Analysis of the Impact of NPM Reform

Abstract

This paper assesses our state of knowl -edge concer ning the ‘New Publ ic Man -age ment ’ (NPM) reforms which seem tohav e been launc hed in so many coun -tries . In the first sect ion it apprai ses thetypes of materials avai lable as a basisfor such an asses sment . I t then con -siders, and re jects, the thesi s that ,bec ause of the improved per formanc ethey bring, these reforms are global inreach and inevi table in nature . Subse -quent ly, the ma in par t of the paperassesse s the avai lable evidence withrespec t to a num ber of key dimens ionsof re form impa cts. It conc ludes that ,whi le there is eviden ce of speci f icimprovem ents in par ticu lar instanc es,the gene ral case for NPM as a solut ionto d iverse probl ems of gove rnance inmany di fferent count ries is far lessfirmly es tab lished than is comm onlysuppo sed.

Key wordsNew Publ ic Manage ment , publ ic man -age ment refo rm, resul ts

IS THE EMPEROR INHIS UNDERWEAR?An analysis of the impacts ofpublic management reform

Christopher Pollitt

Christopher PollittProfessor of PublicManagementFaculty of Social SciencesErasmus UniversityBurg. Oudlaan 50NL-3062 PA RotterdamThe NetherlandsTel: +31 10 408 2134Fax: +31 10 408 9099E-mail: [email protected]

Vol. 2 Issue 2 2000 181–199Public Management ISSN 1461–667X print/ISSN 1470–1065 online© 2000 Taylor & Francis Ltd http://www.tandf.co.uk/journals

INTRODUCTION

The main aim of this article is to assess the available evidence on the results of thepublic management reforms which have swept through the OECD world during thelast ten to � fteen years, and which continue today. To attack such a huge subject in asingle article necessarily requires the discussion to be pitched at a fairly high level ofgeneralization. On the other hand, while the detail is both rich and occasionallyparadoxical or contradictory, on the broad scale attempted here there do seem to besome larger points which are worth making.

To approach the main question of impact, some attention needs � rst to be devotedto three signi� cant preliminary questions. These are, � rst, what kind of evidentialmaterials are available; second, what kinds of reform are we talking about; and, third,what do we mean by ‘results’? These three preliminaries will be tackled sequentially inthe next three sections. Subsequently, the main part of the article will deal with thecentral question of what we know and what we do not know about the results ofreform. Such reforms have been widely trumpeted as both necessary and successful(Pollitt and Bouckaert 2000). In some senses they constitute a new orthodoxy (seebelow). Notwithstanding this display of af�rmation, however, it remains pertinent toask whether the Emperor of Public Management Reform, as he is paraded before usby governments and of� cials, is actually richly robed, thinly clad or even ‘naked asnature intended’?

CHARACTERISTICS OF THE AVAILABLE MATERIALS

A great deal of the available material on public management reform is either promotional(‘look at what we are doing’) or how-to-do-it (‘a guide to . . .’). Governments produceWhite Papers, statements and booklets in which they attempt to convince legislatures,the media, the public and public servants themselves that their reforms are signi� cant,well-intentioned and likely to produce a variety of improvements. Departments producepractical guidelines to help their staffs implement change (e.g. HM Treasury 1992;Department of Finance and Administration 1998). The promotional documents areintended to persuade, and, accordingly, they tend to be heavily freighted with rhetoricand rather light on self-criticism (e.g. Gore 1996, 1997; Chancellor of the Duchy ofLancaster 1997; Chancellor of the Exchequer 1998; Prime Minister and the Minister forthe Cabinet Of�ce 1999). Consultants and other advisers are also either promotional orhow-to-do-it – often they are selling their patented systems (re-engineering, total qualitymanagement (TQM), benchmarking or whatever) and ultimately they are all selling theirservices to governments, public agencies and corporations. The how-to-do-it guides aremore down-to-earth than glossy of� cial reports and White Papers, but they, too(necessarily and understandably) assume that success is possible and that, if staff followthe good advice, of� cially promulgated goals will be reached.

182 Public Management: an international journal of research and theory

It would be unwise, therefore, to assume that, in aggregate, these types of rhetoricand documentation afford a full and balanced picture of what is happening ‘on theground’ throughout the administrative systems of the countries concerned. One doesnot have to be either a cynic or a post-modernist to be aware of the frequency ofrhetoric/practice gaps, or of the extent to which reform talk can take on a life of itsown, somewhat divorced from everyday administrative practice (Brunsson 1989; Hood1998). In this article our main recourse will therefore be to studies which can, primafacie, claim a greater degree of independence than either the promotional or the how-to-do-it literature – particularly to academic studies and to evaluations or otherassessments with a signi�cant degree of independence or ‘distance’.

WHAT IS HAPPENING IN THE WORLD OF PUBLIC MANAGEMENT REFORM?

A �rst answer which is simple but wrong

For some years now there has been a powerful story abroad. It tells that there issomething new in the world of governance, termed ‘the new public management’(NPM), ‘reinvention’, ‘re-engineering’, ‘entrepreneurial government’ or some sim-ilarly dynamic title. This is generally presented as a formula for improving publicadministration and achieving, as the catchphrase for the US National PerformanceReview (NPR) has it, ‘a government that works better and costs less’. In theirin� uential book Reinventing Government, Osborne and Gaebler (1992) put it verystrongly. Referring to what they termed ‘the rise of entrepreneurial government’ inthe USA, they claimed that ‘a similar process is underway throughout the developedworld’ (1992: 328) and that it was ‘inevitable’ (1992: 325). Six years later, Mr Blair’s� rst major white paper on government echoed much of this rhetoric, albeit with a fewlocal variations (Prime Minister and Minister for the Cabinet Of� ce 1999). From thisperspective particular governments or public services can be seen as being ‘well ahead’or ‘lagging behind’ along what is basically a single route to reform. In many of the‘promotional’ publications the characteristics of the reformed public sector organiza-tion have been speci�ed. Typically, these include:

c being close to its customers;c being performance-driven (targets, standards) not rule-bound;c displaying a commitment to continuous quality improvement (again, targets,

standards);c being structured in a ‘lean’ and ‘� at’ way – highly decentralized, with street-level

staff who are ‘empowered’ to be � exible and innovate;c practising tight cost control, with the help of modern, commercial-style

accounting systems;c using performance-related systems for recruiting, posting, promoting and paying

staff.

Pollitt: The impacts of public management reform 183

Furthermore, if these are the characteristics of individual organizations within areformed public sector, ‘reinvented’ governments will also display a distinctive approachto their work in a broader way. They will:

c ‘steer not row’, i.e. become more concerned with strategy and less withcarrying-out;

c act in anticipatory ways – for a host of public problems prevention is better thancure;

c seek to use market mechanisms wherever possible, either in the form of quasi-markets to introduce competition between public providers, or by contractingout or privatizing services which were previously undertaken directly by thestate;

c seek inter-organizational partnerships, both within the public sector (‘joined-upgovernment’) and with the private and voluntary sectors.

This, therefore, is the simple answer to the question ‘what is happening?’. Govern-ments are redesigning institutions and procedures so as to conform to the new modeloutlined above. Everyone is doing more or less the same thing, because they have littlechoice. Powerful forces in the environment are obliging governments to change. Someare further ahead than others.

A second, more complicated, but more accurate answer

The community of scholars conducting comparative analyses of public managementreforms is not large but, over the past decade, it has produced a number of signi� cantstudies covering Western Europe, North America and Australasia (e.g. Pierre 1995;Trosa 1995; Flynn and Strehl 1996; Hood 1996; Olsen and Peters 1996; Kickert1997; Lane 1997; Pollitt and Summa 1997; Peters and Savoie 1998; Pollitt andBouckaert 2000). What these show is a world in which, although the broad aims ofproducing more ef� cient, effective and responsive public services may have beenwidely shared, the mixtures of strategies, priorities, styles and methods adopted bydifferent governments have varied very widely indeed. As Flynn and Strehl said of theirresearch: ‘We quickly found that there were reasons to doubt the idea of convergence’(1996: 4). Guy Peters suggests that although there is ‘a set of relatively commonstimuli for change . . . [ w] hat is different is how political systems have interpreted theideas and responded to the demands and/or opportunities for introducing admin-istrative change’ (Peters 1997: 266). This is a position that is broadly shared in my ownwork, although I would add that reform ideas themselves have also varied considerablyfrom country to country, and certainly the priority given to different components (e.g.privatization, contracting out) has � uctuated enormously (Pollitt and Bouckaert2000).


Part of the explanation is that countries have not started from the same point, eitherin terms of the make-up of their public sectors or in terms of the way they think aboutthe role and character of the state (Pierre 1995; Kickert 1997; Ministry of Finance1997; Pollitt and Summa 1997; Guyomarch 1999). ‘Path dependent’ explanations � tpublic management reform rather well (Pierson 1997; Premfors 1998; Pollitt andBouckaert 2000). Furthermore, governments have not all possessed the same capacitiesto implement reforms. In some countries, such as Germany, changing the centraladministrative structures is politically and legally very dif�cult. In other countries,such as New Zealand and the UK, it has been comparatively easy. A subsidiaryweakness in the ‘simple’ version is that it posits a uniform past, in which ‘traditionalbureaucracy’, like some ponderous dinosaur, ruled the earth. This is simply wrong –much of post-war ‘big government’, in many OECD countries, consisted of sproutingwelfare state organizations that were usually not organized along strict bureaucraticlines. State schools, hospitals and social and community services agencies took on avariety of forms, and in many of these autonomous professionals, not bureaucrats,were the key actors (Clarke and Newman 1997).

Thus, not every country is taking the same route, and, in particular, the radicalreforms implemented during the 1980s and early 1990s in New Zealand and the UKare certainly not universally regarded as a desirable model to emulate. In somecountries leading opinion formers regard the NPM with considerable suspicion (e.g.for France, see Guyomarch 1999; for Germany see Konig 1996 and Derlien 1998; andfor Sweden see Premfors 1998). The USA is an interesting case because, although ithas been a fountain of rhetoric for reinvention and re-engineering, the federalexecutive’s capacity for implementing coherent, broad-scope reforms is severelylimited by the well-known fragmentation of the American political system (Peters1995). Here is a government whose rhetoric frequently outreaches its implementationcapacity – witness the history of some of President Reagan’s reforms during the 1980sor, further back, the disappointing histories of, inter alia, Planning, Programming andBudgeting Systems (PPB), Management by Objectives (MbO) and Zero-Based Budget-ing (ZBB) (Savoie 1994; General Accounting Of� ce 1997). By contrast one mightthink of, say, Denmark, which has carried through extensive public managementreforms but without much publicity or public controversy, or Finland, which, withoutfanfares, launched and sustained a major modernization effort between 1989 and 1997(Ministry of Finance 1997).

DEFINING ‘RESULTS’

Before one can assess evidence about impacts, one has to decide what kind of thing isgoing to count as a result. This is by no means straightforward. For example, a resultcould be any one or more of the following:

c savings (reduced budget appropriations);


c improved processes (e.g. faster, more accessible complaints procedures; quickerturn-round times for repairs or the processing of licences; ‘one-stop shops’offering several services in one place);

c improved ef� ciency (better input/output ratios, e.g. more students graduate perfull-time equivalent member of staff; the same number of drivers’ licences areissued with 20 per cent fewer staff);

c greater effectiveness (less crime, poverty, functional illiteracy; homelessness;drug abuse; gender or ethnic inequality; more new jobs created; more contentedand trusting citizens, etc.);

c an increase in the overall capacity/�exibility/resilience of the administrativesystem as a whole (e.g. through the recruitment and training of more skilled,more committed public servants).

Furthermore, each of these categories contains its own conceptual puzzles, de� nitionalproblems and pitfalls in operationalization (Pollitt and Bouckaert 2000: ch. 5). Someof these will be alluded to further as the evidence on results is examined in latersections.

WHAT EVALUATIONS HAVE BEEN DONE?

There have been surprisingly few independent, broad-scope evaluations of the publicmanagement reforms (Pollitt 1995, 1998). Those which have been conducted tend tosuffer from some fairly fundamental conceptual and methodological limitations (whichhave sometimes been acknowledged, and on other occasions not). Australia carried outone very large study (Task Force on Management Improvement 1992) and NewZealand conducted at least two important reviews (Steering Group 1991; Schick1996). In the USA there are various evaluations in train around the NPR, but the mainstudies were not yet available at the time of writing. However, a series of assessmentsof the implementation of the Government Performance and Results Act (GPRA) hasbegun to appear from the General Accounting Of�ce (e.g. General Accounting Of� ce1998). In the UK, broad-scope evaluations were noticeable by their absence during theThatcher/Major administrations, although there were a number of more focusedassessments of speci�c reforms (e.g. Employment Service 1994; Next Steps Team1998, though these were both ‘in-house’ reviews and therefore their independence canbe questioned). The Blair administration appears to be more committed to the ideathat evaluation should be a regular part of the reform process. A few countries havemade a deliberate effort systematically to review their reform experiences, usingindependent evaluators (Holkeri and Summa 1996) but these initiatives have been verymuch the exception rather than the rule.

The most common limitations to those studies which have been done are asfollows:


c an absence of reliable base-line measures, so that before-and-after comparisonsbecome speculative;

c an absence of benchmarking, e.g. the productivity gains of a privatized � rm maybe praised without it being noticed or admitted that comparable non-privatizedcorporations have made similar gains over the same period (Naschold and vonOtter 1996: 24–5);

c limited or no gathering of the views of service users;c scarcity or absence of data on transitional costs. For example, the � rst major

report on the New Zealand reforms contains no cost �gures at all – SteeringGroup 1991 – and the huge Australian evaluation acknowledges the dif� culty ofassigning savings to the reforms – Task Force on Management Improvement1992. Kettl (1994: 9) makes the same point about the US NPR. Yet implementa-tion costs can be considerable. For example, the implementation costs for theUK national curriculum for schools totalled approximately 1172M eurosbetween 1988/89 and 1994/95;

c scarcity or absence of data on step-changes in transactional costs and/or on othercontinuing ‘side-effects’ such as the loss of trust or a degree of value confusion(see, for example, Kirkpatrick 1999 for the UK and Jorgensen 1999 forDenmark);

c opinion gathering being limited to, or biased towards, senior staff (a number ofsurveys have shown that middle and lower-level staff are often more critical ofreforms than their bosses);

c little analysis of contextual variations which may mean that a similar type ofreform will work well in one situation or locality, but not in another (Pawsonand Tilley 1997);

c limited or no attention to attribution problems. Often several reforms haveproceeded simultaneously, and external conditions have also been changing. Thismakes it very hard to attribute results con� dently to speci�c reforms. As onecommendably honest of� cial evaluation put it: ‘it is unlikely to be possible todisentangle the effects of Agency status from other elements of E[ mployment]S[ ervice] performance: and . . . it is unlikely to be possible to create an exactpicture of what would have happened if ES and Agency status had not occurred’(Employment Service 1994: 10);

c narrow range of criteria applied to the � ndings (e.g. productivity measures only,with no attention to equity, to staff morale or to externalities). In effect, mostof the evaluations fail to distinguish between and/or miss out altogether many ofthe types of ‘result’ listed in the previous section of this article.

There is little sign that these signi� cant limitations to evaluation designs are beingaddressed. Recent evaluations (e.g. Schick 1996; Next Steps Team 1998) appear to bejust as prone to major weaknesses as those undertaken � ve or more years ago.


WHAT RESULTS HAVE BEEN FOUND?

The limitations – or downright absence – of evaluations discussed in the previoussection mean that many important questions cannot be answered, or can only beanswered tentatively or anecdotally, with many quali� cations and reservations. Conse-quently our discussion of results may seem to have a slightly frustrating quality to it.Nevertheless, some aspects are clearer than others, and quantities of evidence liearound the world, inviting scholarly sorting and interpretation. In making a small startto that large labour, I will organize my brief comments under the same headings aswere used in the de� ning ‘results’ section above. Thus we begin with the claims ofreform to achieve economies in the operation of the public sector.

SAVINGS

One German scholar attempted to test what he described as the ‘OECD hypothesis’:namely, that bureaucratic regimes would perform less well in macro-economic termsthan regimes which had modernized themselves according to the NPM prescription.He concluded that:

Confronting our �ndings with the hypothesis formulated by the OECD as to the relationship between

macroeconomic performance (economic growth, productivity and unemployment), on the one hand, and the

regulation regime (bureaucratic governance by rule and its alternatives) on the other, the OECD hypothesis has to

be strikingly refuted: all the countries with bureaucratic governance by rule exhibit with respect to almost all the

dimensions a markedly better macroeconomic performance than the other countries.

(Naschold 1995: 39)

Naschold goes on to acknowledge that there are considerable dif� culties in inter-preting this apparent negative correlation. His caveat should be vigorously endorsed.Even if one con� nes one’s attention to a narrower range of macro-economic indicatorsthan those cited by Naschold, interpretation remains deeply problematic. For example,the OECD database shows that government outlays as a percentage of nominal grossdomestic product (GDP) fell between 1985 and the late 1990s, at least in the majorityof the countries discussed in this article. However, it would be rash indeed to attributethis shrinking proportion to management reforms. There are all sorts of interpretativepitfalls which forbid any such conclusion (Pollitt and Bouckaert 2000: ch. 5). We haveto allow for the state of the economies as a whole – a rising or falling ratio may beas much due to growth or recession in the market sector as to ‘savings’ in the publicsphere – Naschold’s own ‘results’ might have looked rather different only three yearslater. A further problem is that a country-by-country review shows a poor correlationbetween large reductions in the ratio and the depth or breadth of managementreforms. Even more fundamentally, the direction of the arrow of causation is not clear.Instead of assuming that management reform leads to savings we might hypothesize


that forced savings lead to management reform. This second interpretation is stronglysupported by one of the few reasonably sophisticated analyses of public sectorproductivity yet undertaken (Murray’s 1998 interpretation of the Swedish experi-ence).

By the same token, it would be rash to conclude from the parallel OECD � gureswhich show rising public debt (also as a percentage of nominal GDP) for manycountries that management reforms increase state borrowing. A more prudent positionwould be to accept, however reluctantly, that movements in macro-economicaggregates simply will not tell us anything clear and sure about the effects ofmanagement reform.

Therefore we should move away from macro-economic indicators and instead askquestions about the measured savings generated by particular reform efforts. Yet, as wehave already noted, the most sophisticated evaluations are unable reliably to calculateor attribute savings with respect to broad programmes of reform (Steering Group1991; Task Force on Management Improvement 1992). It is only when much morespeci�c and local changes are examined that testable estimates of savings begin toappear. At this level it appears reasonably clear that many reforms have led to savings.The ‘promotional’ government publications referred to in the � rst section are full ofexamples. Vice-President Gore writes of buying government staplers for $4 instead of$54 (Gore 1996: 5). A range of UK executive agencies report reduced unit costs,year-on-year (e.g. Chancellor of the Duchy of Lancaster 1997: 117–18). Some of theseachievements are impressive. Some would be impressive, if they were independentlyvalidated and shown to be a fair and true picture. Many, however, are hard to assess.A ‘saving’ on one dimension may have been offset by increases in expenditureelsewhere, or by quality reductions, or by scope of service reductions, or by shiftingcosts elsewhere in the public sector (and therefore achieving no overall advantage forthe state as a whole). Yet one must not carp too much: when a reform of procurementpolicy frees (or obliges) public servants to purchase simple requirements (staplers,of� ce furniture, security services) from cheaper suppliers, it requires a contortion tosee this as other than an improvement.

One kind of ‘saving’ that appears relatively easy to count is a reduction in thenumber of civil servants (‘downsizing’ in management parlance). The OECD databank(OECD 1999) shows that some countries have been able to make large reductions inthe ratio of public employment to total employment. Other countries have not (seeTable 1).

Once again, however, there is need for great interpretive caution when deployingaggregates of this kind. The OECD statisticians do their best, but varying de� nitionsof government employment plague their comparisons. On this basis the correlationbetween staff reductions and NPM-style reform appears positive, but weak. The mostenergetic implementor of NPM during the 1990s – the UK – underwent proportion-ately less downsizing than the doctrinally more cautious Swedes – the civil serviceeven appears to have grown. The sometimes virulent anti-bureaucratic rhetoric of the


Reagan and Bush Presidencies produced a much smaller effect on the US civil servicethan the apparently more federal government-friendly Clinton Presidency. An exten-sion of these � gures to other countries, such as Finland and The Netherlands, wouldmultiply such curiosities. Despite all these quali�cations, however, it seems reasonablysure that some governments have been able to make substantial reductions in thenumbers of core public servants they employ. Margaret Thatcher downsized the non-industrial civil service by more than one-� fth. Vice-President Gore’s NPR reduced thefederal workforce by 240,000 as of the beginning of 1996 (Gore 1996: 5).

IMPROVED PROCESSES

Again, anecdotal evidence crowds together in the promotional literature. With respectto American pensions: ‘Ten million workers in small businesses could bene� t from the

Tabl e 1: Changes in c ivi l service empl oym ent , the publ icsec tor and the economy as a whole (%)

1985–90 1990–7

AustraliaCivil service 2 7.1 2 22.0Public sector – 2 2.7Economy as a whole 1 16.2 –CanadaCivil service 1 6.2 2 19.0Public sector 1 9.9 2 4.7Economy as a whole 1 8.8 1 15.2FranceCivil service 1 1.9 1 6.3Public sector 1 2.3 2 4.4Economy as a whole 1 3.8 1 2.9SwedenCivil service 2 28.9 2 38.0Public sector 1 5.4 2 18.1Economy as a whole 1 4.9 2 11.0UKCivil service 2 5.8 1 16.1Public sector 2 2.6 2 12.9Economy as a whole 1 9.9 1 9.2USACivil service 1 1.7 2 17.0Public sector 1 11.7 1 5.3Economy as a whole 1 13.2 1 12.1

Source: OECD (1999: Annex 2, p28).


new, simpli�ed 401(k)-type plan – no red tape, just a simple form’ (Gore 1996: 47).The UK Driver and Vehicle Testing Agency repeatedly reduced the average waitingtime for a driving test appointment – although at the same time the waiting time forcar tests with its sister agency (the Driving Standards Agency) drifted upwards(Chancellor of the Duchy of Lancaster 1997: 63–4). Most of us probably also havepersonal experience of improvements: desk staff better trained to handle dif�cultrequests; greater attention paid to the decor, comfort and cleanliness of waiting areas;simpli�ed forms. Techniques such as TQM and re-engineering generate many measuredprocess improvements, sometimes of spectacular proportions. There is now a greatdeal of public sector knowledge about how to improve individual processes, once thewill to do so and the appropriate techniques are brought together in the same place.Modernized management has some real success stories to tell.

Once more, however, there is an argument for viewing these speci�c achievementsin a broader and more ambiguous context. With complex interdependent publicservices, improvements along one dimension may be achieved at the cost of theneglect of another. For example, improving against a target of x per cent of casescleared within a given time period may tempt of� cials to neglect a small percentageof really dif� cult and time-consuming cases (see the UK National Audit Of� ce study ofthe Bene� ts Agency – National Audit Of� ce 1998). Or take the case of the localauthority which invests a signi� cant volume of its scarce resources in automatedelectronic information points, with the aim of providing local residents with a fast,clear 24-hour information service about the Council’s services. For those who use thescreens, these aims may be achieved – the process of informing them has beenunmistakably improved. However, for a variety of reasons, the elderly and certainminorities for whom English is not the � rst language make virtually no use of theinformation points, and women generally use them far less often than men. Meanwhilepossible moves to increase the resources going towards informing these other groups(e.g. through hiring more interpreters in minority languages) have been postponed –indeed resource levels may even have been cut back – because the kiosks are theproject of the moment. In such circumstances (closely adapted from a real case) it isclear that the information process has been improved for young and middle-aged whitemales, but whether it has been improved in general depends on the extent to whichother aspects of the service have remained the same, or been squeezed by resourcediversions to the � agship project (for a general treatment of measurement problems,see Pollitt (1999)).

IMPROVED EFFICIENCY

For two decades, the ef� ciency criterion has lain at the heart of many managementreform initiatives. Reorganizing so as to achieve more outputs per input or the sameoutputs for reduced inputs is one of the core skills of good managers. Achieving a


certain percentage ‘ef� ciency gain’ became an annual routine for UK governmentdepartments and agencies, and for NHS hospitals during the 1980s and early 1990s.Furthermore, in principle at least, the measurement of ef� ciency is somewhat lessdif� cult than that of either quality or effectiveness. With quality measurement there isalways the awkward initial step of trying to � nd some consensus among users as towhat, for them, constitutes ‘quality’. Effectiveness measurement entails research intoimpacts ‘out there’, beyond the organization, and may be both costly and methodo-logically complex. Technical, ‘X’ ef� ciency, however, requires a comparison betweenorganizational inputs and organizational outputs – both of which are usually recorded(or recordable) without too much external research or conceptual head-scratching.

A � rst point to make is that ef� ciency gains may be achieved at the cost of other,less desirable effects. Thus a UK study of the effects of contracting out local servicescame to the conclusion that, while ef� ciency may have risen in many cases, equalopportunities hiring had suffered (Escott and Whit�eld 1995).

Second, while many of the claims of ef� ciency gains are probably perfectlyreasonable and accurate, it would be prudent not to take all assertions at face value.Consider the widely accepted idea that contracting out public services (whatever itsother effects may have been) has regularly led to ef� ciency gains:

claims that that empirical studies �nd ‘consistently’ and ‘without exception’ that contracting out is more ef�cient

than municipal supply are demonstrably untrue. Even taken at face value, only around half of the studies discussed

in the paper {a review of contracting out in US local government} is associated with lower spending and higher

ef�ciency. Furthermore, many of the studies contain speci�c methodological �aws that cast doubt on the evidence

on the impact of service contracts. . . .

(Boyne 1998: 482)

Neither is this an isolated case. Talbot’s research into the performance indicatorsystems used by UK ‘Next Steps’ executive agencies showed that measurements oftheir ef� ciency were both patchy and volatile. Roughly half the declared aims andobjectives were not covered by indicators at all, while two-thirds of the ‘keyperformance indicators’ for a sample of ten agencies had been dropped or replacedwithin a six-year period (Talbot 1996, 1997). A recent Dutch study failed to � nd goodevidence of ef�ciency improvements in a set of recently ‘autonomized’ organizations(Ter Bogt 1999). Embarrassing measures, where performance is declining or stub-bornly low, may be quietly discarded and rede�ned indicators put in their place (as wasfamously the case for the headline UK of� cial de� nition of unemployment during the1980s). Sometimes organizations may even cheat (see, for example, Hencke 1998).Equally, reorganizations themselves may disrupt time-series data by altering the sphereof jurisdiction of an organization or changing data-collection methods or categories.Reliable time series of well-validated ef� ciency measures turn out to be much rarerthan one might have thought (as Pollitt et al. 1998, found for UK hospitals, schoolsand housing agencies). A recent study of � ve European national audit of� ces showed


that even Supreme Audit Institutions, when they conduct performance audits, oftenseem to be able to construct and apply true ef� ciency measures in only a minority oftheir studies. More often they fall back on assessing the presence or absence of goodmanagement practice (Pollitt et al. 1999: ch. 6).

Of course, the frequent absence of ‘gold standard’ measures does not show thatef� ciency has fallen (or risen) but it does cast a different light on the apparentabundance of improvement claims. The celebratory confetti is sometimes so thick inthe air that it is hard to make out exactly what the emperor is wearing.

GREATER EFFECTIVENESS

For at least thirty years civil servants and evaluators have recognized that assessing theeffectiveness of many public policies and programmes is an extremely dif� cult task.Occasionally some particular programme will enjoy the bene�t of an available, validand relevant indicator of outcome, but often the links between programme activitiesand � nal outcomes are tentative or obscure. There are several well-documentedreasons for this complexity. Politicians frequently mandate policy objectives which, inWildavsky’s famous phrase are ‘multiple, con� icting and vague’ (Wildavsky 1979:139). Thus the initial question of what outcomes are being aimed at may be hard toanswer in operational terms. Then there is the problem of timescales: the � naloutcomes of some educational, health and environmental programmes, for example,may lie a long way down the road – longer than many politicians and citizens arewilling to wait before passing judgement. A third common dif� culty is that of safelyattributing observed effects to the programme in question. If unemployment falls, howmuch of that is due to the retraining programme and how much to a generalimprovement in economic conditions? If the health status of a community improves isit the health promotion programme, or improved housing, or increased incomesleading to better diets, or some combination of all these things?

By itself, management reform alters none of these constraints. Indeed, the evaluation of theeffectiveness of management reform itself is subject to precisely these challenges – theaims of the reform are often hard to operationalize, the timescales over which effectsoccur can be long and drawn-out and the attribution of observed effects is frequentlyuncertain (Pollitt 1995).

These dif�culties are widely acknowledged by practitioners and politicians. In NewZealand – a country rightly famed for the thoroughness of its management reformsand the sophistication of its systems – the minister with responsibility for the civilservice opened a 1997 conference of senior public managers with the followingwords:

Quite properly, a great deal of effort was invested in the �rst few years in mastering the new technology – in

making sure that outputs were properly speci�ed, correctly priced, and so on. But within two or three years there


was a sense that we might be becoming, instead of input-�xated, output-�xated. . . . My own impression is that

we still have plenty to do, to make the system work effectively . . . A �rst step towards this should be, I suggest,

a conscious effort by senior Public Service managers to lift their eyes from individual outputs and to spend longer

thinking about the Government’s strategic result areas – about outcomes in other words.

(East 1997)

Similarly, the UK Labour government, after � fteen years of intense and unremittingmanagement reforms, declared earlier this year that henceforth the need was to focuson outcomes rather than just on inputs, functions or value-for-money (Prime Ministerand Minister for the Cabinet Of� ce 1999: 15–18).

We should not be surprised, therefore, when we discover that cases where there isunmistakable evidence of management reform producing more effective governmentaction are rare. The connections between management reform and the effectivedelivery of long-term policy goals are often both distant and complicated by factorsthat lie beyond the control of public managers.

MORE CAPACIOUS/FLEXIBLE/RESILIENT ADMINISTRATIVE SYSTEMS

Systems improvements, it is argued, are not to be judged on the basis of the successor failure of a single project, programme or policy, but rather in a more holistic way.For example, the transformation of a rigid, inward-looking, slow-moving bureaucratichierarchy into a ‘� at’, responsive, multi-disciplinary agency could be said to haveincreased that system’s capacity to cope with new developments in its environment. Asin this example, system changes are broad-scope and will frequently involve bothmajor restructurings and an engineered shift in the dominant organizational culture.

Reformers have made many claims for systems transformations. For example, Vice-President Gore writes that in the US federal government: ‘Many bosses are changingthe way they do their jobs – encouraging innovation and customer service instead ofjust making workers toe the line’ (Gore 1996: 16). Towards the end of John Major’sConservative administration the UK Cabinet Minister for the Public Service wrotethat:

In my current and my previous Ministerial posts I have been struck by the way in which the Next Steps programme

has transformed the civil service. . . . This Review contains many examples of improvements in the quality of

service provided to the public. . . . It also demonstrates very clearly the Government’s commitment to openness,

managerial accountability and better-focused, better-managed systems.

(Chancellor of the Duchy of Lancaster 1997: v and viii)

Such claims may be justi�ed, or they may be largely rhetoric. Distinguishing betweenthe beef and the blather is often dif� cult. One problem is that the currency of systemsclaims’ is hard to cash into speci�c measurements or indicators. Thus claims of ‘better


accountability’ or cultural change or greater responsiveness are relatively easy toillustrate with a sound-bite anecdote but very hard to capture in a general measure.Who has measured cultural change? Who has reliable statistics on changes in‘accountability’? In what units may one count shifts in the ‘responsiveness’ or ‘capacity’of a system of public administration? On those (still quite rare) occasions whensystematic research has been carried out on such matters, the � ndings have by nomeans all pointed towards positive ‘transformations’, though they have certainlyregistered that change is underway (Task Force on Management Improvement 1992;Talbot 1994; Rouban 1995). And if a representative sample of citizens are askedquestions, the picture is no more clear cut. After a period of intensive reform,including the implementation of the Citizen’s charter, and the roll-out of the Next Stepsprogramme to embrace three-quarters of the entire civil service, a UK white papercommented as follows:

Research with the Peoples’ Panel {. . .} shows that more people agree than disagree that our public service

providers are friendly, hard-working and keen to help. But, although the number of people who are satis�ed is

increasing, many services still fall short of expectations. Two out of �ve people think services have got no better

in the last �ve years, and over one in three thinks they have become worse.

(Prime Minister and Minister for the Cabinet Of�ce 1999: 23)

From a strict scienti�c perspective, therefore, the jury is still out on ‘systemstransformations’ – and it may be out for some time. The Emperor says he hasrevolutionized his entire wardrobe, but some of his workers doubt if it is entirely forthe better, and many of the onlookers have not yet noticed much difference.

CONCLUSIONS

In conclusion, what can be said about the state of dress of the Emperor of NPM/NPR?First, it is hard to see him clearly: there are a lot of conceptual problems about quitewhat it is we are looking for, and also some methodological and interpretive puzzlesabout what the available data can be said to show. Second, in so far as we can get agood look, the Emperor does not appear to be naked. Management reform has not allbeen windy rhetoric, by any means. Downsizing has been accomplished in a numberof countries. In many instances measured ef� ciency has de� nitely increased. Thein� uence of published targets in prompting improvement on speci�c dimensions hasbeen demonstrated time and time again. In many cases, also, speci�c services havecertainly become more user-sympathetic and � exible. All these garments can be wornwith pride. At the same time it should be conceded that their full cost – ‘side-effects’and all – has frequently been obscure. Other features of services, which are notmeasured or publicized, may have taken a turn for the worse. Other groups – of staffor citizens – may have suffered degenerating conditions as a consequence of the drive


to improve the more salient aspects of a particular service. It is only rarely that the fullbalance sheet is available for scrutiny.

Two other things can be said about the Emperor. While he may not be naked he isfar from fully dressed. Some of the larger claims heard from time to time for NPM andthe NPR must be judged either false (so far at least) or not proven. Large savings inaggregate public expenditure have seldom accrued from management reforms per se. Agovernment’s legitimacy is not likely to be heavily in� uenced by such reforms either– the chief determinants of public attitudes towards their governments seem to lieelsewhere. The correlation between the implementation of NPM reforms and macro-economic performance is hardly striking, with Germany, Japan and the USA beingthree of the most successful economies (by most criteria) over the last twenty yearsand yet none of these being among the group which has implemented the most radicalreforms (indeed, Germany and Japan have been among the least active in this respect).Furthermore, the achievement of more effective (as distinct from more ef� cient)government is hard to demonstrate. Even in those cases where improved effectivenesscan be demonstrated, there is usually considerable ambiguity as to what it should beattributed. The splendid coat of many colours envisaged by some reformers – slim,fast, effective, decentralized, open, trusted government – still lies more in the realmsof hope and imagination than in demonstrated and warranted reality.

Finally, we should remember that the jurisdiction of this half-clad Emperor islimited. His heartlands have never really extended beyond Australasia, North Americaand the UK. Although some of his methods may have been selectively borrowed byother countries – especially the Dutch and the Nordics – these countries have neverunconditionally accepted his authority. As one senior Dutch civil servant said recentlyto a visiting New Zealander, ‘when will you Anglo Saxons learn that pulling grass upby the roots doesn’t make it grow any faster?’.

ACKNOWLEDGEMENTS

This article is a revised version of a paper presented at the International Summit onPublic Service Reform, Winnipeg, Canada, June 1999. I am grateful to variousdiscussants at that event, to Rune Premfors of SCORE, Stockholm and to theanonymous referees of Public Management.

REFERENCESBoyne, G. (1998) ‘Bureaucratic Theory Meets Reality: Public Choice and Service Contracting in US

Local Government’. Public Administration Review, 58:6 pp474–84.Brunsson, N. (1989) The Organisation of Hypocrisy: Talk, Decisions and Actions in Organisations, Chichester:

John Wiley.


Chancellor of the Duchy of Lancaster (1997) Next Steps: Agencies in Government: Review, 1996,Cm. 3579, London: The Stationery Of� ce.

Chancellor of the Exchequer (1998) Modern Public Services for Britain: Investing in Reform, Cm. 4011,London: The Stationery Of� ce.

Clarke, J. And Newman, J. (1997) The Managerial State, London: Sage.Department of Finance and Administration (1998) The Performance Improvement Cycle: Guidance for

Managers, Canberra, Australia: Department of Finance and Administration.Derlien, H.-U. (1998) From Administrative Reform to Administrative Modernization , Bamberg:

Verwaltungswissenschaftliche Beitrage 33.East, P. (1997) ‘Opening address’ given at the Public Service Senior Managers’ Conference,

Wellington, NZ, October.Employment Service (1994) Employment Service: An Evaluation of the Effects of Agency Status 1990–1993,

London: Employment Department Group.Escott, K. And Whit�eld, D. (1995) The Gender Impact of CCT in Local Government, Manchester: Equal

Opportunities Commission.Flynn, N. And Strehl, F. eds (1996) Public Sector Management in Europe, London: Prentice Hall/

Harvester Wheatsheaf.General Accounting Of� ce (1997) ‘Performance Budgeting: Past Initiatives Offer Insights for GPRA

Implementation’, GAO/AIMD-97-46, Washington DC, GAO, March.—— (1998) ‘The Results Act: Observations on the Department of State’s Fiscal Year 1999 Annual

Performance Plan’, GAO/NSIAD-98-210R, Washington DC, GAO, June.Gore, A. (1996) The Best-Kept Secrets in Government: A Report to President Bill Clinton, Washington DC:

US Government Printing Of� ce, National Performance Review.—— (1997) Businesslike Government: Lessons Learned from America’s Best Companies, Washington DC:

National Performance Review.Guyomarch, A. (1999) ‘ “Public Service”, “Public Management”, and the “Modernization” of French

Public Administration’. Public Administration, 77:1 pp171–93.Hencke, D. (1998) ‘Jobcentres Fiddled the Figures’. Guardian, 8 January p2.HM Treasury (1992) Executive Agencies: A Guide to Setting Targets and Measuring Performance, London:

HMSO.Holkeri, K. And Summa, H. (1996) ‘Contemporary developments in performance management:

Evaluation of public management reforms in Finland: From ad hoc studies to a programmaticapproach’ paper presented to PUMA/OECD, Paris, November.

Hood, C. (1996) ‘Exploring Variations in Public Management Reform of the 1980s’, in H. Bekke, J.Perry and T. Toonen (eds) Civil Service Systems in Comparative Perspective. Bloomington &Indianapolis, IN: Indiana University Press.

—— (1998) The Art of the State: Culture, Rhetoric and Public Management, Oxford: Oxford UniversityPress.

Jorgensen, T. (1999) ‘The Public Sector in an In-between Time: Searching for New Public Values’.Public Administration, 77:3 pp565–84.

Kettl, D. (1994) Re-inventing Government? Appraising the National Performance Review, Washington DC:Brookings Institution.

Kickert, W. ed. (1997) Public Management and Administrative Reform in Western Europe, Cheltenham:Edward Elgar.

Kirkpatrick, I. (1999) ‘The Worst of Both Worlds? Public Services without Markets or Bureaucracy’.Public Money and Management, 19:4 pp9–14.

Konig, K. (1996) On the Critique of New Public Management, Speyer: Speyer Forschungsberichte, 155.


Lane, J.-E. ed. (1997) Public Sector Reform: Rationale, Trends, Problems, London: Sage.Ministry of Finance (1997) Public Management Reforms: Five Country Studies, Helsinki: Ministry of

Finance.Murray, R. (1998) ‘Productivity as a tool for evaluation of public management reform’ paper

presented to the European Evaluation Society Conference, Rome, October.Naschold, F. (1995) The Modernization of the Public Sector in Europe, Helsinki: Ministry of Labour.Naschold, F. And von Otter, C. (1996) Public Sector Transformation: Rethinking Markets and Hierarchies in

Government, Amsterdam: John Benjamins Publishing Company.National Audit Of� ce (1998) Bene� ts Agency: Performance Measurement, HC952, London: The Stationery

Of� ce.Next Steps Team (1998) Towards Best Practice: An Evaluation of the First Two Years of the Public Sector

Benchmarking Project, 1996–98, London: Ef� ciency and Effectiveness Group, The Cabinet Of� ce.OECD (1999) Structure of Civil Service Employment in Seven OECD Countries (http://www.oecd.org/

puma/mgmtres/hrm/index.htm).Olsen, J. And Peters, B. eds (1996) Lessons from Experience: Experiential Learning in Administrative Reforms

in Eight Democracies, Oslo: Scandinavian University Press.Osborne, D. And Gaebler, T. (1992) Reinventing Government: How the Entrepreneurial Spirit is Transforming

the Public Sector, Reading MA: Adison Wesley.Pawson, R. And Tilley, N. (1997) Realistic Evaluation, London: Sage.Peters, G. (1995) ‘Bureaucracy in a Divided Regime: The United States’ in J. Pierre (ed.) Bureaucracy

and the Modern State: An Introduction to Comparative Public Administration, Aldershot: Edward Elgar.—— (1997) ‘A North American Perspective on Administrative Modernisation in Europe’ in W.

Kickert (ed.) Public Management and Administrative Reform in Western Europe, Cheltenham: EdwardElgar.

Peters, G. And Savoie, D. eds (1998) Taking Stock: Assessing Public Sector Reforms, Montreal andKingston: Canadian Centre for Management Development/McGill-Queen’s University Press.

Pierre, J. ed. (1995) Bureaucracy in the Modern State: An Introduction to Comparative Public Administration,Aldershot: Edward Elgar.

Pierson, P. (1997) ‘Path dependence, increasing returns, and the study of politics’ paper presented atthe Jean Monnet Visiting Professor Lecture, European University Institute, (revised version –October).

Pollitt, C. (1995) ‘Justi� cation by Works or by Faith? Evaluating the New Public Management’.Evaluation, 1:2 pp133–54.

—— (1998) ‘Evaluation and the New Public Management: An International Perspective’. EvaluationJournal of Australasia, 9:1&2 pp7–15.

—— (1999) ‘How Do We Know How Good Public Services Are?’, in G. Peters and D. Savoie (eds)Revitalising the Public Service, Montreal and Kingston: Canadian Centre for ManagementDevelopment & McGill/Queens University Press.

Pollitt, C. And Bouckaert, G. (2000) Public Management Reform: A Comparative Analysis, Oxford: OxfordUniversity Press.

Pollitt, C. And Summa, H. (1997) ‘Trajectories of Reform: Management Change in Four Countries’.Public Money and Management, 17:1 pp7–18.

Pollitt, C., Birchall, J. And Putman, K. (1998) Decentralising Public Service Management, Basingstoke:Macmillan.

Pollitt, C., Girre, X., Lonsdale, J., Mul, R., Summa, H. And Waerness, M. (1999) Performance orCompliance? Performance Audit and Public Management in Five Countries, Oxford: Oxford UniversityPress.


Premfors, R. (1998) ‘Re-shaping the Democratic State: Swedish Experiences in a ComparativePerspective’. Public Administration, 76:1 pp141–59.

Prime Minister and Minister for the Cabinet Of� ce (1999) Modernising Government, Cm. 4310,London: The Stationery Of� ce.

Rouban, L. (1995) ‘The Civil Service Culture and Administrtaive Reform’ in B. Peters and D. Savoie(eds) Governance in a Changing Environment, Montreal and Kingston: Canadian Centre forManagement Development & McGill-Queen’s University Press.

Savoie, D. (1994) Thatcher, Reagan, Mulroney: In Search of a New Bureaucracy, Toronto: University ofToronto Press.

Schick, A. (1996) The Spirit of Reform: Managing the New Zealand State Sector in a Time of Change,Wellington: State Services Commission.

Steering Group (1991) Review of State Sector Reforms, Auckland: State Services Commission.Talbot, C. (1994) Re-inventing Public Management: A Survey of Public Sector Managers’ Reactions to Change,

Northants: Institute of Management.—— (1996) Ministers and Agencies: Responsibility and Performance, London: CIPFA.—— (1997) Public Performance: Towards a Public Service Excellence Model, Discussion Paper No.1, Llantilio

Crosseny, Monmouthshire: Public Futures.Task Force on Management Improvement (1992) The Australian Public Service Reformed: An Evaluation of a

Decade of Management Reform, Management Advisory Board, Canberra: AGPS.Ter Bogt, H. (1999) ‘Financial and Economic Management in Autonomized Dutch Organizations’.

Financial Accountability and Management, 51:3&4 pp329–51.Trosa, S. (1995) Moderniser l’Administration: Comment font les Autres?, Paris: Les Editions d’Organisation.Wildavsky, A. (1979) Speaking Truth to Power: The Art and Craft of Policy Analysis, Boston: Little Brown.


Analysis of the Impact of NPM Reform

Documents

Transcript of Analysis of the Impact of NPM Reform