected by industrial restructuring;

��Sweden - an ERDF measure in Norra Norrlandskusten, dedicated to encouraging entrepreneurship among women and young people; and

��Germany - an ERDF measure in Nordrhein-Westfalen, Regional Agencies for Women and Work, which aims to increase employment opportunities for women and support the creation of female-run enterprises. This measure is frequently cited as an example of best practice, and has provided assistance for the creation of regional centres dedicated to enhancing women’s position in the labour market (eg. by facilitating their re-integration into work after breaks or encouraging access to sectors which have traditionally been male dominated).246

The most frequent types of affirmative action invited, almost all of which target women explicitly, are listed below, by Fund.

246 CEC (1997) The new Regional Programmes 1997-1999 under Objective 2 of the Community’s Structural Policies - focusing on Job-Creation, DG XVI of the European Commission, Communication COM(97)524 final, 14.11.97, Brussels. For further detail see Fitzgerald R and Noble P (1997) Integrating Equal Opportunities into Objective 2 Programmes, IQ-Net Thematic Paper, 3(2), EPRC, University of Strathclyde, Glasgow.

209

POSITIVE ACTION UNDER THE ESF

• = Training to facilitate women’s access to sectors in which they are under-represented (eg. Nordjylland, Oberösterreich, Centre, Picardie, Nordrhein-Westfalen, Plymouth). This is sometimes explicitly targeted at high-technology, technical and highly skilled sectors (eg. Aquitaine, Picardie, Poitou-Charentes).

• = Facilitating access to employment-related training for those with caring responsibilities, eg. by the introduction of innovative training delivery systems (Saarland)

• = Ensuring parity among those employed in mutating sectors – providing additional training so that those who have taken career breaks are not in a more precarious position than their colleagues should they become unemployed (Saarland).

POSITIVE ACTION UNDER THE ERDF

ERDF interventions, likewise, frequently involve improving individuals’ access to training or work. Examples here are:

• = Initiatives to overcome the barriers preventing people joining training schemes or entering the labour market, including their caring responsibilities, lack of personal mobility and the impact of career breaks. Interventions typically include:

- introducing or paying for access to childcare and elderly daycare facilities (very widespread, including Aragon, Cataluña, Bourgogne, Languedoc-Roussillon, Poitou-Charentes, Schleswig-Holstein, East and West Midlands, Industrial South Wales),

- transport solutions (eg. Alsace, Thanet, Western Scotland), and

- advice services (eg. Rheinland-Pfalz).

• = Women’s entrepreneurship initiatives, encouraging higher start-up and survival rates among female-run businesses through consultancy or investment (again, widespread, including in the Finnish programmes, Picardie, Bayern, Hessen, Nordrhein Westfalen, Norra Norrlandskusten, East London and the Lee Valley, Thanet and West Cumbria and Furness).

• = Business infrastructure encouraging female employment (eg. incubator facilities with shared services in Alsace, a model start-up centre in Schleswig Holstein and childcare facilities linked to businesses in the West Midlands).

The above range of support is relatively wide, the most frequent measures involving actions to improve access to training and employment among people with caring responsibilities, and encouraging more female entrepreneurs. The types of positive action are in part determined by national contexts. For example, childcare provision is recognised as a major constraint to access to work and employment in the UK, while in Finland, there is no such issue, municipalities being required by law to provide sufficient provision for families.

A small number of actions go beyond targeting women. They address ‘change agents’ such as employers and public agencies, aiming to encourage a better understanding of gender patterns and to influence working practices and norms for the benefit of equality of opportunity. The main examples are:

210

��Planning exercises encouraging greater sensitivity to gender issues, including by defining training needs or identifying the impact of actions undertaken (eg. Nordjylland, Aquitaine, Champagne-Ardennes, Provence Alpes Côte d’Azur, Nordrhein-Westfalen, West Berlin).

��Equality interventions in firms, including the encouragement of flexible working within companies and the exploitation of this to equalise opportunities (Poitou-Charentes, Rheinland-Pfalz, Plymouth), and funding for internal equality strategies in firms (Industrial South Wales).

��Recognition of the contribution of spouses in family firms (Aquitaine, Languedoc-Roussillon).

Among measures promoting equal opportunities, some do so in a way which is tenuous, unclear or tokenistic - in similar vein to some commitments to the environment. The Bayern programme proposes that its economic infrastructure measure (1.1) will respect the principles of equal opportunities, but fails to explain further. Gibraltar’s ESF measure likewise makes a broad claim that some actions will actively promote equal opportunities, but without further clarification. In the next round, it is anticipated that experience to date will enable improved measure-level content, and that the ex ante evaluation and subsequent negotiation with the EC will be used to ensure that this element is improved.

Mainstreaming equal opportunities at the measure level implies gender vigilance across every project in every policy area, recognising that the details of their formulation can affect the extent to which they create opportunities open to people of both sexes (see examples below). A key observation is that fewer projects and policies are gender neutral than might be supposed. At the priority and measure levels, a minority of current programmes show the beginnings of a mainstreaming approach, highlighting the potential gender impact of some actions, as well as specifying existing imbalances which measures could recognise and/or address. Examples include: West Berlin, where interventions are routinely checked for this women-specific dimension; Aquitaine, where certain projects are assessed in relation to the specific characteristics of the labour market including gender differences; and Western Scotland, where explicit consideration of potential gender impacts of projects is among the project selection criteria.

THE ‘HIDDEN’ GENDER IMPLICATIONS OF PROJECTS

Business development: Minimum thresholds for the size of business development grants or loans can unintentionally exclude firms run by women, which on average make smaller applications for support.

Economic infrastructure: Building business premises which are inaccessible or fail to build in precautions for personal safety (eg. with poor access lighting, isolated car parking and no access by public transport) may exclude female entrepreneurs or potential employees from choosing to work from them.

Training: Courses which build in timetable flexibility (eg. with modules for part-time study) or which include some self-teaching modules are much more likely to be accessible to those who have to organise their participation around caring responsibilities.

211

There is a long way to go before the gender dimension is a systematic element of every measure sheet. Policies and projects need to be checked to ascertain whether they have potential unequal implications for men and women, and if so, gender impact assessments (GIA) undertaken at the ex ante stage to identify and mitigate them. Methods and instruments of gender impact assessment are under active development to be applied across all policy fields, to help policy professionals identify potential gender issues, and either to address them or involve gender experts in doing so. Some of the emerging methods, which are still largely in their pilot stages, are described in the Council of Europe mainstreaming study, with a view to disseminating them more widely. Among the tools in systematic use are those of international development programmes, where mainstreaming has been promoted particularly strongly (see, for example, the work of UNDP, the World Bank and the International Labour Organisation). EIA methods can only be operated effectively where appropriate sex-disaggregated statistics are obtainable and where those applying the systems have at least a clear basic understanding of the field. However, they have real potential not only to ensure that gender issues become a standard and integral element of the mainstream policy process but also to impact on gender outcomes.

GENDER IMPACT ASSESSMENT METHODS

CHECKING FOR GENDER RELEVANCE – ‘SMART’ DEVELOPED FOR THE EUROPEAN COMMISSION

To deliver gender mainstreaming, policy personnel who are non-specialists in gender issues need clear and straightforward methodologies to help them to identify whether proposed policies and projects have gender implications and thus whether further stages of investigation and then modification are required. The EC uses an adapted version of the first stages of the Dutch Gender Impact Assessment described below, named SMART (Simple Method to Assess the Relevance of Policies to Gender) 247 to do this.

The first stage of the method is to identify whether a proposal concerns target groups. If it does – and most do – then this implies an uneven distribution of resources among the population, often determined on the basis of the application of rules or regulations. If the proposal does concern target groups, then the second stage is to check for gender relevance, by determining whether there are differences between women and men in the field of the policy proposal. In answering this question, four key areas are checked: differences in rights, participation, resources and norms and values. For example:

• = Participation: Is there a different share of men and women in the target group and among decision-makers?

• = Resources: What differences are there between the resources available to members of the target group in terms of time, income and opportunities, and what are the implications?

• = Norms and values: Could ‘gendered choices’ affect the field in question (eg. their likely education and career choices, choices relating to the division of domestic labour)

• = Rights: What are the different rights of full-time and part-time work, and how do they impact on men and women, as more women work part-time? This may have implications for the proposals.

If a policy is shown to be gender relevant, then a gender impact assessment is the next stage.

247 This method is reported briefly in Council of Europe (1998) op. cit., and described in more detail in: Commission of the European Communities, A Guide to Gender Impact Assessment, available at: http://europa.eu.int/comm/dg05/key_en.htm.

212

TOOLS FOR GENDER IMPACT ASSESSMENT – GIA IN THE NETHERLANDS AND FLANDERS248 Among the best known instruments for GIA are those developed in the Netherlands and

Belgium. The Dutch GIA tool was commissioned by their Equality Division in 1994 for use at national level, and has since been supplemented by related but simpler tools for use at the regional and local levels. As described above, the national GIA system first assesses the gender relevance of a policy or project proposal, then goes on to analyse this in more detail. To do this, the likely impact on the current and future situation in the policy field is identified. This enables potential positive or negative impacts to be identified and corrective adaptations to be made if necessary.

The Flemish GIA tool builds on the experience of the Dutch one, and consists of three stages:

• = identification of the gender dimension of proposals, which involves assessing whether and how gender is taken into account;

• = estimation of the nature and magnitude of possible impacts, assessing whether the policy proposal maintains a discriminating perception of women or men and whether it has discriminatory results for their daily life; and

• = modification of proposals where appropriate, preventing negative impacts and, where feasible or appropriate, promoting positive ones.

The first stage can be undertaken by generalists, but the second and third stages benefit from specialist input. The tool has not yet been extensively used, but has potential not only to lead to positive gender outcomes but also to raise awareness among the ‘generalist’ policy community.

8.6 Programme Implementation

8.6.1 Programme Management Encouraging a critical awareness of the gender balance of programme committee and decision-making structures and of programme applicants and beneficiaries would be positive, and would be expected to lead naturally in time to more balanced structures. However, imposing gender quotas on committees risks being counterproductive.

First, and most importantly, it may encourage the misleading assumption that more ‘gender-aware’ decisions will automatically be generated by involving more women in decision-making. Rather than enforcing quotas, it is more important to ensure that gender issues are a constant part of the programming agenda and that they are addressed consistently and maturely. Gender-aware decisions can be delivered by both females and males provided that they are appraised of the issues and equipped with the necessary tools.

Second, in purely practical terms, a quota system could be difficult to implement as the composition of committees derives to a large extent from the structure of participating organisations and, in some cases, from the choices of the regional or local electorate. Quotas, if they were regarded as artificial or insensitive to regional contexts, might simply lead to false gender accounting, with, for example, gender-balanced lists of ‘official’ committee members co-existing alongside a less balanced composition of individuals actually

248 Council of Europe (1998) op. cit.

213

attending meetings. This would not lead to gender issues being embraced more fully, but instead to an institutionalised tokenism.

8.6.2 Project Generation, Appraisal and Selection The aim of equal opportunities will ultimately be realised through projects, emphasising the importance of project design and assessment procedures. Where the SPD is a primary reference document for those developing projects, consistent, clear and practical information through each tier of the document can help to influence the form and content of projects. There is already evidence that explicit reference to the equal opportunities dimension in the 1997-99 SPDs has, in some cases, sparked debate, and triggered project ideas – for positive action in the first instance - which might not otherwise have come forward.

As noted above, relatively few SPDs include the sort of consistent and clear information required in the diagnostic section, the aims and objectives, and in priority and measure descriptions, including, at a practical level, selection criteria and programme targets. Improvements are required in the consistency, visibility and lisibility of the theme in most programming documents, including:

��a diagnosis which interprets the patterns identified;

��definitions of key terms – including mainstreaming - and clear overall objectives;

��appropriate ideas for positive action; and

��an explanation of how gender should be taken into account in general projects (ie. ones which may have unequal gender impacts, but are not women-specific).

The way in which SPDs are used as reference documents reinforces the necessity for the key points regarding equal opportunities to be included in all relevant sections of the documents. In particular, programmes could use the programme priority and measure selection criteria to state that projects will be favoured which have systematically addressed the issues – preferably by using an agreed gender-impact assessment method. Few programmes currently do this, as Table 8.4, setting out the gender-related selection criteria contained in 1997-99 Objective 2 SPDs, illustrates. At the programme level, the most frequent statement made – but rarely explained - is that ‘all projects should respect equal opportunities principles’. Some programmes choose to use overall selection criteria for positive reinforcement, stating that programmes are open to both men and women in the target groups. A small number use overall selection principles with more force. Western Scotland states that consideration of gender issues will, in some cases, be a condition of eligibility, for example with a requirement that projects provide access to public transport and dependent care where appropriate. In fact, the application of a gender filter could be a condition of eligibility in every programme and for every project. Requiring the issue to be addressed specifically could bring wider benefits: experience in Germany suggests that the more detail requested on the

214

scale and nature of women’s participation, ‘the more targeted are the measures developed and implemented’.249

The table shows that, even at measure level, the use of gender-specific selection criteria under ESF and ERDF measures in current programmes is extremely limited and unsophisticated, largely comprising prioritisation for projects which target women specifically (eg. for entrepreneurship), which have ‘an equal opportunities dimension’, or which aim to widen access to training or employment. The situation on the ground, however, is better than that suggested in the table, because not all programmes set out their selection criteria comprehensively in their SPDs, and gender-related selection criteria have, sometimes, been developed subsequent to programme launch.

Table 8.4: Gender-related selection criteria in 1997-99 Objective 2 programmes

Gender-relevant selection criteria need to be employed more systematically and more specifically, giving applicants a clearer indication of what is being sought. In addition, they could be formulated more powerfully to influence project outcomes, either by offering higher intervention rates to projects which incorporate mainstreaming responses250 or by imposing conditions on projects. Higher intervention rates have already been used in several German programmes where employers have prioritised jobs and/or training places for women. Conditions are another means of influencing projects. In Nordrhein-Westfalen and Schleswig-Holstein, projects supported by the regulations of the Labour Office have to ensure that women participate in them at least to the same level as their percentage among the target group of unemployed.251 To succeed more widely, such approaches require clear guidance to potential applicants, and then consistent tools to assess the equal opportunities content

249 Braithwaite M, Fries R and Fitzgerald R (1998) op. cit., p 5. 250 As suggested in Wulf-Mathies, M (1998) Information Note from Mrs Wulf-Mathies to the Members of the Group of Commissioners on Equal Opportunities for Men and Women and Women’s Rights, DG XVI, Brussels, 24.2.98. 251 Braithwaite M, Fries R and Fitzgerald R (1998) op. cit., p13.

Measure type Gender-related selection criteria Overall principles All projects will respect equal opportunities principles

Projects will be preferred which enhance equal opportunities The programme is open to anyone in the target groups Projects are required to provide access to public transport or childcare provision where relevant

Skills development (ESF) Projects with an equal opportunities dimension will be prioritised (including in-firm training in some cases) Training projects must provide for appropriate provision of dependent care

Labour market studies (ESF) No specific selection criteria Entrepreneurship (ERDF) Support to female entrepreneurs Business development (ERDF) Integration of women into highly qualified positions

Indication of target groups Studies and planning (ERDF) No specific selection critera Care services (ERDF) Projects benefiting women will be favoured

215

of the resulting applications. To maintain the credibility of higher awards, it would also be necessary to significantly improve the monitoring of gender outcomes, and to ensure claw-back for non-delivery. Projects involving ‘positive action’ are easier to assess for their equal opportunities impacts than other projects since these are core objectives. The challenge for other projects could be considerable.

Other channels are also used to generate appropriate project responses. In Auvergne, a technical note was issued for project promoters, encouraging them to integrate equal opportunities into their projects by addressing factors which can constrain or facilitate equal opportunities for men and women.252 In Industrial South Wales, project applicants complete an Equal Opportunities Questionnaire (with the help, if requested, of the European Equality Partnership). In the Swedish context, to help ensure that gender issues are taken into account in all projects, Lindsten253 proposes the ‘gender reflex’ – a set of questions to be used systematically by project proposers and assessors to ensure the gender-awareness of their proposals. All programmes could introduce this type of gender filter, perhaps with questions differentiated by project type.254 Following this up, effective and consistent frameworks are needed at the appraisal and decision-making stages.

8.7 Monitoring and Evaluation To ensure that more systematic account is taken of equal opportunities, programming authorities are being required to establish baseline data on gender, set gender-disaggregated targets and monitor them on an ongoing basis – and to inform implementers of their responsibilities in this regard. Objective 2 measures currently contain few gender-differentiated targets and indicators (Table 8.5). The few which are in place are unsophisticated and rarely quantified. Even where there is quantification (which is present more consistently for the ESF), the utility of these figures is limited by the failure to set out current rates of participation by gender as baseline comparators. Targets tend to refer to the percentage of women to benefit from direct interventions, such as training or employment subsidies, but this is almost meaningless without statistics on the typical gender breakdown of such interventions to date. In the absence of information on which to base their decisions, some programmes have simply set gender targets of fifty percent of beneficiaries being male and fifty percent female. These send out a message in the first instance, but should be replaced with more meaningful targets as baseline information becomes available.

Many of the monitoring problems experienced in the current round relate to the absence of sufficient early work to establish indicators and systems – because of limited time or resources, methodological difficulties or the placing of insufficient emphasis on this task. The poor content of current SPDs, and survey evidence from France, Germany and the UK,255 suggest that further

252 Braithwaite M, Fries R and Fitzgerald R (1998) op. cit., p12. 253 Lindsten S (1997) The REFLEX – Method for analysis of equal opportunities between men and women, Goteburg and Bohus County Administration Board, Sweden. 254 In a related field, Chapter 7 in this volume addresses the checklists which could be applied to assess the environmental implications of projects. 255 Braithwaite M, Fries R and Fitzgerald R (1998) op. cit.

216

guidance is required by many programmes on appropriate indicators and their good use, including the importance of establishing baseline situations.

Table 8.5: Summary table: gender-related targets and indicators in 1997-99 programmes Measure type Gender-related targets and indicators – project level Overall principles Percentage of all jobs created which will benefit women Skills development (ESF) Number of women participants

Percentage of training participants/researchers trained who are female Percentage of women involved in science and technology projects Career prospects of the female beneficiaries Gender breakdown of those benefiting from recruitment subsidies or skills analyses Percentage of jobs created which go to women

Labour market studies (ESF) Number of studies undertaken Entrepreneurship (ERDF) Number of businesses started by women

Number of seminars and events held Percentage increase in women taking up self-employment

Business development (ERDF)

Number of interventions favouring the recognition of the contribution of spouses to family firms. Number of consultancy projects pursuing equal opportunities issues or benefiting businesswomen

Studies and planning (ERDF) Number of studies into equal opportunities in economic development Care services (ERDF) Number of child or elderly care facilities created

Number of grants to improve or create child or elderly care facilities Number of care places created for children and the elderly Increase in the employment of ethnic minorities and women resulting from training and awareness-raising

Once systems have been established, they benefit from ongoing improvement, and several programmes have established structures to facilitate this. Berlin has an expert working group on ‘Women and Economy’, while in Wales, a European Equality Partnership has been put in place, managed by Chwarae Teg (Fair Play).

Following project approval, applicants’ obligations vis à vis gender monitoring can usefully be re-iterated. The offer letter is used to reinforce the message in selected cases (eg. the West Midlands). In collecting monitoring data, quantitative information can usefully be supplemented by a qualitative dimension. Valuable additional insights would emerge from asking projects to report on how they took equal opportunities issues into account and what effect this had.

The widespread systematic assessment of the gender-related impacts of Objective 2 programmes (across the whole portfolio of projects) has not yet taken place. The gender outcomes of the 1989-93 Objective 2 programmes, for example, are not known, as the ex post evaluations carried out in 1997 did

217

not address the issue.256 It would have been unfair and indeed impractical to assess this retrospectively when it was not a formal dimension of those programmes. However, the Council resolution of 1996 on mainstreaming has strengthened the requirement, stating that evaluations should, in future, ‘measure the extent to which the principle of promoting equal opportunities has been taken into account, with particular regard to the involvement of women in general measures and to the implementation, the relevance and the outcome of such measures’,257 subsequently making changes as required.

Following the resolution, the quantity of evaluation evidence on gender issues and outcomes has increased rapidly with the recent interim evaluations and current ex ante evaluations of the draft new programmes. However, the quality of this evidence may be limited as it is dependent on the skills of evaluators (few of whom have specific expertise in gender issues), the abilities of those managing the evaluations and the degree of emphasis placed on this issue in multi-faceted studies. Provided that the emphasis on this issue is sustained, the quality of information available should increase steadily in future rounds of evaluation. This also requires further work to be undertaken to refine methodologies for effective gender impact evaluation – including developing a consistent and clear terminology.

Where they have not already done so, a valuable role of the interim and/or ex ante evaluations for programmes will be to establish clearer baselines, both in terms of gender imbalances in the eligible areas and their causes, and the gender participation and outcomes of specific interventions to date. It is against these that future performance can be measured.

The speed and quantity of information being generated about the gender impacts of mixed Structural Fund programmes, and the need for wide circulation to improve practice elsewhere, provide an excellent case for better exploiting flexible, rapid and low cost means of dissemination – including the Internet and electronic mail.

8.8 Conclusions and Recommendations Women and men in the EU, in assisted areas and beyond, persistently experience unequal access to and benefit from training and employment, both in terms of quantity and quality. In economic terms, this constitutes the sub-optimal utilisation of human resources – an issue of particular importance in economically disadvantaged regions. In order to release the potential of every individual as a ‘vital development agent’, policies – including those for economic development – need to be adapted to enable all groups to gain equal access to productive resources, opportunities and public services. The new policy approach of ‘gender mainstreaming’ provides a powerful tool to systematically screen all policies for their gender implications and impacts at the point of planning, implementation and evaluation, in order to ensure that opportunities are genuinely available to the spectrum of potential participants, and that initiatives have no negative gender impacts.

256 CEC (1997) Ex Post Evaluation of the 1989-93 Objective 2 Programmes – Synthesis Report, Evaluation and documents, No 4, September 1997, DG XVI of the European Commission. 257 CEC (1996) Resolution on mainstreaming equal opportunities of men and women into the European Structural Funds, Official Journal C386 2.12.1996, Article 2a.

218

Recent changes in the wider policy environment, with political will being secured for gender mainstreaming at the highest levels of the EU, in tandem with the inclusion of an equal opportunities obligation in the Structural Funds regulations, have provided important stimuli for equal opportunities to be taken into account by Objective 2 programmes. From being an area of legislation to respect, equal opportunities has newly become, in theory at least, a universal, active dimension of Structural Fund programmes, pursued with the dual rationales of equity and efficiency.

Although in insufficient numbers and with insufficient depth, many current Objective 2 SPDs (1997-99 and, in Autria and Sweden, 1995–99) now contain some, limited, references to gender imbalances in the economic sphere and their potential to address them. This change is important, since the presence of even superficial references to equal opportunities in programming documents has helped to place the issue on the agenda, adding weight to it and opening the possibility for ongoing debate and development at the programme level.

In spite of recent progress, there is further to go: the equal opportunities demands made of Objective 2 programmes are considerable. This theme is expected to be integrated comprehensively as an element of the design, delivery and assessment of all policies, through ‘gender mainstreaming’, and, within this, addressed where relevant by positive action, ie. projects designed by specialist agencies to help remove the barriers to (usually) women’s full participation in the labour market. Both types of response to gender issues are new to most Objective 2 programmes, but gender mainstreaming is also new in the wider policy environment, meaning there are few precedents and little experience. It is not surprising that most responses to date have focused on positive action (frequently, entrepreneurship schemes targeting women, and provision to ensure access to training and employment by those with caring responsibilities).

The lack of targeted evaluation of the equal opportunities dimension of Objective 2 programmes means that little can be said definitively about the extent and impact of gender mainstreaming so far. However, a review of current programmes illustrates that gender mainstreaming has not yet been fully embraced by any programme. Establishing the ‘gender reflex’, in which every action is routinely screened for its gender implications, will have to be a long-term objective. The low quality of many programme responses to date does not necessarily relate to a lack of commitment to the principle of equal opportunities, which is genuine in many cases, but rather to the absence of appropriate knowledge, skills, tools and mechanisms to enable more effective responses.

To move towards true mainstreaming, in which a gender filter would be applied to all policies and projects, assessing their likely uneven gender impacts and making appropriate adjustments, a wider awareness of what this term implies is needed among programme actors. Progress will require the Commission to sustain its emphasis in the next round – including through the provision of improved guidance which goes beyond individual examples of positive action to mainstreaming proper. Programme actors will also have to intensify their efforts. It is notable that those programmes which have made

219

most progress in equal opportunities to date are those which have embraced the issue most actively as an integral part of programming.

8.8.1 Gender Mainstreaming Recommendations: Principles

Gender mainstreaming is new. To secure more of the desired responses earlier, it is recommended that quicker, wider and more effective dissemination of existing best practice and guidance is undertaken. This should prioritise practical information on gender mainstreaming itself rather than positive action as has been the case to date.

Various initiatives have been undertaken by DG XVI and DG V to generate and disseminate best practice on equal opportunities, including studies and conferences,258 but feedback suggests that their results have been under-exploited. Regional and local actors still perceive a lack of practical guidance from EC and national levels on equal opportunities generally, but especially on mainstreaming.

Quicker, wider and more effective dissemination must be a Commission (and Member State) priority to ensure that existing guidance is circulated widely enough to reach those operationalising the issues, who by definition in ‘mainstreaming’, are not equal opportunities specialists and who have limited resources with which to respond to the issue. DG V has used its Internet resource to disseminate gender mainstreaming documentation, but the mainstreaming content of the Inforegio site (DG XVI) is extremely limited.

The main emphasis of best practice information to date has been on positive action. This continues in the draft guidelines for the standard EU programmes in the 2000-06 period. However, if the more abstract issue of mainstreaming is to become integrated into programmes, timely practical advice and best practice examples are urgent and necessary. The instruction that ‘an overall mainstreaming approach for equal opportunities must be introduced into all Structural Funds programming’259 will not be translated into practice without further elaboration.

In general, the ESF community has greater equal opportunities awareness and experience than the ERDF community. To rectify this, there is a need for the more co-ordinated dissemination of relevant information to a wider audience – via their own established channels. Internet resources provide a powerful method of disseminating guidance and ideas in a new policy area. Greater connectivity is therefore recommended not only between the DG XVI and V websites, but also with other information sources of genuine value to Structural Fund actors addressing this field, for example, sources generated by the UN, the OECD, and the Council of Europe.260 A directory of recommended information sources and tools could be devised and updated on

258 Conferences include: Mainstreaming of Equality between Women and Men in the Structural Funds, Portugal, 13-15 September 1998. 259 CEC (1999) The Structural Funds and their Co-ordination with the Cohesion Fund: Draft guidance for programmes in the period 2000-06, Working paper of the Commission, Brussels. 260 Council of Europe (1998) op. cit.

220

a rolling basis to assist programmes in the first steps of establishing their own responses.

Programming authorities should ensure that equal opportunities is embraced actively as an integral part of programme planning and implementation, and that responses are shaped to contribute to core programme objectives.

Several steps can be taken to ensure this:

��Seek to establish commitment to gender mainstreaming among the whole partnership and the programme’s secretariat structures. This can be facilitated through visible commitment to the issue at the highest levels of the programme partnership, and by setting out the economic case for gender sensitivity.

��Follow the issue through, by allocating overall responsibility to an individual, preferably working through and with a dedicated working group of partners, some but not all of whom have specialist gender expertise.

��Improve the quality and efficiency of responses by actively involving equal opportunities specialists on an ongoing basis, including commissioning research where appropriate.

��Undertake an active programme of knowledge and skills development among the wider partnership, including encouragement for the experts associated with the programme to actively disseminate core skills (through presentations at committee meetings, programme newsletters, themed ‘action days’). Positive action requires relatively few people to be aware of gender issues, but mainstreaming requires much wider awareness of gender imbalances and their causes, and skills in assessing the potential gender differentiated impacts of policies and projects. Currently, relevant knowledge and skills are frequently too concentrated, sometimes in a single individual working with the programme on an ad hoc basis and not necessarily entirely familiar with Structural Fund programming.

��Work systematically to integrate gender mainstreaming as an element of each stage in the programming cycle, through programme development, project assessment and selection and monitoring, establishing channels for issue monitoring and feedback loops. For clarity and transparency, the arrangements made to take the issue into account should be set out in the programming document or supplement, stating who has been involved, what follow-up arrangements have been established and what outcomes are expected. To ensure clear thinking and structured follow-up, a checklist which sets out the tasks and issues at every programming stage could be elaborated by an issue group and approved and followed up by the Monitoring Committee. A parallel exercise could be undertaken for environmental sustainability.

221

A greater gender balance should actively be sought in decision-making structures, but ideally not through the imposition of male/female quotas.

Decision-makers on Structural Fund committees should represent as wide a spectrum as possible of the beneficiary community. This includes ensuring a significant representation of women. It is recommended that this be achieved not by imposing male-female quota systems, but instead by actively monitoring the gender balance of committees and providing explanations of the patterns which emerge. This should help to build awareness which in turn should lead to a change in the balance of participation.

It will be important to equip all decision-makers, whether male or female, with the knowledge and skills to identify - and respond to - gender issues.

More ‘gender-aware’ decisions can, but do not necessarily, derive from the gender balance of the decision-makers. Such decisions can be delivered by females or males provided that they are appraised of the issues and equipped with the necessary tools. As such, all members of committees and supporting administrative staff – male and female – should receive appropriate briefings and training, tailored to an economic development context.

One of the barriers to implementing gender mainstreaming is a lack of clarity about what this means in Structural Fund programming. Defining the anticipated scope of equal opportunities responses will help to clarify expectations.

There is a lack of clarity at all tiers of Structural Fund governance about what the optimum scope of equal opportunities action should be within Objective 2 programmes. Specifically, there has been some debate about whether mainstreaming responses should, in future, take the place of positive action. In fact, these are complementary policy responses – positive action being a sub-set of mainstreaming activity - and both should co-exist in future. Currently, positive action projects are not only addressing real issues of inequality (which in many cases have negative economic consequences), but are also helping to prepare the way for fuller ‘mainstreaming’, by familiarising programme participants with gender issues. The following steps would help to build on progress to date.

��Defining mainstreaming: Where programmes have reacted to equal opportunities as a ‘horizontal issue of programming’, this has usually been interpreted as identifying the possibilities for positive action under different measure headings. The true scope of gender mainstreaming is broader, extending to the application of a gender filter to all policies and projects, assessing their likely uneven gender impacts and making appropriate adjustments. To move towards true mainstreaming, a wider awareness of what this term implies is needed among programme

222

partnerships. Also, to realise the concept, tools, procedures and methodologies are needed which can be applied on a day-to-day basis by non-specialist economic development professionals following limited training or familiarisation. It is recommended that programmes experiment with some of the ‘gender impact assessment’ methodologies which are emerging, adapting them to their own requirements. They could build up progressively, addressing first those policy areas where potential uneven gender impacts are greatest or most likely.

��Widening the scope of positive action: The range of types of positive action which programmes specifically invite is currently narrow, with a predominant emphasis on entrepreneurship schemes for women and removing barriers to training and employment by increasing childcare and daycare facilities. Depending on the patterns of gender imbalance encountered, more programmes could highlight a wider range of possible positive action projects. The examples given above, SPDs themselves and projects highlighted in best practice guides provide a menu of possibilities. There has been longer experience of pursuing equal opportunities outcomes through ESF-funded projects, and so it is under the ERDF that the most original thinking and the widest dissemination of new ideas is required.

��Ensuring positive action projects respond to real needs: In identifying ideas for appropriate positive action projects, it is imperative to understand the root causes of gender imbalances. The assumed cause-and-effect relationships may be misleading. For example, while a principal childcarer may have limited access to work or training because of a lack of pre-school childcare, this may also be the result of a shortage of after-school care services for school-age children. In addition, it is important to identify the full range of gender imbalances which could be addressed through positive action. Equal opportunities is not always a ‘women’s issue’. Male patterns of (un)employment may be less favourable than those for women in some cases, requiring their own positive action. More thorough diagnosis would highlight such patterns, including, for example, higher male long-term unemployment among pre-retirement and young people in regional economies formerly dependent on heavy industries.

��Broadening the scope. There is some debate about the categories of individuals which ‘equal opportunities’ should take into account, with some programmes and Member States regarding the focus on male/female equality as too narrow. In some cases, such as Sweden, programmes already reflect a wider view of equal opportunities which takes into account the disadvantage faced in the labour market by a range of groups, including ethnic minorities and the disabled. For example, in Ångermanlandskusten, the feasibility of widening the definition of equal opportunities to groups such as the handicapped is being investigated; Fyrstad also highlights immigrants and young people as groups whose economic integration may require dedicated actions. East London and the Lee Valley further addresses ethnic minorities. The proposed ‘rule of thumb’ is that identified patterns of uneven opportunity – especially where they have economic and social consequences - merit action.

223

8.8.2 Gender Mainstreaming: Programme Development Recommendations

Make gender mainstreaming meaningful by undertaking a thorough diagnosis to identify concrete gender-related economic development issues and then addressing them.

A narrow view of the scope of equal opportunities action persists in many programmes, in part because mainstreaming remains too abstract but also because there is an insufficiently deep understanding of the nature and causes of gender imbalances and of the impact Structural Fund programmes have on these. The most appropriate and effective project ideas can only flow from a detailed diagnosis. As such, addressing the equal opportunities dimension needs to be an explicit requirement of the regional profile and SWOT analysis, with responses describing patterns in sufficient detail to provide a real insight into differentiation and the reasons for it (looking at the quantity and quality of employment and at features of unemployment). The diagnostic phase should capture relevant information and ideas from specialist organisations and others (including social partners and voluntary groups). It should exploit existing information (for example, by involving individuals and organisations with special expertise in an active and timely way), but it may also need to generate new insights where these are lacking, including through dedicated studies. Some programmes have made a commitment to research the issues on an ongoing basis or to establish instruments to generate relevant statistics. This would appear a positive way to instil a longer term commitment and to achieve incremental results.

Where deficiencies are encountered in available statistics, three steps are recommended to programme managers: (a) describe the data limitations faced; (b) highlight known issues or use ‘next best’ statistics – eg. for the region as a whole; and (c) allocate funding for practical solutions to resolve the deficits in the medium term. The current PACA programme, for example, includes scope for new statistical tools to help monitor the equal opportunities outcomes of training. In Western Scotland, rather than setting out a strategy for equal opportunities, the SPD makes a commitment to the forming of a focus group to identify issues more carefully.

To respond more efficiently to the long-term statistical requirements of mainstreaming equal opportunities in economic development, studies should be commissioned at national level to assess the current availability of gender-disaggregated statistics, and to introduce reforms to resolve the main deficiencies. This would benefit not only Structural Fund programming, but the wider introduction of gender mainstreaming across all policy fields, which is being actively pursued from the highest political levels in a number of Member States.261

261 Council of Europe (1998) op. cit.

224

Programmes should be obliged to include gender-disaggregated indicators and targets. In addition, before programmes are finalised, ‘gender filters’ should be applied to policies to identify and address any potentially negative and unnecessary gender implications.

The mainstreaming philosophy applies not only to projects but also to policies. The policy design phase needs to build in time to assess the potential gender outcomes of proposed measures – possibly as part of the ex ante exercise, proposing modifications to mitigate possible negative outcomes. Tools will be needed to achieve this, but they are only in the early stages of development across all policy fields.

Use the programming document to highlight gender issues and their relevance, and present the theme consistently and in an integrated fashion. To assist the assessment of programming documents, a description should be included of the processes by which the equal opportunities elements of programmes were derived and are to be pursued.

Without a body of systematic impact evaluation, it is impossible to say where it is most important to place the emphasis in programming documents. However, it would seem broadly advisable to make the equal opportunities dimension of an SPD more visible throughout, following through from principles and objectives to expected practice. Several recommendations can be distilled from good practice to date.

��For the benefit of partners and applicants, programmes should explain the scope of their equal opportunities role, clearly setting out the difference between positive action and gender mainstreaming and the importance of both.

��Gender-disaggregated statistics should not only be presented, but also interpreted to ensure that their implications are clear.

��Inclusive language is a simple and direct way to emphasise their openness to applications from people of both sexes – ie. referring to ‘men and women’ in describing the scope of measures, (eg. Poitou-Charentes and Nordrhein-Westfalen). What should certainly be avoided is mono-gender vocabulary to describe a measure available to both men and women (eg. businesspeople should be preferred to businessmen).

��In the programme supplement, the following should be considered for measure descriptions.

−= In the same way as measure sheets increasingly include a dedicated paragraph on potential environmental impacts, they should also set out potential positive and negative impacts on gender patterns. This imposes a stage of reflection for each measure which may generate further ideas, and would make the issue more concrete for programme actors, guiding them in what to look for or to avoid in projects and illustrating how negative impacts could be mitigated. Such analysis demonstrates that few measures are in reality ‘gender neutral’.

225

−= Where relevant, explicitly highlight a wider range of possible positive-action project types in measure sheets to help bring forward projects for which there is demand but which may previously not have sought finance under Structural Fund programmes. A review of other SPDs, good practice studies or the sections above, can help to highlight examples.

−= Project selection criteria could be exploited more systematically as a means of influencing project design and assessment. Among the selection criteria could be a specification that applicants should have explicitly considered the potential of their project to take forward (and certainly not to regress) equal opportunities issues (eg. Western Scotland).

−= Consider offering variable intervention rates, with higher rates for projects having a positive impact on gender outcomes.

8.8.3 Gender Mainstreaming Recommendations: Programme Implementation

A conscious, clear and planned communication effort is required to ensure that future programme applicants understand the meaning, importance and relevance of the concept of gender mainstreaming and their obligations in this field.

There is an urgent need to raise applicants’ awareness of mainstreaming, given its novelty, and its early stage of development in most policy fields (although some Member States are further advanced than others in introducing this concept into policy development and project selection). A range of channels can be used, the first being official programming materials: the programming document, the programming supplement, application forms and award letters/contracts. Other active channels include programme newsletters, applicant workshops and other guidance services.

Straightforward methodologies are needed to implement gender mainstreaming. A ‘gender filter’ has to be designed which can be applied by non-specialists to all projects at their development and then assessment stage to identify potential gender-related implications.

To respond to programme expectations, applicants will require a straightforward methodology, applicable across a wide range of projects, to identify – if not to resolve – the possible negative gender outcomes of their projects. Such methodologies are under development in several EU Member States, and their development should be accelerated for Structural Fund programmes.

At the project assessment phase, clear and consistent frameworks are needed to verify that projects have responded appropriately. The same methodology which applicants are encouraged to use could be used by project decision-makers to ensure consistency. The need for consistency is particularly

226

appropriate where variable award rates are being used as an incentive to applicants. Higher awards to selected projects can only remain credible where the necessary back-up is in place, including verification that commitments are realistic, monitoring to ensure that they are delivered, and claw-back for non-delivery.

8.8.4 Gender Mainstreaming Recommendations: Monitoring and Evaluation

Monitoring the outcomes of programmes by gender must become more widespread and systematic, and systems need to be adapted to deliver this. To facilitate evaluation, more base-lines should be established.

The deficiencies in current SPDs suggest that further guidance is required by many programmes on appropriate gender-disaggregated indicators and their good use (including the importance of establishing baseline situations against which to measure progress). The EC should in future require more gender-disaggregated indicators and targets from programmes and enhanced baseline information, and programmes and their monitoring systems should be adapted to collect and collate such information from project implementers. Establishment of more baselines, both for the wider economic situation and the gender-disaggregated performance of past Structural Fund economic development interventions would be a valuable contribution.

Organisations commissioning Structural Fund evaluation studies should include gender analysis as a standard component. Dedicated gender studies should also be commissioned where there are information gaps.

The commitment has already been made to establish gender analysis as a standard component of all Structural Fund evaluation studies, and this should be followed through. In addition, given the novelty of this theme, there continues to be a strong case to commission dedicated thematic studies, especially where this has the potential to make a significant contribution to ensuring mainstreaming is taken forward in a robust and meaningful way. To guarantee the quality of outputs, work should continue to refine appropriate methodologies for gender impact evaluation and to train evaluators in this field. The lessons of all relevant evaluation studies and identified best practice need to be disseminated more effectively than is currently the case, including to the wider non-specialist economic development community.

227

9. MONITORING AND EVALUATION

9.1 Introduction Monitoring and evaluation are two of the most difficult regulatory requirements of the Structural Funds to implement. Both are complex processes, especially when applied to the assessment of the efficiency and impact of Structural Funds operations. Traditionally, monitoring systems were exclusively concerned with financial monitoring for audit purposes rather than as a tool of policy or programme management. Up to the late-1980s, regional policy evaluation was concentrated in northern Europe, the UK and Germany in particular, and was relatively weak elsewhere.262 Since 1988, both monitoring and evaluation have developed considerably, in large measure because of the Structural Fund regulations but also reflecting a more widespread concern with the efficiency (value-for-money) and impact of policy instruments within the economic development field.

The following chapter examines the monitoring and evaluation of Objective 2 programmes. It begins with a review of the evaluation of monitoring and evaluation over the past decade. It then discusses the process of monitoring with a systematic assessment of good practice at various stages in the programme delivery cycle, drawing on experiences of Objective 2 programmes. Lastly, the evaluation of Objective 2 interventions is considered with a comparative perspective on three major evaluation exercises undertaken over the past three programming periods.

9.2 The Evolution of Monitoring and Evaluation In the past, the systematic monitoring and evaluation of EU regional policy was limited.263 Before the 1988 reform of the Structural Funds, it was generally recognised that the monitoring and control of Community regional expenditure was inadequate. In part, this was due to the primacy of the Member States in the field of regional development; ERDF funding was used to co-finance regional development projects, largely subsumed within national regional policy budgets. Also, evaluation research was poorly co-ordinated within the European Commission. Different Commission services determined their own priorities and methodologies with little central control or co-ordination, and there were conflicts with Member States that had different expectations and traditions of evaluation and contrasting methodological approaches. Monitoring and evaluation techniques varied greatly throughout the Community, with little detailed guidance for the Member States. The uneven spread of capabilities across the Community was attributed to a series of ‘compartmentalised schools’ lacking any formalised system of knowledge.264

262 Wadley D (1986) Restructuring the Regions: Analysis, Policy Model and Prognosis, OECD, Paris. 263 Bachtler J and Michie R (1995) A New Era in EU Regional Policy Evaluation? The Appraisal of the Structural Funds, Regional Studies, 29 (8), 745-751. Eskelinen H, Kokkonen M and Virkkala S (1997) Appraisal of the Finnish Objective 2 Programme: Reflections on the EU Approach to Regional Policy, Regional Studies, 31 (2), 167-172. 264 MEANS (1993) Methods to Give Meaning to the Evaluation Obligation: The Conclusions of the MEANS Programme, CEOPS, Lyon, MEANS/93/13/EN.

228

The 1988 reform led to the appraisal and assessment of Structural Fund operations being progressively enhanced in scope, scale and rigour. As part of the new, programme-based approach, the reform introduced a comprehensive monitoring and evaluation system. It required a Monitoring Committee to be set up for each Community Support Framework (CSF) and Operational Programme (OP), comprising representatives of national and regional authorities and the Commission. The 1988 Regulations also required structural operations to be the subject of ex ante and ex post assessment to highlight their impact with respect to the priority objectives (1-5b) and to analyse their effects on specific structural problems. Evaluation was to assess three levels of impact: the overall impact of Community operations in terms of strengthening economic and social cohesion of the Community; the impact of operations under each CSF; and the impact of individual operations.265

During the 1989-93 programming period, the structures for monitoring and evaluation began to be put in place by the Member States and the Commission. However, both monitoring and evaluation represented one of the least-developed aspects of early Objective 2 programming. Data was poor or non-existent, there was an absence of targets and indicators, and monitoring systems (for physical indicators at least) were rudimentary and unsystematic, with major variations between regions.266 In some regions, monitoring was limited to purely financial indicators; more effective monitoring was sometimes seen as being, at best, of secondary importance. Common views were that: “it will be undertaken when it is needed but for now we are too busy getting things done” or “the priority is to make programmes run – then we can worry about making them run well”. Indeed, for many, monitoring and evaluation requirements were seen as irritating, unrealistic and bureaucratic.267

While the value and perhaps even necessity of monitoring was generally acknowledged in principle, there were reservations among Member State authorities concerning the Commission’s expectations of how the activities should be carried out in practice. The Commission requirements had not always been coherent or clearly expressed, and the advice provided on interpreting and complying with them was of variable quality and usefulness. In retrospect, Commission officials acknowledge that they were only in a position to offer clear advice from the mid-1990s. Indeed, the Court of Auditors argued that the system for monitoring operations receiving EC funding had become less efficient and that impact assessments provided insufficient evaluation of the performance and achievement of funding operations.

Various reasons may be suggested for the difficulties experienced. First, the regulatory requirements for monitoring and evaluation were, at times, open to interpretation, and different Member States, regions and services of the Commission interpreted them in the light of their own particular interests at any point in time. Second, the programming approach was new and evolving, with considerable adjustment to their mode of implementation being required from all participants in the process. Third, a key principle of the Structural

265 Bachtler J and Michie R (1993) The Restructuring of Regional Policy in the European Community, Regional Studies, 27 (8), 719-725. 266 Bachtler J and Michie R (1995) op. cit. Eskelinen H et al (1997) op. cit. 267 Kearney C, Taylor S and Bachtler J (1996) Monitoring and Evaluation in Objective 2 Programmes: Progress and Challenges, IQ-Net Thematic Paper, 1(2), EPRC, University of Strathclyde, Glasgow.

229

Funds was partnership, but the individual partnerships which developed were rarely among equals, enabling some national authorities to avoid or block compliance with monitoring and evaluation requirements. Lastly, the opportunity for evaluation had been limited. Because of programming delays following the 1988 reform, some OPs were not approved until 1991, and a significant proportion of expenditure had still not been spent by 1992 (or even 1993).

As the first, post-reform funding period drew to a close, the Commission experienced an upsurge of interest in evaluation. In 1992, “situated at a turning point between the end of the first actions lasting several years and the preparation of the ensuing Community interventions, evaluation work...received fresh impetus”.268 In addition to the regulatory obligations, there were “ever increasing demands from the Commission's services themselves on the one hand, and on the other, other Community institutions who call for more comprehensive information on the implementation of programmes and their effects”.269

The Maastricht Treaty raised the profile of economic and social cohesion and the impact of Community policies. At the Edinburgh Council in 1992, it was agreed that greater emphasis would be given to monitoring and ex ante and ex post evaluation. In the new Regulations approved in 1993, a clearer distinction was drawn between three stages - appraisal, monitoring and evaluation.270 With respect to monitoring, the role of the Monitoring Committees was strengthened. Subject to the overall amount of the Community contribution and within harmonised limits, the Committees were given the flexibility to modify the procedures for granting assistance and the financial plans, including transfers between Community sources of finance and adjustments to the rates of assistance. In terms of the assessment procedures, it was intended that prior appraisal and ex post evaluation should become more objective-based so that assistance could be allocated where appraisal showed medium-term economic and social benefits in line with resources allocated.

Appraisal and evaluation was to be carried out by: “contrasting the goals with the results obtained...and by reference to macro-economic and sectoral objectives and indicators based on regional or national statistics, to information yielded by descriptive and analytical studies and to qualitative analyses”. Furthermore: “appraisal and evaluation shall take account of the socio-economic benefits expected or achieved commensurate with the resources deployed, conformity with Community policy and rules...and the conditions in which the measures are implemented”.

The new Regulations emphasised that the Member States as well as the Commission were responsible for evaluation, and that it was the duty of ‘competent authorities’ in the Member States to ensure that appraisal and evaluation was carried out in the most effective manner. In submitting their regional development plans, Member States were also required to provide more

268 MEANS (1993) op.cit. 269 CEC (1992) Working Programme of the Commission on Structural Policy Evaluation Matters for 1992, internal Commission note (quoted in MEANS (1993) op. cit.). 270 Bachtler J and Michie R (1994) Strengthening Economic and Social Cohesion? The Revision of the Structural Funds, Regional Studies, 28 (8), 789-796.

230

information on the current regional development situation, the impact and effectiveness of Community-financed operations undertaken in the 1989-93 programming period, and the anticipated objectives (quantified where possible) and impact of future operations.

In addition, the Commission took steps to improve the effectiveness of evaluation procedures from 1994 onwards.271 In order to upgrade the data available to evaluators, ‘baselines’ of the regional development situation were to be identified at the outset of the planning period against which the effect of subsequent operations could be assessed. The Commission also established a co-ordinated approach to regional policy evaluation with a specialist evaluation unit created (first within DG XXII, later moved to DG XVI) to have lead management responsibility for ex ante, interim and ex post evaluation of Structural Fund operations in Objective 1 and 2 regions. The MEANS programme, Evaluation Methods for Actions of a Structural Nature, was launched with the task of developing a ‘European evaluation culture’ by improving the quality and utility of evaluations. In addition, exchange of experience among EC countries was organised by the Commission through several fora, including an evaluation working group comprising representatives of the Member States. From the mid-1990s, it was judged that: “European evaluation practice is rapidly expanding, through a snow-ball effect between different levels of government: the French and English cases are relevant examples. In these countries local and regional authorities reacted to national evaluation practices by developing their own evaluation activities. An analogous process can be observed... in Southern European countries”.272

By the second programming period, there were definite signs of improvement, not just in terms of comparisons between the 1989-93 and 1994-96 Objective 2 programmes but also between the regional development plans submitted and SPDs agreed for the 1994-96 period. Many programmes were judged as having satisfactory quality indicators, with the best programmes having indicators at three levels (programme, priority, measure). However, there were still important deficiencies: incomplete indicator lists; lack of quantification at global objective, priority and measure levels; missing targets for monitoring; and unclear assessment indicators.

A comprehensive review of 1994-96 Objective 2 SPDs and those approved in 1995 for the new Member States,273 found that, while most made reference to the envisaged positive employment effects, just under half had absent, unquantified or obviously incomplete employment indicators. Where targets were set, programmes usually failed to specify the type of impacts (eg. direct permanent jobs only or also indirect jobs and temporary employment generated in implementing measures) and their timescale, preventing future evaluators from comparing performance meaningfully with the targets. There was also a lack of precision and consistency in the terminology used to refer to the main types of possible outcome. Where employment indicators were in place, there was little evidence of a ‘scientific’ approach being taken to

271 Bachtler J and Michie J (1995) op.cit. 272 C3E (1995) Centre for European Evaluation Expertise, Information Note. 273 Bachtler J and Taylor S, with Kearney C (1996) Extended Synthesis of Agreed Single Programming Documents in Objective 2 Areas 1994-96, Report to the European Commission (DG XVI), EPRC, University of Strathclyde, Glasgow.

231

quantification. Only one programme explicitly stated the assumptions on which targets were set, and only then for construction jobs. Given the lack of explicit methodologies in use, it was clear that some programmes had set ambitious targets and others chosen to be more conservative. There was little or no attempt to go beyond gross figures to indicate what the net outcomes might be when the negative impact of displacement and deadweight and the positive impact of multiplier effects had been taken into account. The result of these factors was that, when programmes were compared for the promised ‘cost per job’, the variations were considerable, not only between countries, but also between ostensibly similar programmes in the same Member States.

Over the 1994-96 period, a considerable amount of further development work was undertaken associated with: clearer, quantified targets; monitoring and evaluation indicators at all levels of the programme, distinguishing between outputs and impacts; and more efficient monitoring and evaluation organisational arrangements eg. management information systems. Commission services worked closely with programme managers and secretariats to improve the range and quality of indicators as well as monitoring systems. The MEANS programme issued a range of monitoring advice, leaflets and programme manager hardbacks as well as providing support for evaluators. A series of high-profile evaluation conferences was organised by the Commission (in Brussels, Berlin and Seville) to promote the monitoring and evaluation ‘message’ and to encourage exchange of experience among regions.

In several countries, national government authorities were increasingly establishing an overall evaluation framework or common guidelines for evaluation. In the United Kingdom, the Scottish Office and the Department of the Environment formed a sub-group to produce a guidance note, Framework for the Evaluation of Area-Based European Structural Funds Intervention, for all government departments; the guidelines explained the purpose of evaluation and the steps to be considered, and they provided a framework for setting terms of reference for each type of evaluation. In Scotland, more detailed guidance was provided by the Scottish Office which prepared draft terms of reference for evaluation studies, in consultation with programme managers. A similar approach was adopted in France where DATAR produced a Vade Mecum for the Structural Funds backed up by a specific guidance notes on specific types of evaluation (agreed with the Commission) and held a series of seminars with programme managers. DATAR also encouraged regional managers and partnerships to build links with the long-standing government regional evaluation units which provide substantial expertise and resources for evaluating government policies.

The provision of central support and resources was taken a stage further in Austria. The Österreichisches Raumordnungskonferenz (ÖROK) and Federal Chancellery invested heavily in promoting evaluation awareness and interest since evaluation did not have an established culture or tradition in Austria. All Structural Fund evaluations were co-ordinated by ÖROK and involved a range of federal and regional interest groups; exchange of experience meetings between the evaluators for Objective 2 (and other objectives) were organised, and a central resource ‘Checkpoint EVA’ was created by the Federal Chancellery to promote new thinking in the field of evaluation, with three

232

main foci: a library of evaluation literature based in a Vienna research institute, bringing together reference material on evaluation techniques and studies from throughout Europe, as well as providing a depository for all evaluation studies conducted in Austria; a discussion group, comprising experts in the evaluation field; and a consultancy service, aimed at evaluators and public officials.

The reprogramming of Objective 2 interventions in 1996-97 provided an opportunity to strengthen further the quality of monitoring and evaluation systems. Research showed that, in the more advanced cases (the longest-running programmes), there were ambitions of creating fully integrated physical and financial reporting systems, and comprehensive monitoring and evaluation frameworks. Also, regions increasingly accepted the value of monitoring and evaluation exercises. A review of the 1997-99 Objective 2 programmes carried out by the Commission,274 found that quantification of effects for the 1997-99 programmes had improved significantly, 55 of the new SPDs giving detailed measure-level information.

Several increasingly common good practice features can be highlighted among Objective 2 regions. First, as noted in the earlier chapter on job creation, more programmes are specifying what is included in employment targets, detailing some of the assumptions on which these targets are based, and indicating the timescales over which they should be realised. Second, the type and quality of jobs created is being specified with increasing care. There is differentiation between ‘temporary’ employment generated by implementing projects (eg. in construction, often referred to in terms of ‘person years of work generated’) and ‘permanent’ posts (eg. created in the longer term by the use of infrastructure). Also, there is more specification of the nature of the jobs created (full-time, part-time, seasonal, etc). Third, there has been increasing precision in the use of employment-related terminology and concepts, indicating greater familiarisation with the field and increased reflection. Definitions are being used which more accurately reflect the true nature of employment impacts. No longer are people entering employment as a result of enhanced skills being described as ‘jobs created’ (except where they are entering self-employment).

In spite of improvements, these reviews make clear that experience across programmes is still mixed. Several of the above observations relating to the position of Objective 2 regions in 1994 still apply eg. inadequate indicator information, lack of appropriate and practical targets, and lack of consistent and regular data collection systems. Employment targets still often only encompass direct effects, and are expressed in gross terms, the negative effects of deadweight and displacement and positive multiplier effects not yet being taken sufficiently into account. In addition, the assumptions on which targets are based are still not consistently set out. These features are symptomatic of the virtual absence, in many cases, of robust methodological approaches to quantifying programme outputs and impacts. Lastly, it has been suggested that an estimated third of programme resources are wasted - spent on projects

274 CEC (1997) The New Regional Programmes 1997-99 under Objective 2 of the Community’s Structural Policies - Focusing on Job Creation, COM (97) 524 final, Commission of the European Communities, Brussels.

233

which could have gone ahead anyway with less or no Structural Fund support, or whose net outcomes are limited because they simply displace other economic activities in the eligible area.

9.3 A Systematic Approach to Monitoring Having traced the evaluation of monitoring and evaluation over successive programming periods, the next question is what can be learned by comparing the approaches of Objective 2 programmes across Member States. This is of particular concern in the field of monitoring, for which the programming challenges are becoming progressively more intensive. Taking the ‘virtuous circle’ of project delivery presented in Chapter 2 as a starting point, the most highly evolved reaction to monitoring is to integrate it as a continuous process in all programming stages (from strategy development to final evaluation) and with all actors (from the programme manager to final beneficiaries). Success here requires practical approaches to be designed and pursued in partnership. For programme managers and partners, a summary checklist on monitoring issues and good practice is provided below.

9.3.1 Programme Management In preparing and developing programmes, monitoring tasks are complex and technical, requiring specific skills and knowledge. The more programme actors who understand the terminology and methods of calculating, the better equipped a programme will be to respond. The most important tasks which face programme secretariats and other key agencies are:

��assessing the implications of the proposed policy mix at the programme development stage;

��undertaking an ex ante evaluation of anticipated effects which will provide the framework for future assessment of projects and for programme evaluation;

��applying a recognised terminology for effects, and ensuring this is well known among relevant actors;

��providing frameworks which can be used to assess how realistic applicant predictions of impacts are;

��ensuring the monitoring system is sufficient to track outcomes and to feed this information back into the programming process; and

��liasing with specialist evaluators on measuring impacts.

In order to build up in-house expertise, a programme can draw on external sources of guidance, including the Commission and specialist evaluators, recruit specialists themselves, or train existing personnel. DG XVI has focused significant attention on providing targeted and practical advice and support. Guidance has been developed in several fields through the MEANS programme, including the ex ante estimation of the employment effects of programmes,275 and quantification of employment effects in interim or ex post

275 MEANS (1997) Counting the Jobs. How to Evaluate the Employment Effects of Structural Fund Interventions, DG XVI Series: Evaluation of Documents, No. 1, January 1997.

234

evaluations.276 In some cases, national ministries have also worked to provide effective support to programmes. The publication series of the Finnish Ministry of Labour, for example, has featured occasional guidance on both the ESF and the ERDF.277

In addition to having a core of expertise, it is desirable that a wider population of partners has a broad knowledge of the terms and issues, and sufficient familiarity to understand and apply frameworks prepared on their behalf. These public sector actors can then use their knowledge at all stages of programming in which they are involved - in particular, in preparing their own applications for funding, in advising other applicants on how to optimise and realistically estimate the effects of proposals and in assisting programme decision-making. At a further level, final beneficiaries need to be clear about the processes through which their project will affect outcomes (eg. employment), so that they are alert to these effects as they happen and are able to supply better monitoring information.

9.3.2 Project Generation To be followed through effectively into programme implementation, the monitoring imperative has to be made explicit to applicants. The first task is to raise awareness and set out expectations, and the second to enable applicants to respond consistently and effectively to them. Once awareness has been raised, the second stage is to provide effective guidance to applicants. If applicants use a methodology known to the programme, this will be consistent between projects and enable applications to be compared, and reliable information to be gathered for the programme as a whole. It can counter applicants’ tendency to overestimate the effects of their projects; timely advice in this case reduces the chances of unrealistic project targets having to be referred back to the applicant at the initial project appraisal stage, or causing problems when they are not finally realised. Also, encouraging applicants to think through their targets and how they will be produced makes them active rather than passive observers, alert to the direct impacts of their project as they happen.

There are several ways in which guidance can be provided. In some cases, the SPD provides a consistent framework helping applicants to think through their projects. In Finland, where employment impacts are now the main selection criteria, projects are informed that they have to fulfil at least one of five specified employment-related criteria to be eligible: (a) significant direct effects on employment at start-up stage; (b) indirect effects at start-up stage; (c) jobs created during the operating stage; (d) corporate investment creating new jobs or maintaining existing ones; or (e) training, research or development projects improving the employment potential of participants.

276 MEANS (1995) Measuring the Employment Effects of Community Structural Interventions, MEANS Handbook No. 3, Commission of the European Communities, Brussels. 277 The series has included ‘Effectiveness, Dead-weight Effects and Other Concepts of Evaluation’ (1996) produced for Objective 4 operatives, and ‘Assessment of the Employment Effects of Investment Projects Launched under EU Regional Objective Programmes. Pilot Study of the Objective 2 Programme’ (1998, in Finnish).

235

9.3.3 Project Appraisal and Selection Ideally, a systematic and user-friendly framework is required, which mirrors that used in the target-setting exercise undertaken for the whole programme, and which:

��provides a framework to help applicants think through and then quantify the effects of different types of project; and

��enables the internal appraisal and, if necessary, adjustment of applicants’ estimates, so that projects can meaningfully be compared at the selection stage, and their implementation monitored against realistic targets.

Internally, at their most basic level, systems should provide a framework which enables appraisers to assess whether the cost per job which applicants predict is within a range which previous experience suggests is realistic. In Finland, applicants’ estimates of employment impacts are assessed, and, if they are judged to be too high, then the authorities dealing with the application make their own estimate of growth effects so that the proposal can be properly assessed alongside others. The Industrial South Wales appraisal process relies more on applicants, who are asked to justify their targets by specifying the assumptions on which they have been calculated. The targets can then be assessed meaningfully, further information or revision being requested if necessary.

A useful nuance of appraisal systems is for them to take into account the quality as well as quantity of jobs being affected/created by given projects (as in Ångermanlandskusten and Steiermark). There are several ways by which this can be measured, including use of broad categorisations of employment status (permanent, temporary, seasonal, part-time, full-time), the grade of position (unskilled, skilled, technical, professional, managerial), or the sector in which employment will be created.

Rather than every individual programme creating systems, it can be effective to work at national level eg. the Finnish TVA methodologies and handbooks. One appraisal system which is able to bring out the value of projects with limited direct impacts but which strengthen other employment-generating actions, is that which is being introduced under the new English ‘action plan’ and ‘package’ approach to programme implementation.278 Groups of projects under the same priority heading and targeted at the same part of the eligible area are appraised together by sub-groups of partners, at the early stages of programme implementation. This enables potential synergies between projects to be demonstrated, and thus the value of individual projects to be assessed in a more rounded way than if they were submitted individually, over a longer time period and on a more ad hoc basis, to the programme-wide appraisal process.

Two further issues which project appraisal systems need to be able to take into account are deadweight and displacement. These are much more than peripheral concerns: some commentators argue that robust approaches to project selection which minimise deadweight and displacement are critical to

278 Taylor S and Downes R (1997) Meso-Level Partnerships and Structural Fund Implementation, IQ-Net Thematic Paper, 3(3), EPRC, University of Strathclyde, Glasgow.

236

raising the potential employment impact of Structural Funds interventions. This was highlighted, for example, in the recent Yorkshire and Humberside interim evaluation.279 Minimising deadweight and displacement maximises the resources spent on projects which are truly additional to the regional economy, and would not otherwise have been undertaken.

9.3.4 Monitoring The primary function of monitoring systems tends to be a financial management mechanism, pacing the release of tranches of project funding, and ensuring projects are on track, usually in terms of their anticipated timetable and physical outputs, as these are generally the first to be produced. It makes less sense to focus monitoring systems on employment impacts because these, for the most part, emerge over a longer period than the duration of project implementation. Nonetheless, monitoring systems are increasingly used to gather information on the most immediate outcomes, and thus provide information which, whilst imperfect, can be important in helping to inform ongoing processes of programme implementation. Less reliably, but still usefully, monitoring can also provide information on volumes of physical outputs, from which approximate likely impacts achieved can be calculated.

In order for monitoring information to be useful, it is necessary that both physical output and economic impact monitoring information is actually requested from project implementers and then collated. To be entirely valid, the reliability of monitoring data has to be verified. Without programmes of inspection visits, it is difficult to ensure that the project level monitoring information supplied by implementers is a true reflection of reality.

The timeframes over which monitoring takes place are important. Clearly, monitoring needs to captures as many of the direct and short term impacts as possible. As monitoring is only able to capture the most direct impacts (eg. construction jobs and other recruitment directly related to project implementation), it is useful if all of these are captured. However, programme periods are generally too short to encompass medium and longer term project impacts; monitoring usually stops when the last report is submitted at the end of a project’s lifespan. Hence, there is a strong argument for monitoring information to be collected over as long a period as possible, eg. through post-award follow-up questions, as an intermediate information source before ex post evaluation results become available.

The operationalisation of a systematic approach is highly challenging and requires appropriate robust and reliable monitoring systems. In order to illustrate the application of some of these systems in practice, the following three case studies are drawn from contexts relevant to Objective 2 programming: the EVI monitoring systems in Bremen; the monitoring system in Austria; and the PRESAGE system in France.

a. Monitoring Structural Fund programmes in Germany (Bremen) A computerised monitoring system was developed between 1991 and 1993 by the Bremen Economic Research Institute, initially for the evaluation of Phase

279Centre for Training Policy Studies (Ed.) (1997) Review of the Yorkshire and the Humber Region Objective 2 Programme 1994-96: Final Report of the Programme-wide Appraisal, Report to the Government Office for Yorkshire and Humber.

237

1 of the Objective 2 programme. The system, whose development was co-financed by Objective 2, is now used to monitor and evaluate RENAVAL and Joint Task (Gemeinschaftsaufgabe) spending as well as that under Objective 2, and has been copied by several other programmes, sometimes in a simplified form, including informing the development of a new system in Austria under the ERP Fund.

Data collected by EVI includes: location and sectoral information; legal basis for support; details of recipients; evaluation criteria (quantitative and qualitative, and which can be specific to an individual project); the rate of subsidy (total and from all sources); eligible costs; the sources of assistance; the date of award; and progress in payments.

EVI has three functions. First, it fulfils the external requirement for monitoring, providing the basis of reports to national officials and the Commission on the financial progress of the programme, which are submitted twice a year. Second, it supports internal management functions, being used to monitor the progress of the various departments charged with implementing the programme. If demand under certain priorities is insufficient, the Objective 2 co-ordinator can then approach the department in question and discuss the problem, reallocating funds if necessary. Third, at the end of the programme, it provides information for external evaluators, including the geographic incidence of assistance, jobs created, private money contributed, etc.

b. Monitoring Structural Fund programmes in Austria When Austria first became eligible for Structural Funds support, the Austrian authorities had to devise a system to respond to its monitoring obligations. A new system needed to be established as each Ministry and funding agency had different methods and computer systems which were difficult to integrate. The ‘ERP Fund’, based at the Federal Ministry of Science, Transport and the Arts, became responsible for gathering the monitoring information.

The system operates at national level, collating data submitted from regional programmes and then feeding it back to them. Funding agencies submit lists of projects and awards every three months - either directly to the ERP or through a co-ordinating body eg. the EU Geschäftsstelle for Land level departments and funding agencies. Every group of indicators has an individual form - six in total (for industry; tourism; infrastructure; RTD; commercial environmental projects; and consultancy). The same structure is used for all the forms, but with different indicators. This data from the forms is then entered into the monitoring system. The results can be grouped by measure and by aid scheme under each measure, thus directly matching the structure of the SPD. The anticipated financial allocation for the measure and the actual award level to date can then be compared. All this information, together with a list of all the individual projects, goes back to the funding agencies. This informs the Federal Chancellery, as well as the funding agencies themselves, of the stage of the implementation, and the level of demand for measures and aid schemes.

A similar system was developed for ESF monies. Information collected includes the difference between the amounts/conditions agreed and the

238

amounts paid out or conditions met. The federal Ministry of Labour and Social Affairs carries out further monitoring and evaluation on the basis of this information.

The monitoring systems are still rudimentary. For example, environmental indicators were particularly difficult to develop as every new plant or investment is likely initially to have negative environmental effect in terms eg. of emissions. It is also difficult to find indicators which can accurately reflect the effect of investment, particularly given the changing nature of economic development and the shift towards a greater focus on the innovative, rather than the job-creating, content of projects. The evaluation indicators were initially quite mixed in quality, with some being goals rather than indicators. The universal problems of statistical availability and identifying indicators which accurately reflect the economic impact of measures are also evident.

c. Monitoring Structural Fund programmes in France In a project entitled ‘Programme Régional et Européen Suivi-Analyse Gestion Evaluation' (PRESAGE), the French national Objective 2 technical assistance programme is developing a single, standard, integrated computerised management information system to monitor Structural Fund and other programmes across France. The new instrument is being piloted during 1999, and a definitive version will be installed in all regions by 2000, supplanting existing systems.

The system, which will track the implementation of Structural Fund programmes, state programmes and the Contrats de Plan, will make it easier for national ministries to track what is going on in the regions, and will also enable the stronger monitoring requirements of the new Structural Fund programmes to be fulfilled – including making programmes more easily comparable. The database is linked to the Contabilité Publique, and to the ESF database (although currently, information can only be imported from this and not exported to it). The system will feed information from each region to the national level, and contains cascading tiers able to show whole programmes, measures, sub-measures and individual projects, in increasing levels of detail.

At the regional level, the system will be customised to the needs of individual Préfectures, which will be able to determine the accessibility of the network. In theory, in the regions working in the most co-operative ways, a wide range of organisations could input and extract information directly.

The system draws on experience in several regions, including Aquitaine,280 but with further refinements. Projects are recorded at the moment of ‘instruction’ - when they first enter the appraisal system - and are tracked through to completion. Project data are aggregated to measure and programme level. Projects are allocated standard codes so that projects of the same type can be identified easily across very different programmes. The system can be used to generate letters for applicants and implementers, to show where paper files are kept, to keep a trail of communications, and to generate statistical and graphic summaries for reports on programme progress. The PRESAGE database is

280 Taylor S (1998) Objective 2 in Aquitaine, IQ-Net Thematic Paper, 4(4), EPRC, University of Strathclyde, Glasgow; and Kearney C, Taylor S and Bachtler J (1996) op. cit.

239

being shaped to enable it to supply much of the necessary data for the proposed annual implementation reports.

9.4 Evaluation Monitoring is a continuous process, increasingly (and ideally) integral to every part of programming and the project delivery cycle. Evaluation, while a periodic event in itself, conducted at key points in the programming period, can be conceived as an element of this ‘continuous’ process, capitalising and building on the foundation of monitoring data to provide timely additional information on past activity and recommendations for future orientations. The following sections provide a comparative perspective on each of the three types of evaluation undertaken in the context of Structural Fund programming – ex ante, interim and ex post – drawing lessons from the experience of evaluating Objective 2 programmes across the EU.

��1993-94: Ex ante appraisal. The most thorough ex ante appraisal was conducted for the 1994-96 programmes. Appraisal of the 1989 programmes was cursory and patchy, while the ex ante appraisal of the 1997-99 programmes was abbreviated (though standardised and systematic) covering the background analysis, strategy, measures and impact of SPDs.

��1995-96: Ex post evaluation. The only major ex post evaluation of Objective 2 programmes completed to date relates to the 1989-93 programme evaluations undertaken over the 1995-96 period. Ex post evaluations of the 1994-96 programmes are currently underway.

��1996-97: Interim evaluation. Again, only one major and comprehensive interim evaluation of objective 2 programmes has been undertaken. During 1996-97, most Objective 2 programmes undertook an interim evaluation exercise with a view to assessing progress over the 1994-96 period. Interim evaluations of the 1997-99 programmes are currently underway.

9.4.1 Ex Ante Appraisal The 1993 Regulations provided the basis for the most systematic approach to evaluation yet attempted. As noted in Chapter 4, in March/April 1994, the regional development plans for Objective 2 regions were submitted. Each of these plans and documents was subject to ex ante appraisals by the Commission.

A key feature of this process was the Ex-Ante Evaluation Framework produced by the Evaluation Co-ordination Unit of DG XVI. Drawn up by a working group comprising the Commission and Member States, the document was non-binding.281 "Although it relates to a frame of reference that will be used by the Commission's services, it is however essential, in the framework of partnership, to bring it to the attention of the Member States to enable them to be aware of the intentions of the Commission in terms of ex-ante evaluation. It thus permits a clarification of the type and level of information that the Commission judges necessary to be in a position to carry out its regulatory

281 CEC (1993) Ex Ante Assessment Framework, Working Document from DG XVI, Commission of the European Communities, Brussels.

240

obligations in terms of evaluation” The ex ante framework was based primarily on an operationalisation of the regulatory requirements and was compatible with the ROAME technique, an established approach to prior appraisal of programmes, the acronym standing for: Rationale, Objectives, Appraisal, Monitoring and Evaluation. The approach was modified to incorporate additional elements relevant to Structural Fund intervention.

The guidelines began with the requirement for a qualified presentation of the rationale on the basis of which the strategy and objectives of the programme were defined. This encompassed: an analysis of the disparities and backwardness in development/reconversion and of the socio-economic problems; and evaluation of the main results of actions co-financed in the previous CSF (previous programming experience). The next stage was to define the overall regional development strategy, the main axes of the development action and definition of the specific objectives of each of these, indicating their capacity to address the development problems identified.

A cornerstone of the Commission's ex ante evaluation guidelines (reflecting the Regulation requirements) was for the objectives to be quantified and assessed. The purpose was to provide indicators able to assess the anticipated effectiveness at each relevant level. For each development axis, the specific objective was to be accompanied by indicators, broken down into basic infrastructures, productive environment and the development of human resources. Anticipated achievements were to be presented by listing the indicators, baselines (basic values) and target values or, in relation to performance, the initial situation and expected situation after achievement. Appraisal of the impact on the environment was of particular concern, covering an assessment of the environmental situation of the area, the environmental impact of the strategy envisaged, and the environmental impact of proposed actions. Reflecting the need for a prior appraisal of impact, the guidelines also required details of the types of aggregate effects expected as regards socio-economic advantages at both microeconomic and macroeconomic levels. In the language of the guidelines, it should involve carrying out an analysis of the general economy of the programme, the logic behind its structure and the interaction of its various components.

The requirement for external coherence was defined as the extent to which the programme was conditioned by national and Community policies. The issue of coherence covered three main issues: links with Community policies, notably environmental and competition; links with national policies - the extent to which the strategic approach of the programme and its underlying specific objectives strengthened the effects of national economic and structural policies; and additionality - the extent to which the programme entailed an increase in public expenditure.

The final issue was that of implementation covering the issues of: partnership - prior consultation procedures and involvement of regional socio-economic actors; legal context - lead authorities and ownership of the programme; financing - national and regional sources of co-financing; and administrative and financial organisation - financial control, administrative and financial circuits, and monitoring and evaluation procedures.

241

The systematic ex ante appraisal of draft objective 2 SPDs undertaken along these lines during 1994 (1995 for the new Member States) represented a significant step forward for the Commission. It held out the prospect of a more uniform and structured Commission approach to the appraisal and negotiation of Single Programming Documents (SPDs) and the derivation of systematic baselines for forecasting the impact of Structural Fund operations. The impact of the appraisal process alone is difficult to isolate, since it formed only one of the inputs into the negotiation of plans and agreement of CSFs and SPDs, but the appraisals affected the development strategies in three main ways.

First, evaluation reports made a genuine contribution to the negotiations and the development of plans and SPDs, in the form of:

��improved definition and description of strategies, objectives and measures, giving the strategies and actions increased clarity;

��greater coherence within strategies, especially an emphasis on integrated aims and actions rather than investment themes;

��more integration between the Funds, particularly through a greater emphasis on multi-fund objectives;

��the wider inclusion of project selection criteria; and

��organisational improvements through better monitoring and evaluation arrangements, stronger management structures, and more involvement of the social partners.

Second, appraisal assisted with the derivation of systematic baselines; the level of quantification was improved with more focused impact and output indicators. These quantified indicators of performance and impact enabled the Commission to begin aggregating the anticipated impact of Community assistance and to assess the (created and safeguarded) employment associated with Objective 2 strategies.

Third, the appraisals promoted integration (although imperfectly) of environmental concern within economic development. In several cases, attention was drawn to the environmental impact of proposed actions, notably major (potentially intrusive) infrastructure projects such as harbour development. In at least one region, further environmental impact assessment was required before measures were approved.

However, scrutiny of the individual CSFs and SPDs and their ex ante appraisals revealed several important weaknesses in the quantification of impact. Member State programming documents were generally inadequate for evaluation purposes. As noted in Chapter 4, regional development strategies were often loosely related to the problems of the regions that they were meant to address, lacking a clear link between strategies and anticipated impacts. Further, the nature of some designated areas was not conducive to evaluation. Many Objective 2 areas were small, geographically fragmented in some cases, and inconsistent with the territorial units used for statistical data collection. Even basic employment and unemployment data was sometimes not available for such regions. There were also fundamental methodological questions as to whether a prior appraisal of the impact of Objective 2 programmes was

242

possible. In many SPDs, the quantification of objectives was only partially complete with output indicators are provided instead of impact indicators, especially for measures concerned with human resource development and economic infrastructure. The Commission faced particular problems in the UK and Germany with regard to obtaining meaningful impact projections. Finally, ex ante appraisals were not always undertaken in a consistent manner. There was considerable variation in the calibre of independent experts and in the methods used for undertaking appraisals, ranging from desk studies based entirely on secondary sources to detailed interview-based surveys.

9.4.2 Ex Post Evaluation To date, the only comprehensive ex post evaluation of Objective 2 programmes which has been completed was that prepared for the European Commission in 1996, evaluating the 1989-93 programmes.282 The aims of the evaluation exercise were, first, to assess the results obtained during the programming periods 1989-91 and 1992-93, compared with original objectives, and, second, to gather insights contributing to improving the functioning of ongoing and future interventions.

The regulatory basis of the evaluations was contained in the 1988 Structural Fund regulations, notably Article 6(2) of the Framework Regulation, which stated that “In order to gauge their effectiveness, Community structural operations shall be the subject of an…ex post assessment designed to highlight their impact with regard to the objectives…and to analyse their effects on specific structural problems.”283 The Co-ordination Regulation elaborated further284 (Article 26(2)), stating that the studies were to address programme effectiveness at three levels: overall impacts on Community economic and social cohesion, the impact of each type of CSF and the impacts of individual programmes, achieving this by reference to macroeconomic indicators based on regional or national statistics, information yielded by descriptive and analytical studies and qualitative analyses.

In order to deliver a comprehensive and comparable overview, individual ex post evaluations of the sixty 1989-93 Objective 2 programmes were undertaken and then synthesised into reports for eight regional groupings (Metropoles, Industrial Cores South and North, North Sea, Mediterranean, Atlantic Regions, Italy and UK). These in turn were brought together by Ernst & Young into an overall evaluation. The various groups of studies were undertaken with a broadly uniform methodology to ensure comparability, although this had to be adapted in different contexts according to the widely varying information which was available.

282 Ernst & Young (1997) Ex-Post Evaluation of the 1989-1993 Objective 2 Programme: Synthesis Report, DG XVI Series: Evaluation and Documents, No 4, Commission of the European Communities, Brussels. 283 Council Regulation (EEC) No. 2052/88 of 24 June 1988 on the tasks of the Structural Funds and their effectiveness and on co-ordination of their activities between themselves and with the operations of the European Investment Bank and other existing financial instruments. 284 Council Regulation (EEC) No. 4253/88 of 19 December 1988, laying down provisions for implementing Regulation (EEC) No 2052/88 as regards co-ordination of the activities of the different Structural Funds between themselves and with the operations of the European Investment Bank and other existing financial instruments.

243

The research was undertaken in two broad stages. The first, undertaken for every Objective 2 programme, consisted of a general assessment of the situation in the eligible areas before and after the programmes were implemented, and of a top-down overview of available programme data. The second stage, attempted in one in three programmes, but only in fact successfully undertaken in one in six, comprised an in-depth, bottom-up evaluation. The primary aim of this stage was to generate the necessary project-level empirical data to enable the employment impacts attributable to the programmes to be identified on a gross, then net basis.

In the bottom-up evaluations, a range of methods were used to estimate employment effects, depending on the data available. Cost-per-job data were extrapolated from samples of projects in order to obtain programme-wide figures. In the more refined studies, a different cost per job was calculated for different types of ERDF intervention, and for ESF interventions, to arrive at more accurate grossed up figures. Not every in-depth evaluation attempted to calculate net employment outcomes. Of those that did, the most refined used information from project sponsors to define and adjust assumptions relating to additionality, displacement and indirect effects.

The overall synthesis evaluation used the employment effects identified by both the universal top-down studies and the sample of in-depth, bottom-up evaluations as the basis to generate figures estimating the overall impacts of Objective 2 interventions during the 1989-93 period. The study concluded that Objective 2 interventions had played a very significant role in the industrial reconversion process. The programmes were judged to have helped sustain the process of structural adjustment in the regions despite unfavourable macroeconomic trends. The short-term impact on unemployment and longer term contribution to industrial restructuring were assessed to be measurable and significant. Some 850,000 gross jobs were estimated to have been created, saved or redistributed, of which two thirds were accounted for by ERDF interventions and a third by ESF. The distribution of these impacts between Member States broadly followed the distribution of funding which had been available. In terms of net outcomes, it was estimated that, after making adjustments for additionality, displacement and indirect effects, some 450–500 thousand jobs had been created.

In addition to the employment outcomes, the Structural Funds were found to have had a significant leverage effect: it was estimated that for every 1 MECU of Objective 2 commitments, a further 3.4 MECU was contributed by the Member States. Non-quantifiable effects were equally important, especially the stimulus given to the development of regional partnerships and the capacity to plan and manage regional development strategies. In particular, the Objective 2 programmes brought a coherence to regional development strategies in many eligible regions that had been lacking previously. Nonetheless, while the degree of external coherence was high, the Objective 2 programmes often lacked internal coherence, particularly with regard to integration between the ERDF and ESF. Failure to achieve or even to identify potential synergies between the two Funds in most cases related both to policy design and to the different configurations, histories and institutional circuits characterising each Fund. In terms of policy orientations, a considerable adjustment in Objective 2 programmes took place during the programming

244

period, and in particular at its mid-way stage, but this was more often to maximise the absorption of available funds than to respond to changing regional development priorities.

Other findings related to operational and management issues. First, the 1988 reforms to the Structural Funds were found to have provided a strong impetus to the process of partnership building; several partnership models were developed, those with a relatively small number of representatives which had the advantage of being more cohesive and easier to manage, and larger partnerships, which although more representative, were more difficult to co-ordinate. Programme management and delivery systems improved over time. However, in many cases they were underdeveloped at the outset of programming, and often developed more slowly than would have been desirable because compressed timetables meant that energy was diverted into implementation rather than the establishment of formal decision-making and monitoring procedures.

Further important lessons from the ex post evaluation can be derived from a series of methodological problems which were encountered in actually undertaking the studies, some of them characteristic of the nature of ex post evaluations, but others stemming from the inadequacy of the monitoring, administrative and decision-making frameworks established in this first programming period, or to the nature of the eligible areas themselves.

Among the problems which derive from the nature of ex post evaluation – which is undertaken some years after programmes are implemented – were:

��the difficulty or inability of obtaining project and programme records and other information relevant to the evaluation because of: personnel changes in programme management, partner and project organisations; the non-retention or archiving of programming records; and institutional reorganisation; and

��workloads of relevant organisations, including programme secretariats involved in ongoing programme delivery, which made evaluations of a past programming period a low priority.

Further problems encountered by the ex post evaluators related to the quality of early monitoring and evaluation frameworks or to the nature of Objective 2 areas themselves in the first programming period.

��There were gaps in the monitoring data available. In some programmes the monitoring systems and the indicators they tracked were weak or non-existent (eg. Franche Comté and Veneto). Others had only introduced systems part-way through the programming period or only used them to amass financial rather than output and impact data. In some cases (eg. Cumbria and the North West), data sets had to be built up from scratch using raw material before evaluation could be undertaken. Where monitoring data was available, considerable time was sometimes involved in compiling it from multiple sources.

��In spite of the time lag, not all programmes had been completely closed. For example, the North West had been granted an extension of its closing date and some project files were still open at the time of evaluation.

245

��Lastly, the measurement of progress since the start of the programming period was impeded by the fact that baseline studies had not always been carried out. Compiling a picture of baselines retrospectively was complicated by the fact that some eligible areas did not readily correspond to areas for which statistics were collected.

9.4.3 Interim Evaluation285 Among the three types of evaluation of EU regional development programmes, interim evaluation is in many respects the most important. In theory, it is the only evaluation phase that can simultaneously assess the effects of a programme and influence its operational orientation and balance. The practical decision whether to undertake interim evaluation in Objective 2 areas at the mid-point of the 1994-99 period was left to local partnerships, some of which were reluctant to take on an additional administrative burden, especially in view of the delayed start of the 1994-96 programming period. Nevertheless, with the exception of some UK regions, most Objective 2 programmes underwent some form of interim evaluation.

The starting point for assessing interim evaluation is the regulatory position. Based on the 1993 amendments to the Council Co-ordination Regulation (Article 25), the monitoring of SPDs is to be backed up by interim evaluation so that any adjustments required to the programme and operations may be made: “Interim assessments include a critical analysis of the data collected through monitoring, including those forming part of annual reports…[and]… measure the extent to which the objectives pursued are gradually being attained, explain any discrepancies and forecast the results of the operation. They also express a view about the validity of the operation in progress and its relevance to the objectives selected.” From an EU perspective, it is clear that interim assessment is an important part of the monitoring process. This is explicitly restated in the Commission guidelines on monitoring and interim evaluation: “the main purpose of interim evaluation is to improve the monitoring mechanism and arrive at a tool which can be useful for decision-making and effective management”.286 Although the statutory obligation was mainly intended for operations lasting longer than three years, Commission services encouraged Objective 2 regions to undertake ‘mid-term reviews’ wherever possible, for several reasons.

First, the quality of monitoring systems and data in many regions was still considered unsatisfactory, a point reinforced by the 1996 ex post evaluation exercise. Second, the Commission was interested in assessing how the 1994+ programmes were proceeding, especially those launched for the first time in the ‘new’ Member States. Third, Commission services wanted to obtain information that could guide their approach to the 1997-99 programmes, particularly in promoting the new areas of emphasis - job creation, RTD, environment and equal opportunities. Equally, considerable stress was placed on the fact that interim evaluation should be seen as an asset to programme

285 Bachtler J, Michie R and Downes R (1997) The Interim Evaluation of Objective 2 Programmes, IQ-Net Thematic Paper, 2(1), EPRC, University of Strathclyde, Glasgow. 286CEC, Common Guide for Monitoring and Evaluation, Commission of the European Communities, Brussels.

246

management and not solely as a response to Commission requests or regulatory requirements.287

In 1996-97, evaluation studies were carried out more systematically than was the case during the 1989-93 programming periods, but the current coverage was still patchy. Belgium (apart from Liège and Aubange), the Netherlands and France had comprehensive evaluation coverage; among German Objective 2 regions, external evaluations were conducted for major programmes, but only internal assessments for the smaller-scale Objective 2 programmes. Finland effectively conducted an ex post evaluation for the 1995-96 Objective 2 programme, similar to the approach in Austria and Denmark with a single evaluation covering all of the Objective 2 regions in the country. Sweden evaluated Objective 2 programmes in two groups, encompassing three northern and two southern Objective 2 regions. In the UK, interim evaluation was largely limited to Scotland, the other UK regions proposing to produce enhanced annual reports instead.

There was no standard approach to managing and organising evaluation. The regulatory position is that interim assessment is the responsibility of Monitoring Committees. The MEANS handbook is more specific: Monitoring Committees should establish an ‘evaluation mandate’ which defines operational responsibility for the evaluation, including budgetary resources, timing of the evaluation and the selection and positioning of the evaluation team; the evaluation mandate should incorporate a “degree of distance between programme managers and those carrying out the evaluation”; and Monitoring Committee members should “adopt an active approach to evaluation”: either by setting up an evaluation steering group, comprising external evaluation advisers, or by establishing an evaluation sub-committee drawn entirely from the Monitoring Committee.

In practice, Member States took different approaches to managing and organising evaluation studies. In France, central government recommended that all regions should establish specialist committees to supervise and guide the evaluation programme; these would be sub-committees of the Monitoring Committees. This recommendation was only partly adopted in Objective 2 regions and with variable success. A more ‘arms length’ approach was evident in Austria, where all four Objective 2 programmes were being evaluated by one evaluation team, and where a single evaluation working group was established under the chairmanship of ÖROK. The Danish and Finnish situations were similar, but with a more consultative role for the steering groups. In Finland, there was also one evaluation covering all the regions and a single evaluation steering group representing government ministries, Commission services and the evaluators. By contrast, the Objective 2 regions in Germany adopted a more direct management approach. The smaller regions undertook internal evaluation exercises, conducted by the Land government departments responsible for implementing the Funds. The larger Objective 2 programmes were evaluated externally.

287 MEANS (1995) Organising Intermediate Evaluation in the Context of Partnerships, Practical Advice for Programme Managers, MEANS Handbook No. 1, Commission of the European Communities - C3E, Lyon.

247

The aims and objectives of individual Member States and Objective 2 regions varied considerably, reflecting the political and administrative priorities of the programme management authorities. In practice, it was possible to identify six objectives, represented to varying degrees among Objective 2 interim evaluations:

��progress monitoring: the measurement and assessment of the outputs achieved so far;

��assessment of impact: a preliminary assessment of impacts where they could be evaluated;

��review of relevance: a reappraisal of the appropriateness of the programme and its component parts in relation to the objectives of the programme;

��assessment of management performance: an assessment of the efficiency of programme management, implementation and partnership arrangements;

��development of the evaluation framework: the derivation of indicators or the development of targets as preparation for the ex post evaluation study; and

��programme modification: proposals for adjustment or ‘corrective actions’ to the programme or individual elements in line with the evaluation results.

In addition, evaluation requirements incorporated a range of secondary concerns such as the environmental impact of the programme, synergies between funds and organisations, the impact on particular sectors or the performance of specific management organisations.

Evaluation studies ranged from data collection exercises to comprehensive assessments of programme progress and performance. Among the more substantive studies were the interim evaluations conducted in Austria. As the first attempt to analyse how EU funding was being implemented in the Objective 2 regions of Niederösterreich, Steiermark, Oberösterreich and Vorarlberg, a one-year study was conducted encompassing seven ‘evaluation blocks’: (i) the development of targets and indicators for the programmes, including the definition of indicator tables for individual measures; (ii) an analysis of monitoring system financial and physical output data to assess effectiveness and efficiency; (iii) an analysis of qualitative and quantitative environmental effects; (iv) an assessment of operational practice, including programme management and organisational arrangements for project delivery, the synergy between funds and between SPDs and Community Initiatives; (v) an assessment of the strategic context, notably coherence with national and EU policies, as well as relevance to regional socio-economic developments; (vi) the benchmarking of Austrian Objective 2 programmes with those in other European regions; and (vii) proposals for programme reform and adaptation.

A similar substantial evaluation was undertaken for three northern Objective 2 regions in Sweden. The study of Ångermanlandskusten, Bergslagen and Norra Norrlandskusten included: a background analysis of regional baseline data; a process analysis of implementation; project-level questionnaire surveys and case studies of best practice; an analysis of impact at SPD level and project level with a view to identifying total impacts on new jobs,

248

entrepreneurial climate, regional growth and new firm formation; and recommendations for programme development.

Many programmes concentrated heavily on evaluating the process of Structural Fund implementation and management, particularly in regions still establishing implementation structures, where new management arrangements were required, or where the timing of the study meant that outcomes could have little influence on policy. The Benelux, Austrian, Swedish and Danish evaluations, for example, emphasised the development of indicators to improve the evaluation of impact at a later date; and a Dutch project established harmonised indicators between the five Dutch Objective 2 regions to facilitate inter-regional comparisons of progress. The Western Scotland study also concentrated on process and management issues. In addressing issues such as absorption, the payments system and monitoring system, it included a partnership survey of 182 local and regional organisations with experience of the programme.

In Finland, evaluation concentrated on administrative and monitoring issues partly because of the major problems experienced with programme implementation: slow transfers of funds from central government to the regions; multiple layers and slowness of decision-making; and unsystematic information processing and storage. A broader assessment of outputs and impacts was impeded because the programmes were still young, but also because the monitoring system was not yet fully operational, delaying the production of statistics and indicator data.

A particular characteristic of the Danish evaluation was its thematic emphasis, arising from the strong interest on the part of local partners to focus on issues of particular concern to the programme partnership. The interim evaluation exercise was divided into two parts - a general evaluation common to both Objective 2 regions (Nordjylland and Lolland) and a thematic element specific to each programme. In Nordjylland, the evaluation focused on two themes: (i) the private sector target group of manufacturing enterprises which contribute to internationalisation; and (ii) support for large and small enterprises, especially SMEs. In each case, the evaluation was to determine whether these programme priorities were appropriate in the light of long-term objectives of growth in production and employment. This approach also characterised some of the French evaluations. The regional development agency, DATAR prepared a Vade Mecum of the Structural Funds which termed interim evaluation as ‘évaluation globale’ to distinguish it from the thematic studies that were also being promoted as a secondary stage. Among other countries, the Objective 2 evaluation in Austria also had a thematic element, focusing particularly on technological infrastructure and the new system of regional management offices introduced to mobilise awareness of Structural Fund opportunities and to assist with project applications.

An alternative to in-depth evaluation was the broader approach undertaken in Nordrhein Westfalen. The study was given the remit of looking beyond the 1994-96 period to examine the long-term results of earlier programming periods. As in Austria, evaluators were asked to make comparisons between the NRW programme and other Objective 2 programmes, to consider their programme components and implementation strategies, with a view to

249

presenting the broadest possible selection of alternatives potentially transferable to Nordrhein Westfalen. The report also considered the effects of the RESIDER, RECHAR and SME Community Initiatives and included an assessment of the effects of all Federal and Land programmes in the Objective 2 areas, especially their effects on diversification and job creation.

In retrospect, the interim evaluation of Objective 2 programmes contributed to a significant increase in awareness and understanding of the value, purpose and conduct of evaluation among national, regional and local actors. There was a considerable improvement in the scale and quality of evaluation being undertaken, and regions were able to demonstrate that effective structures for co-operation in evaluation were being built. The increase in actors with a knowledge and understanding of evaluation improved the ability both to deal with the evaluation exercise itself, and to benefit from it, ultimately improving the quality of the programmes.

There was no obligation for regions to act on the results of the evaluations, but there are several examples of evaluation conclusions leading to improved programme management and implementation. Among several Dutch, French and Danish Objective 2 programmes, for example, the interim evaluation was conducted early enough to influence the strategic focus and structure of the 1997-99 programmes, including modifications to the strategic orientation of programmes, and the reorganisation of problematic priorities and measures. A further important effect of interim evaluation was to improve the credibility of programmes with the wider regional partnership; the involvement of partners in the exercise provided a conduit for the exchange of information, which might not have occurred otherwise. A related issue was the effect on project selection procedures, awareness and publicity leading to applicants becoming more rigorous in preparing the rationale, structure and content of project proposals, in line with programme aims. Lastly, a common conclusion from interim evaluation studies was the need for monitoring and evaluation systems to be improved. Evaluation studies led to the establishment of new monitoring systems, the identification of new impact indicators and the collation of improved monitoring data.

9.5 Conclusions and Recommendations It is evident from Agenda 2000 and the proposed new regulations that the monitoring and evaluation of the impact of programmes will continue to have an increasingly high profile and that the performance of programmes will come under more intense scrutiny. The Commission will be emphasising the importance of employment creation, among other strategic priorities, during the programme design phase and supervising compliance more closely. It is inevitable that the improved quantification and measurement will be central concerns in the definition of monitoring indicators, annual reporting on implementation and the evaluation of performance. Programme mangers and partnerships will therefore have to dedicate greater resources to this area of programming.

There is now widespread appreciation in the Member States and Regions of the various ways in which monitoring and evaluation should be of benefit in the implementation of these programmes. First, the most fundamental benefit

250

of monitoring is that it provides hard data which backs up intuition for both tracking progress (important in view of time pressures on allocating and spending funds) and assisting in the development of solutions to problems. Second, monitoring helps programme managers develop more accurate project selection criteria and may also provide them with evidence for making a case to national or Commission officials with regard to adjusting a programme. Third, monitoring and evaluation should provide all the partners with the information which they need to contribute to the effective implementation of their programme; regular reporting of monitoring information and access to evaluation findings are required for this to be the case. Lastly, the steady tracking of progress (even if it is not dramatic) through regular reports not only boosts the morale of the partners, but also allows for the provision of accurate and up-to-date information to the media (important in many cases to promote absorption of assistance). It can also provide a public demonstration of the fairness of the management of the programme.

Regions have increasingly accepted the value of monitoring and evaluation. Despite continued misgivings about the frequency and timescale of evaluations, especially given the short programming periods, and the number of monitoring reports which have to be generated during programme implementation, positive commitment to monitoring and evaluation are now evident. Nevertheless, the developmental work has not yet been translated widely into practice. From a Commission perspective, programme monitoring and evaluation remain inadequate, with systematic monitoring procedures limited to financial indicators and little systematic tracking of physical outputs and impacts or verification of data collected from implementers; it is questionable whether programming authorities have the expertise or resources to undertake efficient project monitoring.288

Monitoring systems have a primary function of financial monitoring but can also help to collate physical output information from which employment outcomes can be deducted and to record the most immediate direct impacts. The potential of monitoring systems to fulfil these roles is not yet being fully exploited; they could be oriented to provide more useful information on employment in particular. If impacts are to be maximised, this priority needs to be followed through into all stages of programming from initial development to final evaluation. Much progress has been made, but most Member States and programming authorities are still just getting to grips with the field, which poses considerable planning, information, resource and skills implications.

As with the programming approach extending into other national and regional policy areas, the evaluation culture is filtering through into non-Structural Fund activities in some regions. The impact or applicability of evaluation appears to be related partly to ownership of the evaluation process. Where ownership is local/regional, reports appear to be more intensively exploited. As regions become more actively involved in evaluation, and better at specifying terms of reference and commissioning and supervising studies, the usefulness of evaluations to programme administrations has improved. One of the major benefits arising from the regular evaluation exercises has been the

288 CEC (1997) op. cit.

251

improved credibility of programmes within wider regional partnerships. The participation of partners in the evaluation exercises themselves has had a positive impact, improving information channels and heightening awareness about the programmes within the regions.289

The future of evaluation has improved but is still not yet secure.290 There is a danger of increasing hostility from those opposed to the perceived ‘excessive demands’ of the Commission and the regulations. Future development of evaluation will require greater focus, relevance and continuity in evaluation requirements.

9.5.1 Monitoring Recommendations

Most programmes have achieved good financial monitoring, but still suffer from limited ongoing output and impact monitoring beyond minimum obligations. In future, the more systematic tracking of project activities, outputs and impacts should be prioritised.

The need to demonstrate programme efficiency and effectiveness in order to compete for the Performance Reserve Fund may ensure that, in future, greater priority is placed on wider monitoring activities. Among the factors which have inhibited their development in the past have been the lack of resources to develop or maintain such systems, their low political priority and, where such practices are theoretically possible, the treatment of indicator updates as an optional, or certainly non-compulsory, element of project reporting.

The most highly evolved approaches to monitoring have arisen from a philosophy of integrating monitoring and evaluation as a continuous and cumulative dimension of all programming stages from programme development to final evaluation. In-built feedback loops can inform improved programme delivery and help to refine indicators, quantification and monitoring and evaluation processes themselves.

Adopting a cumulative approach to building a monitoring system is effective in a frequently changing policy context in which there may, initially at least, be limited knowledge and experience of monitoring. It enables a system to be established and then progressively refined as knowledge and expertise grows and needs evolve.

To underline how monitoring systems can be strengthened and refined at every programming stage, the checklist below sets out the key tasks to be designed and undertaken at each stage of the programming cycle and highlights best practice ideas in each. Note the benefits from undertaking a thorough preparatory stage during which realistic and meaningful targets and workable data collection and collation processes are established, the significance of

289 Taylor S (1998) op. cit. 290 Bachtler J, Michie R and Downes R (1997) op. cit.

252

feedback loops in enabling refinements to the system, and the value of prioritising pragmatic and practical responses balancing aspirations with available resources.

Investing in improvements to the quality of monitoring data generated at the project level benefits the programme as a whole by ensuring that the programme monitoring and evaluation system has firmer foundations.

Tools and advice are required to help applicants identify potential project outputs and impacts, to set targets and to measure achievements. Such investments help implementers to think through their proposals more fully at an early stage and improves the quality of information they generate. Programmes could consider providing a simple spreadsheet form to project implementers, to provide them with a consistent framework within which to record project activities and outcomes, rather than expecting each applicant to establish their own system from scratch.

If the provision of environmental and gender-related monitoring data by project implementers is made compulsory, it would help to ensure that programme commitments to the environment and equal opportunities are followed through at the project level.

If project implementers claiming tranches of funding fail to provide environmental and gender-related monitoring data, this should be questioned by monitors, (who could return monitoring forms to implementers as incomplete). The issues would then come to be perceived externally as integral, established parts of the programme, rather than add-ons. In turn, this would influence the priority given to these dimensions in project implementation.

253

9.5.2 Checklist of Monitoring Tasks throughout the Programming Cycle

Programme preparation

�Establish definitions for outputs, results and impacts. Think through how they will be identified and quantified.

Establish definitions and methods early, perhaps bringing in specialists to ensure the quality of frameworks. This will structure subsequent responses, and increase the utility of information generated.

�Establish assumptions enabling realistic targets to be calculated by collating existing information about the outputs and impacts of past projects per MECU spent.

Assumptions do not have to be based on complex studies. Some programmes have undertaken rapid telephone surveys to compile the necessary data. These assumptions can be refined later on.

�Ensure the programme has the necessary expertise and resources to respond to the implications of pursuing and quantifying effects.

Recruit or train a specialist in monitoring issues (ideally someone with a wider monitoring or project development role). Expertise can be ‘bought in’ on an ad hoc basis, but in-house expertise integrates the issues more effectively into day-to-day programming.

Familiarise a wider group of programme actors with terms, methods and issues and ensure that they can apply the frameworks.

Programme development

�Identify how each measure type affects outcomes to facilitate an informed debate about policy choices. Work from the volume of physical outputs to the numerous likely short and longer term effects.

�Consider the likely impact of deadweight and displacement. These can reduce net impacts, making some measures less good value for money.

Systematic comparisons of the potential (eg. job creation) of different measures at the programme development stage are rare, yet this could derive a policy mix with greater potential.

Widen the scope of the policy debate to innovative policies which could develop new effects (eg. sources of employment).

There may be good reasons to pursue measures with high displacement – eg. entrepreneurship schemes for the long-term unemployed which have a social rationale. However, these trade-offs should be explicit.

�Quantify ex ante targets once the programme has been decided.

�Use the definitions from the preparatory stage in setting targets.

During ex ante quantification, focus on deriving reliable gross impact estimates – eg. using prior performance information. These are essential, while net predictions are only desirable at this stage.

Specify the assumptions on which targets are based and state the limitations of the data used. This provides a stronger framework for critical monitoring and subsequent evaluation.

Project generation �Inform applicants about the programme’s emphases.

Promote the programme message vigorously through the SPD itself, promotional materials, seminars, written guidance, ad hoc advice etc. This should place strong emphasis on employment outcomes, but should also highlight horizontal themes, notably sustainable development and gender mainstreaming.

�Then provide specific guidance to applicants on identifying and quantifying targets for their projects.

�Draw on the standard terminology and approaches developed at the ex ante stage. This will help to generate consistent information.

Use discussions with applicants to help minimise deadweight and displacement by establishing that (a) projects are genuinely dependent on Structural Fund support to proceed, and (b) will not displace other activities (eg. employment).

The more thoroughly applicants have thought through effects at an early stage, the more likely they are to track progress and report on it accurately.

254

Project appraisal and selection �Appraise projects in a rounded way for their likely effects. This may require a structured checklist.

�Use the established frameworks to assess the realism of targets. If targets are unrealistic, revise them using the agreed assumptions so that projects can be compared fairly.

Applying the selection criteria in a rounded way involves considering the wider implications of each project, eg. not just the number of jobs they promise. This should • = also take into account the quality of jobs created

(sustainability, skills, sectors, appropriateness to target populations);

• = recognise projects with modest employment outcomes but which will help realise other programme priorities (eg. internationalisation);

• = eliminate projects whose employment effects are cancelled out by displacement, except where there are other reasons to pursue them; and

• = avoid projects with high deadweight (which would have been undertaken with less or no Structural Fund support - here, no real employment impacts can be attributed to the Structural Funds).

Monitoring project implementation

�Establish and maintain a monitoring system to record not only financial but also output and impact data (insofar as this is possible in the timescale available for monitoring).

�Use systems operationally

�Make systems available to partners

The ability of monitoring systems to track project outputs and impacts can be verified using a basic checklist: Is information requested from project implementers about outputs and immediate impacts? Is it returned consistently? Is it reliable? Can it be manipulated? Is it collected over as long a period as possible?

Good advice to implementers at the application stage prepares them for measuring and reporting back on outcomes at the implementation stage, leading to a better quality foundation for evaluation.

To ensure that the issues are followed through into projects, it should be made clear that the provision of environmental and gender-related information is obligatory.

Evaluation �Use the targets set at the ex ante stage, and the assumptions on which they were based, as the framework to assess programme performance. Exploit monitoring information on physical outputs and economic impacts as the starting point for the assessment.

� Use the evaluation to produce insights feeding back into effective programming.

Evaluation committees which are well-informed on evaluating impacts are better able to ensure the quality and relevance of any studies undertaken.

To ensure specialist input of sufficient quality, it can be advisable to award large, broad-based evaluations with specialised components to a series of evaluators, each with appropriate skills in specific fields.

255

9.5.3 Evaluation Recommendations

Evaluation is now an established part of the Structural Funds environment, but there is some way to go before a robust and self-sustaining ‘culture of evaluation’ is embedded. At this stage, a more strategic, co-ordinated and integrated approach to evaluation is recommended.

A longer-term approach to evaluation is already being taken in some regions or countries, with the establishment of evaluation frameworks which provide links between evaluation studies, strengthens the indicators on which studies are based, ensures quality control and systematises dissemination. Such frameworks view evaluation as an on-going, regularly updated process rather than a series of unrelated, one-off exercises. This also helps to reduce the perceived evaluation burden and to integrate the practice into the programming process.

To embed evaluation into programming, it can be beneficial to lay the foundations for it as early as possible in the programming cycle.

The UK has embedded evaluation more fully into policymaking routines by piloting a process whose acronym is ‘ROAME’, in which five elements are required to be elaborated before policy finalisation: policy Rationale, Objectives, project Appraisal methods, and Monitoring and Evaluation arrangements.

There is a need for more systematic dissemination and follow-up of evaluation findings, closing the feedback loop.

Evaluations only achieve their potential if there is greater dissemination and use of the results. Regions must find a balance between transparency and confidentiality, using different documents and different media to reach different types of audience, and instituting a phased dissemination process.

Programmes should be encouraged to act on evaluation results by making the process more relevant in terms of substance and timing but also by ensuring rapid Commission approval of changes arising from evaluation studies.

256

Engendering a sense of ‘ownership’ for evaluation studies at the programme level, through active steering and wide partner participation helps the activity to be embraced more seriously, and makes it more likely that any recommendations will be acted on.

Engendering a sense of ‘ownership’ of an evaluation process at the programme level can be facilitated by several factors. How a study is designed, commissioned, steered and funded are all important issues in determining how programme actors respond. Where responsibility for studies can be devolved to the programme level (as in many of the Objective 2 interim evaluations) this has a strong influence. Where this is not done, all other organisations commissioning studies, including at EC and Member State level, should be sensitive to the need to involve programme-level partnerships at all stages as stakeholders in the evaluation process rather than as the objects of evaluation.

The ‘style’ of evaluation is a further key dimension. ‘Formative’ evaluations are a relevant option for ensuring that evaluations actually impact on programming, with evaluators acting as ‘process consultants in learning exercises involving all relevant stakeholders, providing advice and recommendations as well as independent analysis’.291 This makes consultants facilitators, helping programme actors actively to reflect, exchange and learn from their experiences.

The number of evaluations could be rationalised – perhaps to one major study per programming period.

The diet of evaluations could be rationalised to one major in-depth study per programming period. This would be timed to feed into the ex ante evaluation for the next programme, the interim evaluation of the current programme and the ex post evaluation of the previous one. There is a large-scale precedent for such an approach in, for example, the first five-year assessment of the Framework Programmes for science and technology, which was an ex post of the third, interim of the fourth and intended to generate lessons for the fifth.292 This approach promises not only to be more efficient but may also improve the standing of evaluation at programme level, by placing greater emphasis on the primary relevance of evaluation to programme actors – to improve current and future systems by better understanding past activity. Lessons from the Framework Programme evaluation study indicate that timing is critical to ensure that outcomes can influence decisions on future programming.

291 Guy K et al (1998) Strategic Options for the Evaluation of the R&D Programmes of the European Union, Study for the Scientific and Technological Options Assessment Group, European Parliament DG for Research (available from: www.technopolis.co.uk/reports/stoa/title.htm). 292 Ibid.

257

Beyond the obligatory scope of evaluation, the practice could be called upon more, either independently or within wider studies, to address issues where there is a particular need for additional information or improved practices.

Evaluations could more frequently address process or thematic issues of particular concern, rectify information deficits or analyse specific policies at priority or measure level where absorption or impact is of particular interest to the programme partners. Issue driven, such evaluations can be particularly valued by programme actors.

259

10. SUMMARY AND CONCLUSIONS

10.1 Introduction With agreement on Agenda 2000 at the Berlin Council, the shape of the Structural Funds budget for the 2000-06 period is now clear. A reformed Objective 2 will encompass both regions undergoing industrial reconversion as well as those experiencing rural underdevelopment. The map of Objective 2 areas will include a mix of industrial, rural, urban, and fisheries areas. A simplified and decentralised approach to programming will place greater responsibility for programme development and management in the hands of Member State authorities but with stricter obligations for monitoring and accountability of performance. Strategies will need to give increased emphasis to sustainability, equal opportunities, competitiveness and job creation.

In preparing for the new programming period, a particular concern is that Objective 2 regions build on the lessons and experiences of the past. The aim of this study has been to provide a definitive assessment of the achievements of Objective 2 interventions under the Structural Funds as input for the implementation of future programmes. As noted in the Introduction, the distinctive nature of the study is that it focuses primarily on the process of Objective 2 programming, examining both commonalities and contrasts between countries and regions as well as changes over time. In doing so, it provides a better understanding of the long-term evolution of programme interventions addressing industrial reconversion, especially the progressive improvements in plan preparation, strategy development, programme management, partnership, programme delivery, monitoring and evaluation.

In assessing the experience of Objective 2 over the decade 1989-99, the project has had five specific objectives:

��to review Objective 2 reconversion strategies supporting industrial restructuring;

��to analyse the main strategic options deployed in the Objective 2 programmes to support specific horizontal themes;

��to assess issues relating to programme management;

��to examine the socio-economic impact of Objective 2 interventions; and

��to derive policy implications for future programmes.

The study has drawn on a mix of published information in the form of Commission papers, evaluation studies, programme documentation and academic literature, as well as research undertaken by the EPRC under various research programmes, in particular the IQ-Net project (Improving the Quality of Programme Management in Objective 2 Regions through Exchange of Experience) which involves a network of Objective 2 regions from seven Member States, together with the European Commission.

The eight core chapters of this study have examined, in detail, the institutional arrangements for implementing Objective 2 programmes in the relevant 12 Member States as well as the relationships between national and EU regional policy structures and trends in programme management, partnership and

260

programme delivery. The study has traced the evolution of Objective 2 regional development plans over successive programming periods – 1989-91, 1992-93, 1994-96, 1997-99 – highlighting the changes in presentation, structure and content of programmes. In a series of four chapters, it has also considered how the ‘horizontal priorities’ – job creation, RTDI, environment and equal opportunities - have been addressed in the design and implementation of programmes. Lastly, the evolution of monitoring and evaluation over the past decade has been reviewed.

This final chapter draws together the main results of the study, highlighting conclusions and some key recommendations for future programming of Objective 2 interventions. Further detail on specific recommendations can be found in relevant chapters.

10.2 Effective Programme Management and Delivery The study has discussed in detail the complexity of institutional arrangements for managing Objective 2 interventions. Each Member State has a different approach to implementation, reflecting different structures of territorial administration, the division of powers and policy responsibilities between national and sub-national authorities, and the relationship between Structural Fund and national regional policies. Furthermore, there is considerable regional variation within countries, and the regional divergence appears to be growing.

It is a feature of virtually all EU Member States that established government authorities - national governments in most cases, regional authorities in federal countries - play a dominant role in managing Structural Fund implementation. Indeed, most Member States have sought to limit the impact of EU regional policy on their systems of governance. Whereas some countries consider EU regional policy to be complementary to their own national policies, others have sought to restrict the scope of the Funds. In certain cases, EU regional expenditure has been channelled rigidly through the existing funding circuits of regional policy, while elsewhere the Structural Funds and national regional policy have been kept completely separate. An interesting point is that, over time, there has been some evidence of convergence between EU and Member State regional policies, although more in terms of procedural adaptations, such as changes in eligibility conditions for financial incentive schemes, than direct effects on national administrative structures and systems.

One of the main causes of problems for Member States arises from different actors (at national, regional and local levels) working at different stages of the implementation process, often with conflicting goals and agendas, and with differing relationships with Commission services. In this context, the level of sub-national involvement in programme management is a hotly contested issue. Lower levels of government have often perceived the Structural Funds as an opportunity to gain more autonomy with regard to their economic development responsibilities and resources (especially in unitary states), and this has increased the awareness and involvement of regional and local authorities with respect to strategic planning and development.

Over time, the Structural Funds have contributed to a change in attitude and structures. In some regions, participation in Structural Fund programmes has

261

led to deconcentrated services of the State, formerly accustomed to operating independently in accordance with their own discrete remits, becoming more conscious of how their actions fit into an overall regional strategic context. This applies to the Government Offices in the UK, Regional Prefects in France, CABs in Sweden, and to the District Offices in Finland. Rationalisation of central government administration with greater integration of (or co-operation between) the regional offices of sectoral ministries is also apparent, as in England and Finland. Elsewhere, implementation of Fund administration has strengthened the devolution of economic development, for instance to the Provinces in the Netherlands, the Regional Councils in Finland, and the Regions in Italy. A similar process is evident at sub-regional level in Germany with the creation of 12 ‘regional conferences/committees’ in Nordrhein-Westfalen. Novel institutional arrangements have been established for collective participation in strategy-making, implementation and monitoring. Of particular note is the use of independent ‘programme management executives’ - pioneered in the Western Scotland Objective 2 area but now being used in other Objective 2 regions in Scotland, Wales and Sweden (and currently under consideration in other Member States). In Denmark, political interests have been actively incorporated into Fund management in the Nordjylland Objective 2 region by creating a Regional Steering Committee to ensure political backing for development issues and to co-ordinate Objective 2 with other economic development initiatives. In Austria, a network of Regional Management Offices has been formed to support Structural Fund implementation at the local level, encouraging information dissemination, project generation and co-operation, especially among municipalities.

It is equally clear that there is still a considerable gap between the advocacy of regionalisation and partnership, outlined in Chapter 2, and the reality of regional development. The management of programmes continues to be dominated by the compartmentalisation of priorities and measures; substantial shares of programme expenditure are still allocated to government departments or agencies over which programme managers have limited influence and no control. This is not exclusively a problem of central government institutions; considerable funds are channelled through existing schemes of the sectoral departments of the Objective 2 Länder in Germany and Austria and the Assessorate of the Italian Regions with questionable strategic orientation and policy additionality.

With respect to the process of administration, in the early stages of programming, implementation of the Funds was characterised by management deficits at every level. At regional level, it was often anticipated that programme management could be accommodated within government departments or agencies, many of which had no experience or frame of reference for designing regional strategies or managing multiannual programmes. At Member State level, many national regional policy officials were often highly suspicious and critical of the policy objectives and implementation requirements of the Structural Funds and, in some cases, resisted complying with the principles of the Funds. In retrospect, officials will freely admit that it took them 4-5 years to appreciate the terminology and procedural requirements of the Funds and introduce appropriate management

262

and monitoring systems. Institutional inflexibility hindered integrated approaches to implementation. At Commission level, the management and resource capabilities of Commission services were inadequate to deal with the administrative complexity of the Funds, resulting in the familiar litany of complaints – endemic delays, complicated decision-making processes, bureaucracy, lack of co-operation between Directorate-Generals, and inconsistent advice.

Not all of these problems have disappeared. The management of the Funds has improved greatly – at all levels - over the past decade, but problems still remain. For example, there is still unwillingness in some Member States to accept that there is a different approach to regional policy from long-established national practice. This translates into resistance to adopt a positive, pro-active approach to implementing the Funds, and the attempted sidelining of concepts like partnership. Also, institutional inflexibility persists at Commission and national levels, most obviously evident in the difficulties of integrating ERDF and ESF in practice. Further, information flows are still inadequate, with considerable variation in information and awareness of policy and programming developments among regions. Information flows between Member States and regions, and between the Commission services and regions, differ from country to country. A fundamental problem faced by regions is how to ensure standard implementation of the Funds across the programme, and across government departments at central, regional and local levels in a manner that complies with the Structural Fund regulations and national administrative and accounting practices. There are inevitably great difficulties in maintaining a transparent overview of programmes, especially in decentralised/regionalised systems.

Initially, programme management was essentially department or agency based, each administering department operating programme measures through established structures and with little cross/inter-departmental liaison or communication. Early approaches to programme delivery were opportunistic, reflecting the approach to programme management. The perception among partners of the programme as a vehicle for drawing down funding was shared among beneficiaries who saw the programme in terms of ‘grant opportunities’. The need to commit resources rapidly, following a delayed start to the CSFs/OPs led to a widespread tendency to take projects ‘off the shelf’.

Implementation systems have matured significantly. Progressively programme delivery has become more pro-active, attempting to take a more integrated, developmental approach in the generation and selection of projects. During the 1994-96 programming periods, programme administration moved from a ‘list of projects’ approach to a more systematic and co-ordinated method of allocating resources. The programming concept became more firmly embedded: the strategic management of multi-annual programmes began to be accepted as more conducive to efficient and effective planning. Pre-application advice is being provided by increasing numbers of programmes to improve the relevance and technical quality of project applications even before submission.

In the best cases, programme delivery is moving towards an holistic, strategic approach along the lines of the ‘virtuous circle of programme project delivery’

263

(see Chapter 2). Faced with increasing numbers of applications and, in some cases, decreasing resources, the processing of applications has become more professional and systematic. In particular, formal and systematic appraisal and selection procedures have been introduced by many programmes, with some developing sophisticated methods for scoring, weighting and ranking projects, or involving independent experts and/or specialist panels in the selection process. Lastly, effective computerised management information systems have been introduced which assist in tracking projects from initial proposal submission to completion.

10.2.1 Programme Management Recommendations ��More effort needs to be made to promote attitudinal change within

Member State authorities, in particular to encourage flexibility and new thinking in strategic approaches to industrial change.

��The key to successful programme management is to understand the inter-relationship between the different elements of the programming cycle and to manage them as an integrated ‘virtuous circle’.

��In order to improve programme management, information flows, especially about required programming practice, need to be improved at all levels and made more consistent.

��Successful Structural Fund programming is increasingly demanding in terms of resources and expertise. Drawing on existing practice, a more explicit debate is required at all levels about how programme development and delivery should be organised and resourced and how those involved should be trained to meet the specific challenges of this changing environment.

��Better use needs to be made of the partnership resource. This means not just involving more partners but using them more effectively at appropriate stages of the programming cycle.

10.3 Preparing Plans and Strategies The presentation of regional development plans and Objective 2 strategies has progressed significantly over the past decade. Early plans had little or no strategic coherence, being viewed mainly as mechanisms for accessing EU funding. Analyses of stated needs were basic, and requirements for quantification or prior appraisal were neglected or ignored. Development objectives and axes were frequently general and vague and programmes sometimes seemed to be little more than selections of eligible projects.

Successive phases of programme preparation have seen a progressive increase in sophistication among Objective 2 strategies. Some of the best examples have systematic and in-depth socio-economic analyses (including SWOT assessments), and embody a clear, unifying mission statement, quantified objectives, specification of targets, use of benchmarking and a coherent set of measures, targeted at specific economic development problems and related to other national and regional initiatives, collectively supported by a monitoring system capable of determining physical outputs and impacts as well as financial progress. The links between programme elements – objectives,

264

priorities, measures – are becoming more explicit in the presentation of strategies, with the relationship of individual measures to overall programme goals being demonstrated more clearly

The Funds have clearly generated greater interest in strategic thinking and planning when this was on the decline or did not exist in some countries. This has been of particular relevance to Objective 2 regions, which typically have a greater number of agencies or initiatives operating across a range of sectors. The programmes have also widened the range of organisations able to exert an influence on economic development - now in many cases including the voluntary sector and trades unions in economic development thinking and decision-making. The Funds have required a more integrated approach to planning and implementing projects than has been typical in the field of economic development.

One of the major impacts of the Structural Funds in Objective 2 regions has been to encourage innovation in economic development. Examples of new ideas being disseminated through EU programmes include regional technology policies, community economic development and integrated environmental and economic measures. Perhaps the most important innovation promoted by the Structural Funds is in the field of community economic development. Some of the most deep-rooted economic and social problems lie in areas within urban/industrial agglomerations suffering from generational unemployment, poverty, and marginalisation or exclusion from labour markets and economic activity. Objective 2 interventions have been used for pilot programmes to enable excluded social groups to re-enter the labour market and to change attitudinal culture through community responses to skill development and provision of local services.

Virtually all Objective 2 regions have seen a shift away from promoting ‘traditional’ economic development activities such as physical infrastructure towards softer, more innovative measures. In the RTDI field, for example, investment in RTDI infrastructure is being superseded and complemented by a range of imaginative measures to exploit RTDI facilities through advice and consultancy support, enterprise and employee training, environmental technology promotion, network building and awareness raising. The promotion of technology and innovation policies appropriate to the needs of firms which are neither technologically oriented nor innovative has filled a particularly significant gap in the policy framework of many areas undergoing industrial reconversion (see Section 10.6 below).

Reflecting Commission pressure, environment and equal opportunities are becoming more prevalent within programme documentation. A shift towards integration of economic development and environment has been most necessary - and most evident - in Objective 2 areas where a clear momentum has been building up to strengthening environmental features of SPDs as a horizontal rather than compartmentalised feature of programme design and implementation (see 10.7). While full gender mainstreaming has not yet been achieved in Objective 2 programmes, commitment to equal opportunities is beginning to be in evidence, with responses ranging from the assessment of female employment opportunities within programme objectives, to specific measures or actions targeted at women (see Section 10.8 below).

265

Despite the improvements that have been made, there is still considerable scope for better programme preparation and strategy development, in particular for good practice to be spread more widely. First, the most challenging goal is for strategy development to be undertaken in a more holistic fashion, whereby the various elements of the strategy are developed as an integrated whole. Currently there is a widespread tendency to assemble Objective 2 strategies as a collection of ‘prefabricated units’, with the regional analysis, assessment of previous programme experience, strategy, environmental profile, prior appraisal, etc, being prepared separately and then collated within a programming document. Better interconnection between these building blocks, from the ‘bottom up’, would ensure better linkage between the analytical base and the strategy, the integration of sustainability and gender issues, and synergies between EDRF and ESF. Strategy development also needs to give more consideration to the conceptual basis for objectives and outcomes.

Second, the strategy development process needs to be seen as part of a continuum, linked to other parts of the programming cycle. Even in 1996-7, the mis-timing of evaluation studies meant that interim or ex post evaluations were rarely available for programme preparation. The content, form and exploitation of evaluation studies also needs to be organised to take account of reprogramming requirements, for instance targeting thematic or process issues of concern.

Third, the design, management and delivery of Objective 2 programmes needs to be integrated more fully into the regional economic development context, especially with the need (even at this stage) to look beyond 2006. The initiatives being taken in some regions to engage in strategic analysis and consultation about the wider regional context within which the new Objective 2 plan will fit, and complemented by diagnostic exercises and partner input about the programme itself, are to be recommended.

Lastly, while there have been long-term shifts in expenditure allocation among Objective 2 programmes, there is generally considerable continuity between successive programme periods. This is likely to be the case again in reprogramming from 1997-99 to 2000-06, but there is a strong case to promote new thinking and policy directions where this is justified. As this study has shown, an enormous reservoir of programming experience has been built up in Objective 2 regions over the past decade; more partners than ever before are engaged as experienced actors in different parts of the programme development process. It is important that regions harness this expertise and experience, and that they are prepared to initiate new or improved policy orientations. At the very least, the next programming period should be used for experimentation and testing new initiatives that may contribute to more effective strategies.

10.3.1 Programme Development Recommendations ��In developing their programmes, regions should be encouraged to exploit

the next programming period to create meaningful regional development strategies with scope for experimentation and innovation.

266

��The most challenging goal is for strategy development to be undertaken in a more holistic fashion, whereby the various elements of the strategy are developed as an integrated whole.

��The programme development process is too complex to be undertaken easily in the time available for it in the formal programming calendar. To resolve this, the process of strategy development needs to be seen as an ongoing process linked to other parts of the programming cycle.

��The design, management and delivery of Objective 2 programmes needs to be integrated more fully into the wider regional economic development context. In particular, this may become the only way to consistently achieve coherence with parallel human resource development programmes.

��As programme development becomes more inclusive at the regional level, the final programme negotiations need to become more sensitive to this, for example by being undertaken with greater transparency and consistency.

10.4 Making Partnerships work better Partnership has been one of the most disputed and difficult principles of the Funds, basically because it means people from different organisations and institutional cultures, with different interests, priorities, expectations and ambitions, having to co-operate and work together in pursuit of shared goals. The decade began with little scope for partnership, and more exclusion than inclusion. Early attempts at partnership were limited or token, and typically characterised by vertical and horizontal institutional tensions and the exclusion of key groups. Since then, the partnership principle has become embedded. In many cases it has moved beyond a formal, regulatory requirement to involve “a complex nexus of strategic and operational relationships…which represent a significant capacity for multi-organisational action”.293 This is not universally the case: the impact of partnership in some areas remains limited, especially where programmes are small, and where eligible areas are fragmented or transcend administrative boundaries. Nevertheless, it is possible to identify several distinct developments.

First, the membership of partnerships is expanding – they continue to be dominated by national government authorities (both at a central and territorially deconcentrated level) as managing authority and implementing agency, but a wide range of partners at the regional level is now routinely involved. The breadth and depth of partnerships is becoming more inclusive, with participation of elected members, social partners, and voluntary and community groups, although this is certainly not fully developed everywhere and is often only in a consultative capacity. Second, there is clear evidence that partnerships have significantly improved communication and co-ordination between actors at regional level. Especially in unitary states, collective regional strategic thinking, planning and co-ordination has been introduced for the first time or rediscovered. The Funds are promoting better

293 Stern E (1997) The ‘Partnership Principle’ in European Structural Funds, Report to the European Commission (DG XVI) Tavistock Institute, London.

267

contacts between the public and private sectors, economic development organisations, voluntary organisations and other bodies. Further, the boundaries between government ministries (and departments within ministries) are being broken down in the interests of effective implementation.

Third, it is no longer possible to view the ‘partnership’ as a single entity. Partnership has become multilayered. Special thematic groups and local, sub-partnerships are playing a more active role in drafting and implementing strategies. Reflecting the trend towards more sub-regional programming and thematic foci, a more targeted participation of meso-level partnership groupings is emerging in a limited number of regions to increase the quality of partnership input at the most effective points. Lastly, the influence of the partnership principle is spreading beyond Structural Fund programmes into other areas of national and regional policy; the Growth Agreements being introduced in Sweden are one of the recent examples.

Looking to the next programming period, there are several challenges. Improved approaches to partnership management are needed, learning from the ‘good practice’ that exists, to involve more partners more actively at more stages in the programming process where their input could improve the quality of programming. The key issue in assessing partnerships is the extent of ‘strategic capacity’ - whether they make decisions or rubber-stamp them. Partners tend to be principally involved either in the early stages, consulted on the drafting of the SPD, or later by participation in programme-wide decision-making and strategic committees, or as applicants themselves. The optimum potential for partner participation might involve more or different phases of involvement. However, with an ever wider group of partners involved in Structural Fund implementation, the practical functioning of partnerships has become increasingly problematic, especially as more and smaller local organisations, with less familiarity with the Funds, are involved. Broader partnerships are more representative but more difficult to manage, making it advisable for programme authorities to identify which partners, or groups of partners, can most effectively contribute which functions at what stage of the programming process.

A key question is how to establish and maintain coherent and effective partnerships, working strategically to shared aims, focusing on strategy rather than projects. Evaluations frequently report a strong feeling of an ‘inner core’ of major partners whose organisational and policy interests determine the shape and direction of programmes, supplemented by an outer circle of (more numerous) minor players. Inevitably, the remit and resources of some partners afford them more influence. There is therefore the risk of a programme’s ‘committee community’ effectively becoming an ‘exclusive club’ which could alienate key groups such as the social partners. This underlines the importance of having open proceedings and publicising them widely.

Not all partnership experience has been productive: in some regions partners have left the process through misunderstanding or frustration at their inability to access resources or influence strategies. Private and voluntary sector bodies often have difficulty with the complexity of the process, especially at local level; there is a need for more transparent and targeted information to enable them to be involved more easily. A further challenge has been engaging

268

partners in reflection on the strategy as well as projects. For many regions an important issue is the lack of interest in participation among relevant actors and the wider community; the operations of programme managers and the proceedings of partnership committee meetings may be open but poorly supported and attended. Press releases addressing the wider public and the business community may also not be taken up. This raises the important question of how to promote awareness, interest and participation.

In institutionalising partnership, the common feature of all programmes is the establishment of a Monitoring Committee. As noted earlier, some countries operate a single national committee representing all Objective 2 regions, while other states have created an MC for each Objective 2 programme. A frequent complaint is that MC meetings do not invite open discussion on strategic direction because they are dominated by technical issues. The technical character of meetings also suppresses the input of some partners (eg. political figures, minor partners) as they are intimidated by the process, lack understanding of the issues and feel they have no part to play. Solutions advocated to promote a higher quality and different style of debate include the separation of strategic issues from financial/technical matters eg. between MCs and Programme Management Committees.

10.5 Creating Jobs Objective 2 areas are defined primarily on the basis of negative employment change. As such, employment creation has always been one of the main aims of these programmes, and the primary indicator for assessing their achievements. In every reprogramming exercise since 1989, the emphasis placed on job creation has grown steadily. This objective has become increasingly prominent as high-level political commitment to addressing persistent and worsening European joblessness at EU and Member State levels has grown. This commitment was reflected in: the so-called ‘Essen Track’ proposals which called on Member States to take action in defined fields to increase employment; a major new chapter of the Amsterdam Treaty and, most recently; and, in the introduction of National Action Plans for Employment, built around the four pillars of employability, entrepreneurship, adaptability and equal opportunities, which further develop the Essen employment programmes. By the time Objective 2 programmes were being developed for the 1997-99 period, job creation had become the overriding Commission priority.

Over the last decade, the content of Objective 2 programmes has been adapted in three main ways to respond to the employment imperative. First, their policy content has evolved. While the overall aim of Objective 2 programmes is to improve the employment situation, most have moved towards balanced combinations of interventions which will help to achieve three separate employment-related objectives: to establish the conditions for long-term economic growth by raising competitiveness, to increase the job intensity of growth and to take a pro-active approach, where appropriate, to social solidarity.

In terms of specific policy orientations, investments in basic infrastructure, which offer relatively poor value for money either in terms of permanent

269

employment created or impacts on competitiveness, have been reduced over time in favour of measures releasing the potential of indigenous businesses by intervening directly in firms (especially SMEs) or - more durably - improving the collective business environment (eg. by establishing incubator units, technology brokers, etc). Business development schemes have also progressively been refined as experience has been gained to ensure value for money, eg. by requiring them to more clearly establish demand and to monitor supported firms more closely. The ESF by its nature creates relatively few jobs directly, but serves an important complementary role in getting people into the new jobs being created or preventing their passage into unemployment. ESF interventions have also been refined and adapted in response to the need to address unemployment in more focused ways, with increased resources being dedicated to tailoring human resource development to the needs of specific firms and individuals, and investments in labour market intelligence services ensuring that the skills and qualifications delivered under programmes are genuinely those in demand. Community economic development policies have also become more prevalent, responding to the need to consider not just the number but also the quality and distribution of jobs where there are concentrated problems of labour market exclusion.

The second response of Objective 2 programmes to the employment imperative is that they have almost universally made this theme more explicit. In programming documents, its importance is now spelt out in the regional profile, the strategy statement and, at priority and measure level, in rationales, selection criteria and impact targets. The most striking presentational changes took place between the 1994-96 and 1997-99 programmes as a result of Commission guidance. In addition, Commission level policy documents and initiatives have also helped to raise the profile of employment, and especially its role in Objective 2, for example through the ‘Jobs Challenge’, an international competition to distil transferable lessons from projects which have been particularly successful in creating employment.

The third response has been action to improve the ability of Objective 2 actors to predict and then demonstrate employment outcomes, using enhanced prior quantification and monitoring. Across the 1989-93 programmes, only one in five programming authorities included employment targets in programming documents. Coverage and quality have both improved successively since then, with each programme engaged in a process of learning and building on past experience, helped by the (albeit erratic) dissemination of methodological frameworks elaborated with the support of the EC. A small number of Member States still resist detailed prior quantification because of the uncertainties inevitably involved, but in general the principle is accepted and the necessary expertise to apply it is accumulating.

By the 1997-99 programmes, three-quarters of SPDs were providing detailed, measure level estimates of employment effects, the approaches taken to supplying them varying from relatively sophisticated, comprehensive exercises commissioned to address the issue across groups of programmes, to very pragmatic solutions addressing primary weaknesses in single programmes. In addition, the definitions used in setting out targets have become more consistent and robust. In a small number of instances, targets

270

have become more transparent, with SPDs setting out the assumptions on which they are based.

In addition to the three changes in the presentation and content of Objective 2 programmes, their delivery has also been adapted in an effort to ensure that employment impacts can be achieved and demonstrated: first, improved methods have been being used to communicate the jobs message to potential applicants; second, support has been given to applicants at the project development stage to help them to maximise potential employment impacts; and, third, monitoring systems have been improved to provide more feedback about the actual employment outcomes of projects and measures.

The need to improve the employment impacts of Structural Fund programmes will only grow in future. The following recommendations set out some principles by which they can respond.

10.5.1 Job Creation Recommendations ��All programmes should at this stage be required to achieve a threshold

level of quantification of targets, including for employment - both in terms of their quality and their coverage.

��In designing and implementing future programmes, employment outcomes should be prioritised, but in a balanced way, taking into account potential impacts over the short, medium and longer term and the quality, durability and distribution of the jobs to be created.

��The ERDF primarily addresses demand-side issues in the labour market, creating or safeguarding jobs. Programmes must continue to seek coherence with ESF interventions – whether they are under Objective 2 or operated in parallel – which respond to the supply side, by training people in appropriate skills to take up the vacancies being created.

��Programmes could significantly improve value for money in terms of their job creation outcomes by filtering out policies and projects likely to have limited net employment impacts because of high deadweight or displacement effects. They should place a premium on developing the necessary systems to do this.

��Further efforts should be made to equip programme applicants to respond to the jobs challenge.

��The employment-related lessons of monitoring and evaluation should feed continuously into improving all stages of programming.

10.6 Encouraging Innovation The RTDI content of Objective 2 programmes has evolved over the last decade, in line with growing awareness about the importance of technology transfer and innovation to regions seeking to regain competitive advantage, and deepening understanding of the barriers to establishing these dynamic, interactive processes. Across successive programming periods, rising expenditure has been dedicated to RTDI policies, across an increasingly diverse and sophisticated range of interventions. The policy orientation has changed, with the science and technology focus of early programmes being

271

replaced by a stronger emphasis on applied policies, targeted at raising the ability of regional firms to compete. Initiatives increasing demand for technological solutions to business problems have been complemented by policies helping existing regional scientific and research organisations to respond.

In addition, the balance of interventions being supported has changed, with a shift away from the focus on infrastructure which characterised the early programmes (science and technology parks, equipping of research-oriented laboratories), towards softer and more demand-side measures, including support services to assist businesses to identify and implement projects to raise their technology and innovation levels. ERDF policies have also increasingly routinely been complemented by human resource development measures – unusual among ESF interventions in that they tend to target those who are already highly qualified in relevant fields - helping employees to raise or adapt their technical skills, or introducing technical or research staff into firms. An element of policy which addresses a particular deficit of Objective 2 areas, and is receiving increasing attention, is the diffusion of established technologies to lagging firms. In addition, policies explicitly promoting the adoption and exploitation of new communication and information technologies are seeing strong growth.

The overall positive picture, however, masks variations between programmes in terms of the strength of their rationales for RTDI policies, the design and internal coherence of these policies and the quality of delivery systems (including selection criteria, project development support and monitoring indicators). The depth of diagnostic analysis could be increased in most programmes, and further efforts made to encourage projects that are mutually reinforcing and have genuine potential to raise regional innovation capacity - for example focusing efforts on specific clusters or fields of unique regional expertise.

A further recommendation is that deeper strategy development exercises be adopted on a more widespread basis. Among the realisations of the last decade with regard to policy design has been that RTDI-related regional policy is distinctive. It not only has to help accumulate the many components necessary to enable innovation and technological advancement, based around regional specificity and unique knowledge, but it also has to ‘breathe life into them’, establishing a self-sustaining dynamic of continual learning and change among all the actors involved. In order for policies to achieve such an impact, they require a deeper level of strategic reflection, focus and co-ordination than most policy types.

Detailed regional (and often sector) specific analyses are needed to identify weaknesses in the innovation milieu and design policies able to address them successfully and establish new and durable behaviours and systems. While Objective 2 programmes do facilitate the partner interactions which can drive this type of exercise, the compressed timetables and broad scope of Objective 2 programmes in many cases inhibit exercises of sufficient depth, focus and continuity – except where the context is particularly favourable. An important innovation by the Commission, which has facilitated dedicated strategy development and partnership building in innovation and technology transfer in

272

many Objective 2 areas, has been the launch of RTP, RIS and RITTS projects. These strategy-building projects have shown considerable potential in helping regions to address the complex issues surrounding low technology and innovation levels. In many of the relevant cases, they have significantly reinforced regional capacities to begin to establish the dynamics which might one day enable them to join Europe’s ‘islands of innovation’. The momentum gained by some of these strategic exercises, whose outcomes have been operationalised with the help of Objective 2 co-financing, is already such that they are likely to become durable elements of the economic development environment. Given the significance of RTDI to industrial regeneration, these exercises could be made a universal element of every Objective 2 programme.

10.6.1 RTDI Recommendations ��RTDI is a demanding policy field and the depth of diagnostic effort needs

to be improved across more Objective 2 programmes, to design and deliver more effective policies.

��The Objective 2 programme development timetable is not conducive to the necessary diagnostic and policy-making practices. Therefore, parallel, partnership-based RTDI strategy-building initiatives are highly recommended: the Commission’s RIS and RITTS studies have proved highly effective in many cases.

��Elements of RTDI policy are relevant to every Objective 2 area, and every programme should explore how they could be used. Policies should be tailored to address the specific characteristics of lagging firms as well as those with greatest innovative potential, and should include parallel human resource development measures for best effect.

��This is likely to be the last Objective 2 programme of this scale in most regions, and a priority should be placed on actions able to deliver durable, self-sustaining effects.

��Raising a region’s technology and innovation capacity requires a wide range of issues to be addressed through diverse policy instruments. Relevant policies should be organised and presented prominently in Structural Fund programmes, with a clear expression of policy priorities.

��Pro-active and inclusive approaches are required to generate sufficient good quality project proposals in the time available under Objective 2.

��RTDI project proposals are some of the most technical and difficult applications to appraise. Provision should routinely be made for specialist involvement in this task.

��The selection criteria used by most programmes could be improved to become more focused and specific. This could help to emphasise further the priority placed on regionally relevant projects for which there is proven demand and an applied economic rationale.

��The ability to capture the true effects of RTDI actions, particularly with regard to their impact on regional innovation capacity, is under-developed and requires further work.

273

��Programmes may benefit from being closely and supportively monitored, given the complex, innovative and unpredictable nature of RTDI projects, the opportunities for spin-off projects, and the intangible or long-term nature of their outputs/impacts.

��RTDI merits special provisions in programme-wide evaluations. ��Exchange of experience between programmes and with related initiatives

outside Objective 2 has considerable potential for the dissemination of good practices.

10.7 Promoting Sustainable Development In response to the EU's increasing orientation towards environment and sustainable development, Structural Fund programmes are progressively incorporating new features that support the aspirations of the Fifth Environmental Action Programme, Towards Sustainability. In parallel, the environmental impetus provided both by the revised Structural Funds Framework Regulations and the practical guidance issued through studies directed by the Commission has resulted in a much clearer focus on the means of realising environmental integration.

In reviewing the successive programming periods, three different orders of integration can be identified. First order integration relates to the stage in which programmes treat environment as a horizontal issue, in practice often meaning that environment receives no explicit priority, but rather is assumed to feature as a component of other (economic) priorities. One characteristic of dealing with environment in this manner is that external consultations carried out with relevant departments or agencies may mean that time constraints severely limit the level and usefulness of external input. In recent years, the majority of SPDs have fallen into this category. Some appear to be making considerable progress in overcoming hurdles to environmental integration.

Second order integration relates to the higher profile treatment of environment as a vertical component of the regional strategy. Only a small number of the reviewed programmes have these characteristics. Vertical integration means that the environmental objectives are expressed as a separate priority of the programme, given more status, and as a result monitored effectively. At this level, a committee of environmental experts might work as a programme advisory group making recommendations on project selection. Such approaches are more likely at this stage than the previous one, when the horizontal approach would typically result in an individual ‘environmentalist’ sitting as a member of other project advisory groups. There is also scope for a clear link to a specific environmental budget heading, which can assist implementation and project accountability. However, a disadvantage of this approach is that partners tend to assume that environmental issues relate only to projects within the environment priority: consequently, horizontal integration may lost.

The scenario of third order integration would comprise the adoption of a strategic approach to the exploitation of environmental issues with competitive purpose. Bringing together both horizontal and vertical integration, it would render environment a very significant element of the economic development strategy. This signifies a substantial change in the perception and role of

274

environment, potentially overseen by internal environmental expertise during the three phases of programme design, implementation and evaluation. Employing both qualitative and quantitative targets for environmental improvement, this programme orientation would set environmental excellence as a central objective, subsequently giving rise to environmental competitiveness. In the current round of programmes, this third level of integration is principally aspirational. Although several programmes appear to be moving towards this stage, none has yet progressed sufficiently to generate regional environmental competitiveness.

The following recommendations highlight practical ways in which Objective 2 programmes can more fully integrate the principles of sustainable development into regional development practice.

10.7.1 Sustainable Development Recommendations ��To achieve sustainable development, the environmental strand of

Structural Fund programmes needs to be viewed as integral to the programme, rather than a detached sub-component.

��Environmental factors should be followed through consistently and meaningfully at every stage of the programming cycle.

��Environmental profiles should be enhanced to improve their ability to inform and direct programmes, bearing in mind the business development focus of SPDs.

��Following through from the environmental profile, all priorities and measures (both mainstream and environment-focused) should state their environmental relevance, and set output targets for their value added to the environment.

��In integrating environment as a horizontal issue, priorities and measures must be designed in ways which facilitate environmental integration, so enabling mainstream economic projects to contribute towards wider sustainability goals.

��Consideration should be given to wider and more systematic use of Strategic Environmental Assessments at the programme development stage to derive useful early estimates of the potential environmental impact of programmes.

��Project selection and scoring systems should reward projects which meet environmental criteria. This element of project assessment needs to be consistent, understandable and justifiable for applicants.

��Monitoring environmental outcomes presents particular methodological problems. Nonetheless, the collection of even relatively basic indicator information can be extremely valuable – if it is undertaken systematically, and if it is seen as integral to the programming process and not a voluntary, one-off optional exercise.

��Implementing a ‘Cascade System’ for indicators is recommended for the more effective tracking of environmental outcomes. Here, indicators and targets which are meaningful at the overall programme level are

275

disaggregated into a number of indicators relevant to individual projects which combine to help to achieve the overall targets.

10.8 Encouraging Gender Mainstreaming Women and men in the EU, in assisted areas and beyond, persistently experience unequal access to and benefit from training and employment, both in terms of quantity and quality. In economic terms, this constitutes the sub-optimal utilisation of human resources – an issue of particular importance in economically disadvantaged regions. In order to release the potential of every individual as a ‘vital development agent’, policies – including those for economic development – need to be adapted to enable all groups to gain equal access to productive resources, opportunities and public services. The new policy approach of ‘gender mainstreaming’ provides a powerful tool to systematically screen all policies for their gender implications and impacts at the point of planning, implementation and evaluation, in order to ensure that opportunities are genuinely available to the spectrum of potential participants, and that initiatives have no negative gender impacts.

Recent changes in the wider policy environment, with political will being secured for gender mainstreaming at the highest levels of the EU, in tandem with the inclusion of an equal opportunities obligation in the Structural Funds regulations, have provided important stimuli for equal opportunities to be taken into account by Objective 2 programmes. From being an area of legislation to respect, equal opportunities has newly become, in theory at least, a universal, active dimension of Structural Fund programmes, pursued with the dual rationales of equity and efficiency.

Although in insufficient numbers and with insufficient depth, many current Objective 2 SPDs (1997-99 and, in Autria and Sweden, 1995–99) now contain some, limited, references to gender imbalances in the economic sphere and their potential to address them. This change is important, since the presence of even superficial references to equal opportunities in programming documents has helped to place the issue on the agenda, adding weight to it and opening the possibility for ongoing debate and development at the programme level.

In spite of recent progress, there is further to go: the equal opportunities demands made of Objective 2 programmes are considerable. This theme is expected to be integrated comprehensively as an element of the design, delivery and assessment of all policies, through ‘gender mainstreaming’, and, within this, addressed where relevant by positive action, ie. projects designed by specialist agencies to help remove the barriers to (usually) women’s full participation in the labour market. Both types of response to gender issues are new to most Objective 2 programmes, but gender mainstreaming is also new in the wider policy environment, meaning there are few precedents and little experience. It is not surprising that most responses to date have focused on positive action (frequently, entrepreneurship schemes targeting women, and provision to ensure access to training and employment by those with caring responsibilities).

The lack of targeted evaluation of the equal opportunities dimension of Objective 2 programmes means that little can be said definitively about the extent and impact of gender mainstreaming so far. However, a review of

276

current programmes illustrates that gender mainstreaming has not yet been fully embraced by any programme. Establishing the ‘gender reflex’, in which every action is routinely screened for its gender implications, will have to be a long-term objective. The low quality of many programme responses to date does not necessarily relate to a lack of commitment to the principle of equal opportunities, which is genuine in many cases, but rather to the absence of appropriate knowledge, skills, tools and mechanisms to enable more effective responses.

To move towards true mainstreaming, in which a gender filter would be applied to all policies and projects, assessing their likely uneven gender impacts and making appropriate adjustments, a wider awareness of what this term implies is needed among programme actors. Progress will require the Commission to sustain its emphasis in the next round – including through the provision of improved guidance which goes beyond individual examples of positive action to mainstreaming proper. Programme actors will also have to intensify their efforts. It is notable that those programmes which have made most progress in equal opportunities to date are those which have embraced the issue most actively as an integral part of programming.

10.8.1 Gender Mainstreaming Recommendations ��Gender mainstreaming is new. To secure more of the desired responses

earlier, it is recommended that quicker, wider and more effective dissemination of existing best practice and guidance is undertaken. This should prioritise practical information on gender mainstreaming itself rather than positive action as has been the case to date.

��Programming authorities should ensure that equal opportunities is embraced actively as an integral part of programme planning and implementation, and that responses are shaped to contribute to core programme objectives.

��A greater gender balance should actively be sought in decision-making structures, but ideally not through the imposition of male/female quotas.

��It will be important to equip all decision-makers, whether male or female, with the knowledge and skills to identify - and respond to - gender issues.

��One of the barriers to implementing gender mainstreaming is a lack of clarity about what this means in Structural Fund programming. Defining the anticipated scope of equal opportunities responses will help to clarify expectations.

��The programme development stage is critical to laying the foundations for gender mainstreaming. At this stage, the issue should be made more meaningful by undertaking a thorough diagnosis to identify concrete gender-related economic development issues – then designing responses to address them.

��In programme development, programmes should be obliged to include gender-disaggregated indicators and targets and a description of how gender issues have been and will be taken into account. In addition, before programmes are finalised, ‘gender filters’ should be applied to

277

policies to identify and address any potentially negative and unnecessary gender implications.

��A conscious, clear and planned communication effort is required to ensure that future programme applicants understand the meaning, importance and relevance of the concept of gender mainstreaming and their obligations in this field.

��Straightforward methodologies are needed to implement gender mainstreaming. A ‘gender filter’ has to be designed which can be applied by non-specialists to all projects at their development and then assessment stage to identify potential gender-related implications.

��Monitoring the outcomes of programmes by gender must become more widespread and systematic, and systems need to be adapted to deliver this. To facilitate evaluation, more base-lines should be established.

��Organisations commissioning Structural Fund evaluation studies should include gender analysis as a standard component. Dedicated gender studies should also be commissioned where there are information gaps.

10.9 Monitoring and Evaluation It is evident from Agenda 2000 and the proposed new regulations that the monitoring and evaluation of the impact of programmes will continue to take an increasingly high profile and that the performance of programmes will come under more intense scrutiny. The Commission will be emphasising the importance of employment creation, among other strategic priorities, during the programme design phase and supervising compliance more closely. It is inevitable that the improved quantification and measurement will be central concerns in the definition of monitoring indicators, annual reporting on implementation and the evaluation of performance. Programme mangers and partnerships will therefore have to continue dedicating greater resources to this area of programming.

There is now widespread appreciation in the Member States and regions of the various ways in which monitoring and evaluation should be of benefit in the implementation of these programmes. First, the most fundamental benefit of monitoring is that it provides hard data which backs up intuition for both tracking progress (important in view of the time pressures on allocating and spending funds) and assisting in the development of solutions to problems. Second, monitoring helps programme managers develop more accurate project selection criteria and may also provide them with evidence for making a case to national or Commission officials with regard to adjusting a programme. Third, monitoring and evaluation should provide all the partners with the information that they need to contribute to the effective implementation of their programme; regular reporting of monitoring information and access to evaluation findings are required for this to be the case. Lastly, the steady tracking of progress (even if it is not dramatic) through regular reports not only boosts the morale of the partners, but also allows for the provision of accurate and up-to-date information to the media (important in many cases to promote absorption of assistance). It can also provide a public demonstration of the fairness of the management of the programme.

278

Regions have increasingly accepted the value of monitoring and evaluation. Despite continued misgivings about the frequency and timescale of evaluations, especially given the short programming periods, and the number of monitoring reports which have to be generated during programme implementation, positive commitment to monitoring and evaluation are now evident. Nevertheless, the developmental work has not yet been translated widely into practice. From a Commission perspective, programme monitoring and evaluation remain inadequate, with systematic monitoring procedures limited to financial indicators and insufficiently widespread systematic tracking of physical outputs and impacts or verification of data collected from implementers; it is questionable whether programming authorities always have the expertise or resources to undertake efficient project monitoring.

Monitoring systems have a primary function of financial monitoring but can also help to collate physical output information from which employment outcomes can be deducted and to record the most immediate direct impacts. The potential of monitoring systems to fulfil these roles is not yet being fully exploited; they could be oriented to provide more useful information on employment in particular. If impacts are to be maximised, this priority needs to be followed through into all stages of programming from initial development to final evaluation. Much progress has been made, but most Member States and programming authorities are still just getting to grips with the field, which has considerable planning, information, resource and skills implications.

As with the programming approach extending into other national and regional policy areas, the evaluation culture is also filtering through into non-Structural Fund activities in some regions. The impact or applicability of evaluation appears to be related partly to ownership of the evaluation process. Where ownership is local/regional, as with many interim evaluations, reports appear to be more intensively exploited. As regions become more actively involved in evaluation, and better at specifying terms of reference and commissioning and supervising studies, the usefulness of evaluations to programme administrations has improved. One of the major benefits arising from the regular evaluation exercises has been the improved credibility of programmes within wider regional partnerships. The participation of partners in the evaluation exercises themselves has had a positive impact, improving information channels and heightening awareness about the programmes within the regions.

The future of evaluation has improved but is still not yet secure. There is a danger of increasing hostility from those opposed to the perceived ‘excessive demands’ of the Commission and the regulations. Future development of evaluation will require greater focus, relevance and continuity in evaluation requirements.

10.9.1 Monitoring Recommendations ��Most programmes have achieved good financial monitoring, but still suffer

from limited ongoing output and impact monitoring beyond minimum obligations. In future, the more systematic tracking of project activities, outputs and impacts should be prioritised.

279

��The most highly evolved approaches to monitoring have arisen from a philosophy of integrating monitoring and evaluation as a continuous and cumulative dimension of all programming stages from programme development to final evaluation. In-built feedback loops can inform improved programme delivery and help to refine indicators, quantification and monitoring and evaluation processes themselves.

��Section 9.5.2 of this report provides a checklist of monitoring tasks throughout the programming cycle, and highlights best practice approaches at each stage.

��Investing in improvements to the quality of monitoring data generated at the project level benefits the programme as a whole by ensuring that the programme monitoring and evaluation system has firmer foundations.

��If the provision of environmental and gender-related monitoring data by project implementers is made compulsory, it would help to ensure that programme commitments to the environment and equal opportunities are followed through at the project level.

10.9.2 Evaluation Recommendations ��Evaluation is now an established part of the Structural Funds

environment, but there is some way to go before a robust and self-sustaining ‘culture of evaluation’ is embedded. At this stage, a more strategic, co-ordinated and integrated approach to evaluation is recommended.

��To embed evaluation into programming, it can be beneficial to lay the foundations for it as early as possible in the programming cycle.

��There is a need for more systematic dissemination and follow-up of evaluation findings, closing the feedback loop.

��Engendering a sense of ‘ownership’ for evaluation studies at the programme level, through active steering and wide partner participation helps the activity to be embraced more seriously, and makes it more likely that any recommendations will be acted on.

��The number of evaluations could be rationalised – perhaps to one major study per programming period.

��Beyond the obligatory scope of evaluation, the practice could be called upon more, either independently or within wider studies, to address issues where there is a particular need for additional information or improved practices.

10.10 And finally… In all aspects of implementation discussed in this study, there is clear evidence of overall improvement in strategic direction, programme management, project delivery and partnership. But there are significant differences between programmes; so-called good practice is not always widespread, and many regions are still lagging in their programming arrangements. In the next programming period, programming will undoubtedly be still more demanding and complex. More specialisation and professionalisation will be necessary,

280

especially in view of forthcoming efforts to increase internal coherence between national and EU funding priorities.

A positive feature of the past decade of experience is that there is clear evidence of learning. The challenge is how to promote and enhance this learning process through training and dissemination. There is an enormous pool of experience available as a resource, and the aim must be to encourage the exchange of experience of Structural Fund implementation.

With decentralisation of implementation, there will be new pressures on management systems. The localisation of economic development initiatives creates challenges for effective programme management. Monitoring and evaluation will need to improve in response to pressures for decentralisation and sound and efficient management. There are considerable variations in monitoring capability between programmes; some smaller programmes appear to have little more than basic data. Programmes will have to undertake systematic monitoring of priorities and measures, monitoring not just in terms of jobs but moving towards more value added issues. Also, they will need to implement more consistent definitions of indicators, and distinguish more systematically between outputs and impacts; establishing benchmarks and monitoring horizontal issues will require further attention.

Finally, for the first time since the reform of the Structural Funds, the new programming period offers Objective 2 regions the stability and certainty of a lengthy (seven-year) period of EU support. With the range of economic development expertise available in Objective 2 areas, and building on ten years of EU programme experience, these regions have a real opportunity to use the Funds to make a difference to their regional development environment. Equally important is the impact that can be made on the process of regional development – the theme of this study. The next programming period is likely to be the last period of substantial assistance for Objective 2 areas, and programme managers need to be thinking, already now, about how to embed structures, systems and initiatives to the long-term benefit of the regions. Programmes need to include plans for self-sustaining economic development initiatives, so that the result is durable economic structures for the future.