Work package B: preliminary report 2011 ESSnet … · Web viewCountry Activity NACE code...

in partnership with

FSO

ESSNet Profiling of large and complex Multinational Enterprise GroupsFRAMEWORK PARTNERSHIP AGREEMENT N 30621.2009-2009.470

Specific Grant Agreement N° 30621.2009.001-2010.456A project funded by the European Union

ESSnet on profiling large and Complex MNEs

Work package B:

Conceptual framework, methodology,rules and standards

Preliminary report 2011

Version 1.1.p1, 7 February 2012

Drafted by : Partners in the ESSnet on profiling large and complex Enterprise GroupsVersie : 1.1.p1Versiedatum : 07-02-2012

in partnership with

FSO

ESSNet Profiling of large and complex Multinational Enterprise GroupsFRAMEWORK PARTNERSHIP AGREEMENT N 30621.2009-2009.470

Specific Grant Agreement N° 30621.2009.001-2010.456A project funded by the European Union

Drafted by : Partners in the ESSnet on profiling large and complex Enterprise GroupsVersie : 1.1.p1Versiedatum : 07-02-2012

ESSnet on profiling large and complex MNEs Version 1.1p1Date 07 February 2012

document.doc

Version control

Version Date Description Author0.1.p1 15/02/2011 First draft Jean Ritzen0.2.p1 01/05/2011 Updated first draft Jean Ritzen0.3.p1 15/06/2011 Review first draft on language by ONS

Review first draft by INSEEAddition of chapters 3.4a, 4, 13 and 22 by INSEEReview of chapter 4 by CBSComment on added para 3.4a by StatfinAddition of chapters 6, 24 and 25 by StatfinAddition of chapter 12 by ISTATReformat document by CBSUpdated draft

Linda ScottPierre TeilletPierre TeilletJean RitzenTarja HatakkaTarja HatakkaSimone AmbroselliMarcel WieserJean Ritzen

0.4.p1 01/07/2011 Addition of chapter 12 (now 5) by ISTAT Giuseppe Garafolo0.5.p1 20/07/2011 Comments on version 0.3.p1 (except SPEs and

R&D), e.g. tuning with EGR namings (GEN/TEN)Inserting replacement chapter 6.1 WPBUpdated draft

Pierre TeilletPierre TeilletJean Ritzen

0.6.p1 15/12/2011 Updated draft a. with comments on version 0.5.p1 from Helsinki discussions, 30-31 August 2011, e.g. changed order of chaptersb. Some more additionsResult: draft preliminary report 2011

Jean Ritzen

0.7.p1 17/01/2012 Updated draft with comments received on the 0.6.p1 version (Especially chapters 11, 14, 24, 25, 26 and 27

Jean Ritzen

1.0.p1 26/01/2012 Final preliminary version 1.0, updated with re-ceived comments on the version 0.7.p1 (Especially chapters 6.2, 11, 14, 24, 25, 26 and 27

Jean RitzenPierre TeilletTarja Hatakka

1.1.p1 07/02/2012 Replacement of example in the annex VI Dominique Francoz

Version Date Distribution0.1.p1 15/02/2011 ESSnet partners0.2.p1 01/05/2011 --0.3.p1 15/06/2011 ESSnet partners + input workshop Leiden 20-21 June 20110.4.p1 01/07/2011 Not distributed0.5.p1 20/07/2011 ESSnet partners as a basis for the Helsinki meeting, 30-31 August 20110.6.p1 15/12/2011 ESSnet partners for comments (before 6-1-2012)0.7.p1 17/01/2012 ESSnet partners for London meeting, 19-20 January 20121.0.p1 31/01/2012 Coordinator of the ESSnet with a copy to ESSnet partners1.1.p1 07/02/2012 Coordinator of the ESSnet with a copy to ESSnet partners

© Eurostat 2011 Page i


document.doc

Table of content

Preface.................................................................................................................. 1

I Common operational methodology.................................................................31 Exploring references to existing descriptions in the BR manual; Mak-

ing clear the differences and/or the deviations......................................32 The units model described in more detail, with focus on the interna-

tional dimension........................................................................................33 Statistical units in the model and other related units (observation, re-

porting): importance and use...................................................................64 Exploring the concepts of control and proposals for use in the con-

text of EGR and Profiling; the concepts to be used: UCI, GGH or other85 The concept of autonomy in international context, what does it

mean?......................................................................................................106 Classifications (activity and institutional): which, why, where and

how? Single and/or multiple?................................................................207 The core variables and the roles in the BR/EGR context....................288 Classification of changes: which, why and how? Continuity rules.. .359 Analysis of usefulness of units of the new model for SBS, STS, FATS,

Prodcom and NA.....................................................................................3510 The concept of autonomy related to main information and variables:

meaningfulness and availability of necessary information/data. How relates global autonomy to geographical autonomy at MNEs? What are the consequences?..........................................................................36

II Consistency and breaks................................................................................3611 Relationships with used statistical units (Ent/KAU/LKAU) in present

statistics (SBS, STS, FATS, Prodcom, NA) and related to Consistency ESSnet.....................................................................................................36

12 Analysis of potential breaks in time series in changing the used units, aspects of continuity and consistency; implications of chan-ging..........................................................................................................36

III Consolidation aspects..................................................................................4013 Description of basic organisational principles of large MNEs and the

appearances of these.............................................................................4014 How relates consolidated information at the (global) group-level to re-

quired geographical statistics? How to decompose and how to deal with internal flows of goods and services within global groups, in-cluding the aspect of transportation?...................................................42

IV Modes in profiling.........................................................................................4215 The standard and basic steps in profiling large and complex MNEs. 4216 Description of the general principle approach.....................................4417 The difference between top-down and bottom up profiling and how to

combine these modes?..........................................................................44

© Eurostat 2011 Page ii


document.doc

18 The different intensities of profiling (Intensive, light, manual, auto-matic).......................................................................................................47

19 The selection of groups to be distinguished (criteria and numbers), see also WP C..........................................................................................48

20 Priorities and follow up strategies: which group in which intensity with which frequency..............................................................................48

21 Relationship with updating strategy of the EGR..................................48V EU-parts of non-EU MNEs.............................................................................4822 Description of the problem.....................................................................4823 The impact of partly profiling on the statistical units model...............49VI SPEs and other special cases.....................................................................5024 SNA principles and requirements of Special Purpose Entities (SPEs)5125 Treatment of SPEs in profiling; the global and national perspective.5426 Auxiliary activities and R&D according to SNA....................................5727 The treatment of R&D and auxiliary activities, with focus on crossing

border services.......................................................................................5928 Branches and the question of VAT positions.......................................59List of abbreviations.........................................................................................61ANNEX I: BR recommendation manual 2010 chapter 21, the points 21.31

to 21.46.....................................................................................................63ANNEX II: Background information related to autonomy...........................68ANNEX III: Variables for reconciliation with SBS statistics, NA use and

their links with accounting standards:..................................................72ANNEX IV: SPEs: classification tables..........................................................76ANNEX V: R&D problem statement examples..............................................79ANNEX VI: Example of profiled GEG and the impact on business statistics84

© Eurostat 2011 Page iii


document.doc

Preface

The ESSnet on profiling large and complex Multinational Enterprise groups star-ted at the end of 2009. One of the work packages (WPs) is on methodology, the WPB.

The objectives of the Work Package B

To describe the theoretical background on statistical units (enterprise, en-terprise group, reporting unit, observation unit, analytical unit), the rela-tionships between units, the core variables (employees, turnover, value added, economic activity) and their relevance to consistency as well as to the concepts of control (ultimate controlling institution vs. group head), country of ‘decision making centre’, continuity and other ‘coordinating’ characteristics.

To develop a common operational methodology for profiling that is based on the theoretical foundation and takes into account practical is-sues like availability of information in different countries.

To define different approaches to profiling like top-down and bottom-up, the selection of relevant units (cut-off-criteria).

To delineate the non-financial sector (at this stage, profiling will focus on the non-financial sector as far as possible).

To describe the treatment of specific cases like joint ventures, financial vehicles, holdings, auxiliary activities and units, R&D, SPE, families, funds, foundations, associations, co-operatives etc. and also the treatment necessary to profile the EU truncated part of non EU MNEs.

The main 2010 result of this work package is the report on statistical units. The proposals done in this report provides a fundament for the chosen approach in the start of profiling in 2011. A further methodological foundation had the atten-tion in 2011. This are described in this preliminary report. Many issues are dealt with, but also some remain for the attention in 2012, for initial research and de-scription or for further elaboration.

Proposals are not restricted to the analysis of the statistical structure of the Global Enterprise Groups (GEGs). More precisely, also a description is given how the proposed units can support an efficient data collection for the main busi-ness statistics domains; this has lead to the definition of core variables (including priority variables such as number of employees, turnover, value added, eco-nomic activity) linked through accounting relations. The developed methodology is being tested with large and complex MNEs.

© Eurostat 2011 Page 1


document.doc

The results are a product of good co-operation between the working and review-ing partners and with Eurostat within the Work Package. The content of the re-port is shown in the table of contents. Some issues are put in annexes, especially the references to other manuals and/or some background descriptions.

The approach of profiling and the consequences of it are shown in the last annex.



document.doc

I Common operational methodology

Aim: To develop a common operational methodology for profiling that is based on the theoretical foundation and takes into account practical issues like availab-ility of information in different countries. The different levels of statistical units should be considered (Global enterprises (GENs) defined as statistical divisions of the group that can cross the borders, truncated enterprises (TENs) defined at the country level) as well as the feasibility of the collection of information for the enterprise, at least for Option D. An assessment will be done, in this ESSnet, between this proposed type of "enterprise" and the present theoretical or used en-terprise unit.

1 Exploring references to existing descriptions in the BR manual; Making clear the differences and/or the deviations

1.1. In the context of the ESSnet on profiling the chapters numbered 5, 19, 21 and 22 of the 2010 BR recommendations manual are very relevant. Parts of the chapter 21 on Enterprise Group are further elaborated in a separate recommenda-tions manual (September 2010) in which the definition of the UCI (Ultimate Controlling Institute) is elaborated in more detail. In this last mentioned docu-ment also other important terms are defined, like Global Group Head (GGH) and Global Decision Centre (GDC). In the BR manual the glossary provides defini-tions of frequently used terms. Of these, the terms are only defined again if the definitions in this report deviate from the definitions in the BR manual.

1.2 The present BR regulation and BR recommendations do not take in account the most recent developments as achieved from the ESSnet on profiling and the EGR modelling of MNEs1. After formal adaptation of the proposed statistical unit’s model the mentioned chapters must be revised and/or adapted. The concept of the Global Enterprise (GEN)2 and of the Truncated Enterprise (TEN) must be added, with the units definitions and their characteristics.

2 The units model described in more detail, with focus on the international di-mension

2.1 In the report on statistical units, prepared in the first year of the ESSnet on profiling, a statistical units model is presented that can be used for compiling statistics in which data of MNEs are included in a consistent way. For these units

1 The terms MNE and GEG will be used as a kind of synonyms many times in this report, al-though MNE is mostly used in the real world and GEG is a more proper statistical term.2 Abbreviations GEN and TEN are in accordance with the used abbreviations in the EGR



document.doc

the delineation of the Global Enterprise (GEN) is essential with its geographical (national) components, the Truncated Enterprises (TEN).

Because the GENs are the result of the economic and statistical analysis of the structure of the Enterprise Group, and consequently each Truncated Enterprise is viewed purely as the trace of the Global Enterprise in a country in which the GEN is present, a Global Enterprise cannot have more than one truncated enter-prise in one country, just like a global Enterprise Group (EG) can not have more than one truncated enterprise group in one and the same country (on this topic, see also paragraph 3.5).

Diagram: The model

2.2 The proposed definition of the enterprise group (see report WPB 2010) is the following:

The enterprise group is a set of legal units under common direct or indirect con-trol. It mostly appears as a combination of legal units bound together by legal and/or financial links. An enterprise group is to be constituted regardless national borders. Here we speak of a global group. If the group operates in more than one country, this group is mentioned as a Global Enterprise Group (GEG) or Multinational Enter-prise Group (MNE). The territorial or geographical (mostly national) parts of the MNE are the Truncated Enterprise Groups (TEGs). A global group can not have more than one truncated group in one defined geographical area, e.g. one coun-try. A distinction can be made in:


Economic/statistical world (global)

Global Enterprise group

Legal Unit

Legal/administrative world (global)

Local (legal) unit

Global Enterprise

Local unit

Economic/statistical world (sub-global)

Truncated Enterprise group

Truncated Enterprise

Local unit

Legal/administrative world (sub-global)

Legal or operational unit (sub global)

Local unitlegal or operational

SPE SPE


document.doc

- Multinational Global Enterprise Groups (MNE or GEG) and- All resident (or full domestic) Enterprise groupsWith all resident enterprise groups the global group and the truncated group are identical. Global profiling is then identical to national profiling.

According the 2008 BR Regulation enterprise groups can be identified through the links of control between their legal units. In order to delineate enterprise groups, the definition of control set out in point 2.26 of Annex A to Regulation (EC) No 2223/96 shall be used3.“Control over a corporation is defined as the ability to determine general cor-porate policy by choosing appropriate directors, if necessary. A single institutional unit (another corporation, a household or a government unit) secures control over a corporation by owning more than half the voting shares or otherwise controlling more than half the shareholders' voting power. In addition, government secures control over a corporation as a result of special legislation decree or regulation which empowers the government to determine corporate policy or to appoint the directors.In order to control more than half the shareholders' voting power, an institu-tional unit need not own any of the voting shares itself. A corporation C could be a subsidiary of another corporation B in which a third corporation A owns a majority of the voting shares.Corporation C is said to be subsidiary of corporation B when: either corpora-tion B controls more than half of the shareholders' voting power in corporation C or corporation B is a shareholder in C with the right to appoint or remove a majority of the directors of C.”The concept of control is further elaborated in the chapter 5.

2.3 The proposed definition of the enterprise (see report WPB 2010) is:

The enterprise can be either:- a single legal unit (including a natural person) if this LeU is

“independent” or sufficiently autonomous within the enterprise group, or

- the whole of an enterprise group as a set of legal units undercommon control if the group is managed globally and not more autonomous segments can be distinguished, or

- an “autonomous” part of an enterprise group, producing goods or services.

3 Council Regulation (EC) No 2223/96 of 25 June 1996 on the European system of national and regional accounts in the Community



document.doc

It benefits from a certain degree of autonomy in decision-making, especially for the allocation of its current resources4. The enterprise will mostly appear as an organisational unit that can provide meaningful data for statistics.

The enterprise group is the starting point to derive or establish enterprises. The enterprise group can be split into more enterprises if there are more autonomous decision centres. In the case of not two ore more legal units under common con-trol, the enterprise is equal to the single legal unit (including a natural person).

The appearance as an organisational unit also indicates having its own manage-ment.

From this definition, the distinction between the global enterprise and the trun-cated enterprise is: The global enterprise may be part of an all-resident enterprise group or be a single legal unit, if it is not part of a set of legal units under common control. The truncated enterprise is the national part of the global enterprise. This be-longs to the “truncated enterprise group” of the “multinational enterprise group” (MNE). Within an all-resident enterprise group, the global enterprise and the truncated enterprise are one and the same unit. It is very important to stress that the truncated enterprise is not the result of an independent profiling action or analysis of the truncated enterprise group, but it is the national part of a global enterprise, that then automatically is a part of the truncated enterprise group. In profiling this is the most important change in the derivation the statistical structure of a MNE. If a Group is all resident or fully domestic, the global enter-prise and the truncated enterprise are equal.To get internationally consistent statistics also the belonging characteristics to be assigned to the global enterprise and to the truncated enterprise are qualifying. For this see the beneath chapters 6 and 7.

3 Statistical units in the model and other related units (observation, report-ing): importance and use

3.1 In the model of the “diagram 1” three types of statistical units are presented, the Enterprise Group, the Enterprise and the Local Unit. These three units are apparently the same than those to be registered in the statistical business re-gisters according the 2008 BR Regulation.

4 From this definition it may be clear that an enterprise can be a single legal unit



document.doc

But this new model is introducing the international dimension, what is of grow-ing interest in the context of increasing globalisation.

Statistics that are internationally consistent need to be based on relevant (changed from the presently used) concepts. Because many European statistical concepts are prescribed from regulations, these regulations should be adapted too. An example is the Structural Business Statistics Regulation. In this domain, the enterprise statistical unit is the prescribed basic unit for national collection and statistics compilation. It should be changed into the truncated enterprise.The change has no conceptual consequences for the registering of the local unit5.

3.2 In the definition of the enterprise no links with legal units are mentioned. The enterprise structure should show the real economic situation to be described in statistics. The global enterprise structure is the result of an analysis of the real economic and operational structure of the Enterprise Group.

Most logic starting point in the analysis is the presentation of the group itself in its components. We can find the basis of this analysis in the way the group has organised itself in Group Operating Divisions (GODs). These are operational segments displayed in its financial reports. For management reasons, it is expec-ted that the operational segments will have internal supporting information sys-tems for all kinds of product and process information. These are to be used in the analysis of the structure, if these are available for these purposes.

3.3 Next to the statistical units structure of a group we must know the units that can report for the established statistical units. These units are called “reporting units”. These units can be inside or outside the group, e.g. an accountant.

3.4 For some kind of information, the requested data cannot be directly collected for the target statistical unit, but the data may be available from aggregation of underlying (sub) units, e.g. local units of enterprises. These subunits are then called “observation units”. The data of the observation units must be aggregated into data of the statistical unit thereafter. For some physical data the collection must be done at the physical local places, nevertheless this is done by the report-ing unit itself or by the statistical institute according to agreements passed with the group.

3.5 When studying the “feasibility” of profiling, we met what we called the “classification dilemma”: a Global Enterprise which main activity is manufactur-ing industry could at first glance have two activities in a non-UCI country: e.g. one manufacturing and one trading (retail or wholesale); we can classify them either according to the principal activity of the GEN they belong to or according to their own.

5 Nevertheless, for statisticians using legal units, one local unit refers to one LeU. In the BR reg-ulation, the local unit is linked to an enterprise => it mostly results then from the merging of all the local units of the legal units which are included in the same “enterprise”.



document.doc

If we would interpret the two activities as two different TENs, this might lead to a contradiction to the principle included in § 2.1 according to which there is only one TEN per GEN and country. We propose to solve this double dilemma (de-lineation and classification of units) by sticking to the 2.1 choice and to allow creating two KAU (kind of activity units) in the concerned country. As in the 3.4 case, the statistical results can be obtained directly from the TEN reporting unit, or re-built by the statistical institute from other “reports”. This issue will be elab-orated further in the paragraph 6.1

3.6 In profiling we should keep in mind a very important basic principle related to units classification. First of all the statistical unit must be established or con-stituted according to the criteria mentioned in the definition. After establishing the statistical units, these are to be classified according to the appropriate classi-fication systems like size-class and NACE.

3.7 The most important criterion in the definition of the GEN is that of the autonomy in decision making, especially for the allocation of its current re-sources. Activity homogeneity is not mentioned as a criterion as this is also not in the definition of the Enterprise in the 1993 Statistical Units Regulation. Nev-ertheless, it can be mentioned that operational divisions within Enterprise Groups may appear strongly correlated with activity homogeneity in many cases, but not in all.

So: first the unit and after that the classification of the unit!

3.8 The relationships with the other statistical units defined in the 1993 Statist-ical Units Regulation will be dealt with in the beneath paragraph 11.

4 Exploring the concepts of control and proposals for use in the context of EGR and Profiling; the concepts to be used: UCI, GGH or other

4.1 In the BR recommendation manual 2010 a separate chapter 21 deals with en-terprise groups. The content of this chapter is very useful and can stay un-changed for the description of the concepts of control. All these are based on the general definition of control as mentioned in the paragraph 2.2 before. A part of the chapter 21 is attached in the ANNEX I of this report, the points 21.31 to 21.46, and most parts will not be repeated here.

4.2 In the section 21.45 of the BR manual chapter 21 is mentioned that “the statistical concept of the enterprise group is different from the accounting concept, as can be derived from the Seventh Council Directive (see the Annex). In fact, as it is stated in explanatory note 3 of section III C of the Annex to the SU Regulation “this definition (of accounting groups …) is not suitable for statistical analysis because they do not constitute mutually exclusive, additive groups of enterprises. A statistical unit known as ‘enterprise group’ based on the ‘accounting group’ concept must be defined by applying the following four



document.doc

amendments:- consider accounting group at the highest consolidation level (group head) ;- include in enterprise group units whose accounts are entirely integrated in those of the consolidating company;- add majority-controlled units whose accounts are not included in the overall consolidating by virtue of application of one of the criteria allowed by the Seventh Directive, i.e. difference in the type of activity or small relative size;- discount temporary links of less than a year.”

4.3 These four amendments are very important and are confirmed with all IN-SEE 2011 tests with 3 remarks to be added:- the highest displayed (displayable) consolidation level is not always the Global Group head (GGH): this is the case for many “family groups” in which the “family holding” does not reveal any data; this is also the case for state-owned groups, the National State defining several Groups and not consolidating them all; in this type of cases we take the highest displayed level of consolidation as the basis of the delineation of the MNE we intend to profile.- in the present uses of the Directive, “proportional integration” does exist, ac-cording to which a Joint Venture is split 50%/50% between 2 MNEs; it is then very difficult to understand which of the 2 MNEs, if any, has the real control; and it is very difficult to obtain separate information for the JV; but this excep-tion should disappear in 2013, so that we will not try to find a long-term solution for all the related cases.- a concrete criterion for “small size” has still to be defined

4.4 In the paragraph 2.46 of the BR manual operational rules are described for identifying control links: “Control is a complex concept in economic terms. Statistical operational rules need to rely on observable criteria: Proof of control. Therefore it is sufficient that at least one of the following applies, in order to identify a link of direct or indirect control between two legal units:1) A legal unit directly owns more than 50 % of the voting rights of another

legal unit (direct control);2) A legal unit indirectly owns more than 50 % of the voting rights of another

legal unit, through subsidiaries (indirect control);3) Existence of special legislation decree or regulation, which empowers the

government to determine corporate policy or to appoint the directors of the legal unit;

4) A legal unit fully consolidates the accounts of another legal unit, according to the criteria of the Seventh Directive, and no other legal unit consolidates the same legal unit (control by virtue of full consolidation);

5) Administrative sources, collecting declarations in application of specific laws for market regulation, provide the information that a legal unit controls one or a set of legal units, even though it owns less or 50 % of its voting rights (effective minority control) and no other legal unit owns more.



document.doc

Note — It may be possible that two rules, e.g. both cases 1 and 4, could apply simultaneously. As one unit cannot be controlled by two different units, the de facto controlling unit should then be chosen. Case 4 can in general be considered as weaker than 1, because there can be consolidation situations with less than 50% ownership and situations with over 50% ownership without consolidation.”

End of quotation out of the BR recommendations manual.

In global profiling, the concept of control is not restricted to national borders, but is globally border crossing. This solves the problem of identifying relationships of separate domestic subholdings or subsidiaries which are linked together by a common foreign control and so belong to one and the same global and truncated enterprise group.

4.5 More and different appearances of MNEs can be distinguished. In the chapter 13 most frequent appearances are described. The way a MNE is organised, appears and presents itself can influence the resulting way of approaching in profiling. It can be concluded that in specific cases it can be advised that consolidation and/or /control is to be related to autonomous subgroups, e.g. for only financially related groups.

5 The concept of autonomy in international context, what does it mean?

5.A. Introduction

5.1 Autonomy is a crucial criterion in the delineation of the enterprise statistical unit. This criterion must be elaborated further therefore. Autonomy can be ap-plied at several levels, e.g. at the Enterprise Group level and at the Enterprise level.

5.2 From a semantic point of view autonomy (from αυτο – auto, "self" + νόμος - nomos, "law") means "one who gives oneself its own law6“.Autonomy is not a synonym of the word “independence”, its broadness is re-stricted within the constraints defined by the owner and/or the controller. It more relates to the use of a “broad mandate” regarding defined aspects.Kant develops in his philosophical theories7 the concept of autonomy in terms of relations of a person and his freedom: a person is free when bound only by her own will and not by the will of another; her actions then express her own will and not the will of someone or something else.

5.3 In a similar fashion, we may think of a person (physical or juridical), en-gaged in economic activities, as autonomous when it may take decisions (eco-6 Wikipedia7 Stanford Encyclopaedia of Philosophy: Kant Moral Philosophy, 2004



document.doc

nomic - productive – financial – organisational) in its own will and free from ex-ternal influences. A general definition of autonomy could be the following:

DEFINITION 1 - A person (physical or juridical) is defined as autonomous when he (or it) has the control of the use (it may not be the owner from a legal point of view) of the whole productive means, of the whole productive process, of the whole productive outputs of the economic activities in which it is en-gaged.

5.4 This concept of autonomy is too general to be applied to the complex reality of the economic operators. It is necessary to face with a range of issues:1. The boundary of the “autonomous person” especially in the global market.2. The area (domain) in which the autonomy is applied (financial area, eco-

nomic area, organisational area).3. The interactions between the legal constrains and the economic decisions.4. The time period in which decisions are valid (short, medium, long).

5.5 More complex aspects have to be taken in consideration when we need to ap-ply the concept of autonomy to the statistical data collection: 1. concepts of observational/non-observational unit (identification issue), 2. the coherence of data acquired in different contexts – e.g. short term, struc-

tural and national account statistics (statistical coordination issue), 3. temporal data consistency (stability issue), 4. the data coherence between different territorial boundaries, e.g. regional (in-

tranational), national, EU 8 (geographical or spatial issue).

5.6 Last but not least, in the scope of the European Statistical System, the devel-opment of practical and consistent criteria (and not only theoretical definition) to identify the autonomy of a global economic actor has to face with the present Statistical European Regulations, and especially those that were recently ap-proved, like the new System of National Account and the NACE rev.2. and to compare and confront them with the practical use of “operating segments” (GODs) as basis for delineating the “enterprises” (see also the chapter 6).

5.7 This paper, after considering some short elements on the present European Statistical Regulation (see ANNEX II) and presenting some reflections on the concept of autonomy in the global economy (paragraph 5.B), focuses the atten-tion on the operational criteria to identify autonomy in case of complex multina-tional (and national) units (paragraph 5.C). Further aspects dealing with the global/national points of view (paragraph 5.D) and the set of minimum informa-tion needed to guarantee “meaningfully statistical data” are proposed.

5. B. The concept of autonomy in the globalisation world8 in the published Annual report or Reference Document MNEs refer also to “regions” as over-national zones such as Latin America or EMEA (Europe Middle-East Africa). It is the same when reports on globalisation are concerned.



document.doc

5.8 In the modern economic world the enterprises have increasingly chosen to organize and conduct their business operations in the form of a cluster of vari-ous separate corporations (legal units), rather than as a single corporate entity (legal unit)9. The largest enterprises typically evolve as a complex, large-scale business network, where the different parts of the business are allocated to a group of affiliated corporations (subsidiary corporations); global co-ordination is obtained through the submission of such “legally independent” parts to a com-mon economic strategy. The management of the whole is exercised by headquar-ters. “The globalisation describes a process by which regional economies, societies and cultures have become integrated through a global network of communica-tion, transportation and trade of goods and services” (ESSnet on International profiling, WPA report 2010).

5.9 The main element characterising the global economy is the “speed”. Not only in transfer and allocation of financial resources but also as regards the phys-ical ones. The transfer of capital, labour and technology takes place with increas-ing speed. Consider, for instance, as the transfer of entire production plants take place relatively quickly (in less than a year) unthinkable only at the end of the last 20th century. Also the very rapid developments of the Information and Com-munication Technology heavily contributed to this and created many necessary conditions. The decisional processes have necessarily to be accelerated. For this reason the new organisational systems of the complex and global enterprise are the so called “short chain organisation”, that permits to the top management to collect all needed information in short time to take its decisions. In this situation the control of parent corporation over the subsidiaries is not lim-ited on the “financial or strategic” aspects but focuses more on decisions of re-sources allocations, markets where to produce and sell goods, price strategies, communication and advertising strategies,….

5.10 In the global economy the short/medium/long time and strategic/current de-cision loose their meaning: by definition the speed reduces the time. In this way there may be an increasing reduction of autonomy (and responsibil-ity) of the subsidiary corporations in favour of the headquarters or of higher but intermediate hierarchical levels of the group where the decisions (at short/me-dium/long time) are being taken.

In case of multinational enterprise (or national enterprise group) the headquar-ters of the EG are the only decision centre that may take all types of decisions – strategic and current - in own will and free from external influences, because it has the control of the whole productive means, of the whole productive process,

9 on this topic see also chapter 13 on basic organisational principles and appearances of large Multinational Enterprise Groups



document.doc

of the whole productive outputs and of the whole financial processes of the eco-nomic activities in which it is engaged.

5.11 This cluster vision (instead of the atomistic one) of the economic actors it is not only a “common sense” in a theoretical and economical point of view, but above all, it is a reality that must be governed by laws and regulations. In fact, several European countries in their legislation recognize the validity of the so called “domain contract” or “hierarchical coordination contract”. With these contracts the parent company buys the right to issue directives to the subsi-diary company also detrimental to the latter and, in some case (or in some na-tional laws) without any need for a compensatory advantage. “These typologies of contract create an alienation of the corporate government characterized by the fact that: a). A company acquires the right to give directives to the directors of another (subsidiary) company even if detrimental (and in contrast with the social interest of the subsidiary), provided that the interests of the dominant society or of an-other company of the group are honoured; b) Members of the governing body of the company (subsidiary) are obliged to execute these directives, going free from liability for damages that may result from their implementation.”10 Here we also see the appearance of legal constructions to avoid or minimise risks and to minimise other “disadvantages” like tax paying. These constructions are mainly beyond pure real economic reasons.

5.12 Another important area of intervention in national legislation (and in the European Union) is the correct pricing identification in the transfer of production means (goods, services, labour and financial capital) between companies in the same group. Through the “transfer price” mechanism the prices of intra-group transactions are determined on the basis of evaluation parameters anchored to the general needs of the multinational group from a management, organizational and tax viewpoint, instead of under market conditions. Many governments have adopted transfer pricing rules that apply in determining or adjusting income taxes of multinational taxpayers. OECD has adopted guidelines followed, even if in part, by many of its members. Rules are provided to permit the tax authorities to adjust those prices where the prices charged are outside an arm’s length range.

The law recognizes and governs the fact that the global economic actors central-ise their decisions and operate transfer of production means between their subsi-diary units without pure economic evaluation.

5.13 Considering the accounting point of view, the IFRS 10 has the objective to establish principles for the presentation and preparation of consolidated financial statements when an entity controls one or more other entities. The definition of 10 U. Tombari - Autonomia privata e gruppi di imprese - http://www.notariato.it/it/primo-piano/studi-materiali/


http://www.notariato.it/it/primo-piano/studi-materiali/

http://www.notariato.it/it/primo-piano/studi-materiali/


document.doc

consolidated financial statements is: the financial statements of a group in which the assets, liabilities, equity, income, expenses and cash flows of the par-ent and its subsidiaries are presented as those of a single economic entity

“The Standard: requires a parent entity (an entity that controls one or more other entities) to

present consolidated financial statements defines the principle of control, and establishes control as the basis for con-

solidation sets out how to apply the principle of control to identify whether an investor

controls an investee and therefore must consolidate the investee sets out the accounting requirements for the preparation of consolidated fin-

ancial statements.” [IFRS 10:1]

5.14 The main concept to realize the consolidated financial statements is the Control of an investee: “an investor controls an investee when the investor is ex-posed, or has rights, to variable returns from its involvement with the investee and has the ability to affect those returns through its power over the investee”. [IFRS 10:5-6; IFRS 10:8]Where power means: the existing rights that give to the investor the current abil-ity to direct the relevant activities of an investee. Power arises from rights. Such rights can be straightforward (e.g. through voting rights) or be complex (e.g. em-bedded in contractual arrangements).“A parent must not only have power over an investee and exposure or rights to variable returns from its involvement with the investee, a parent must also have the ability to use its power over the investee to affect its returns from its involve-ment with the investee.” [IFRS 10:17].

The definitions of “control” and “power” proposed by the International account-ing principles show as the subsidiaries (identified as investees) are considered just a way for the investor (the global group headquarter) to affect its returns having the total control of the activities of the group.

5. C. Operational criteria and rules to identify autonomy

5.15 Issues in identification of the statistical “autonomous” units

The ESSnet on profiling (WPB Report on statistical units, 2010) has proposed the following definition of the Enterprise Group:The enterprise group is a set of legal units under common direct or indirect con-trol. It mostly appears as a combination of legal units bound together by legal and/or financial links.



document.doc

and the following definition of the enterprise:“The enterprise is a single legal unit (including a natural person), an enterprise group as a set of legal units under common control, or a part of an enterprise group, producing goods or services. It benefits from a certain degree of autonomy in decision-making, especially for the allocation of its current re-sources. The enterprise will mostly appear as an organisational unit that can provide meaningful data for statistics.”

Two levels of autonomy can be distinguished, one for the Group and one for the Enterprise. The group has full autonomy at the highest level and exercises this mainly and at least for strategic and long term decisions. The group head only needs to give account and to report for its decisions to the owners or to the board of shareholders of the group and in direct contact. A second and lower level of autonomy is that related to the allocation of current resources in the production processes. This is to be understood from a more operational view. If the Enter-prise is a part of the Group the Enterprise management has to give account and to report to Group management In terms of statistics: for statistics related to large and strategic, mostly financial, decisions, the Enterprise Groups is the most eligible statistical unit and for stat-istics related to current production the Enterprise is the most eligible one.

In some countries both kinds of statistics are part of the statistical programs and statistical units are distinguished and applied in this way. The distinction does not appear in the statistical program and regulations of the EU (Eurostat) till now. The European system is based on the enterprise related statistics.

5.16 The main criteria to identify autonomy

On the levels in a (large and complex) organisation where the important de-cisions are taken, the following kinds of decisions are assumed to be relevant and could be considered to be criteria in checking.

For the Enterprise Group:- mergers and deconcentrations, i.e. the authority to buy (stocks of) other

Legal Units (being or not autonomous enterprises) or sell parts of the Group itself;

- income assessment, i.e. decisions on keeping reserve funds, large unex-pected depreciations, etc.;

- income distribution, i.e. the allocation of the generated income, e.g. from production, to shareholders, investments or reservations;

- financing of new projects or termination of ongoing ones.



document.doc

For the Enterprise:- fixed capital investments within certain limits for the running production;- recruitment or dismissal of personnel;- raises in salaries and other remunerations;- choice of the assortment of products;- size of the production;- provisions for doubtful debtors;- pricing of the output/products;- choice of the suppliers.

The Enterprise Group is market oriented by definition. For the Enterprise market orientation is also a requirement (with some exceptions when quasi-enterprises within a Group must be distinguished)

5.17 It is, of course, to be realised that none of the mentioned criteria are abso-lute. A production manager may be allowed to decide to invest in new ma-chinery but he will not be permitted to build a whole new complex. Actual sales prices will, taking into account the market, be set by the sales department of the production unit, but the price policies may be made by the overall management of the production unit as a “strategic” issue. The board of directors of the Group probably has the authority to withhold large amounts of reserves but the share-holders may not like it, if they are not paid any dividends at all. So everything has to be seen in its perspective, which it is exactly what makes it so difficult to establish straightforward rules. Autonomy and the level of it mostly seem from management contract between the Enterprise management and the Group man-agement. On the one hand we have mandated responsibilities, on the other duties to give account and to report on results and performances.

5.18 In order to be able to operate as an autonomous unit it is necessary to have available information systems to control and to report. From these systems meaningful real economic figures can be provided, also for statistics. If informa-tion systems, maybe only for internal use, are lacking, then the existence of an autonomous unit is very doubtful. A first hold in the analysis of a large Enter-prise Group into Enterprises is the existence of Group Operating Divisions (GODs). These can nearly almost be considered to be autonomous. But, depend-ent of the statistical needs, it can be necessary to explore the capability of GODs to be “Enterprises” or reversely Enterprises can be defined at a lower level within the GODs, keeping the basic principles for criteria to be autonomous. A good conclusion is only possible after consulting the Enterprise Groups and agreeing with it, but with the requirements from statistics in mind.

5.19 The analysis of multinational enterprise groups should start regardless the territorial boundaries. If territorial boundaries belong to the organisational cri-teria of the Enterprise Group, they still have to be analysed to check if they fit



document.doc

with requirements from statistics. It is often easier to judge that a potential unit is not an enterprise because it does not fulfil some basic requirements (negative ap-proach) than to follow a positive approach to confirm it as adequate.

5.20 The BR recommendation manual (chapter 19) suggests three different oper-ational criteria can be applied to identify an enterprise as autonomous actor in the economic world:1. An enterprise has accounts at its disposal2. The operations of an enterprise are managed in an integrated manner3. The enterprise is market-oriented.

In some way these criterions summarise the more detailed previous treatment re-lated to the aspect of autonomy. Also here the distinguished criteria must be con-sidered in their combination and not solely.

See further the BR recommendation manual 201011.

5.21 Identification rules

The following operational rules meet mainly the need to identify autonomous statistical units. The aim is to produce statistical information consistent at EU level, and as close as possible to the economic reality.

Identification rule 1 - A single legal unit not controlled by any other legal units is an autonomous enterprise.

Identification rule 2 - An enterprise group (Multinational or National) is autonomous in itself; it will be equal to an Enterprise if there are no reasons to split it into more autonomous parts.

Some groups have a complex structure: under a “financial holding” they can present more sub-groups that operate in different and distinct economical domains. These sub-groups often have sub-consolidated accounts and are listed on a regulated market: Identification rule 3 - In case of a conglomerate group12 each sub-group, if and only if it produces sub-consolidated accounts, has at least to be treated as an in-dependent Enterprise and the intra-flow transactions between the sub-groups are not take in consideration

Some (Multinational or National) enterprise groups (or sub-groups) may decide to organise their activities in different “Group Operating Segments” (“GODs”) or “business lines”.11 Business Register recommendation manual, § 19.32.12 a conglomerate groups exists if the enterprise group is engaged in different activities that have no links except of shareholding with eachother: no product, no process, no salesforce, no brands etc in common.



document.doc

Identification rule 4 - If the GODs are observable (their existence are proved in the reporting system of the Group), they are the starting point to identify and delineate the “autonomous enterprises”.

Considering the market oriented criterion:Identification rule 5 - If one GOD inside a Group (or sub-group) represents a way on which the “group” itself performs a specific productive activity (it means that it contains all legal entities and their parts serving to this scope), it will be considered as a separate (autonomous) global enterprise. .

Identification rule 6 – If inside a GOD there are more (market oriented) autonomous parts for which meaningful statistical data can be provided, these have to be considered as autonomous units and should be proposed to be global enterprises. Identification rule 7 - If one GOD perform support activities (as ancillary unit: administration, IT, wholesale, transport… or as artificial subsidiary corporation: own land, building, equipment or being nominal employer) for other GODs and the main part (more than 50%) of its transactions is with other GODs it has to be considered as a serving GOD. In this case it is not considered an autonomous enterprise. The outputs of the serving GOD have to be considered as intermediate consumption for the other GODs and its accounting data have to be consolidated with the data of the autonomous enterprise (global or truncated) it serves, in principle. However, there can be exceptions, e.g. for R&D and for holding activities which are provided at market prices. Here it can be considered to define quasi enterprises, which are dealt with in statistics as enterprises, see rule 8.

Taking in consideration the NACE13 manual, the following rule is valid:Identification rule 8 –The R&D activities, if separated business line inside the EG, have to be considered “autonomous” GEN and from there TEN.

Where possible the SNA14 and NACE15 manuals should be taken in consideration in the decisions on the establishment of the enterprise statistical unit

5. D. Autonomy at global and national level (still to be elaborated further)

5.22 The market orientation is an important criterion to identify the autonomous global enterprises. The question is then: in which market and for which market we need to define these units.It is possible to speak in terms of local, regional, national, European, global market. 13 Nace introduction guidelines, § 5314 System of National Accounts – 2008, 5.21-5.28 15 Nace introduction guidelines, § 77



document.doc

In particular, and for our purposes, autonomy at global and national level could not be considered in the same manner. The boundaries of the European country have to be taken into account and to be related to the concept of “residence” in the National Accounts.An example can help:A, B and C are three legal units controlled by the same MNE. A and B are manufacturing units; C is a wholesale unit. They are part of the same GOD. The combination of A, B and C has to be considered as one Global Enterprise.A is resident in country X, B and C are resident in country Y. If there are intra-flow transactions between B and C, C can be considered as ancillary unit and the Truncated Enterprise is the sum of B and C. The NACE TEN code = NACE GEN code. If there are transactions between A and C (and not with B) this transactions are import/export activities and they are evaluated at arm’s length (for tax purposes). In this case we still have one GEN with one TEN in country X (legal unit A) and one TEN in country Y (combination of legal units B and C).

5.23 Autonomy at the global level is determining in the delineation of the Enterprise statistical unit GEN. The truncated national or geographical parts of this are not necessarily autonomous because these are under the operational leadership of the GEN management, even if this is under local legal conditions.

5E. Autonomy in relation to the provision of meaningful statistical data

5.24 The provision of meaningful economic figures is the purpose of business economic statistics. The best way to comply with this purpose is to find a fit with the information that is compiled for management purposes within the MNE and for the operational management of it. Also information for shareholders and/or the board of commissioners reflects real information on thee operations of the MNE. Managers of autonomous units will have meaningful and detailed in-formation available from accounting systems and from other internal informa-tion systems. The best way to get meaningful data is to connect the need for stat-istical information to the available accounting and information systems and to extract data from these in data collection. At the highest level and for the Groups Operating Divisions accounting systems are being standardised according the IFRS, but more detailed information systems related to more detailed autonom-ous units may be available.

6 Classifications (activity and institutional): which, why, where and how? Single and/or multiple?



document.doc

6.0 As is mentioned in the chapter 3, first of all the statistical units must be con-stituted and established according to the criteria in the definitions. Once these units have been established, the units must be classified according to the appro-priate classifications. The main ones are the NACE and the size-class. In this chapter the issue of the activity classification (NACE) will be dealt with in more detail.

6.1 The issue of activity classification

6.1.1 Concerning the MNEs we face the question on how to classify the global enterprises (GENs) and the national parts (TENs) of global enterprises. As pro-posed in the feasibility study, we will in practice classify the national parts (TENs) either according to the principal activity of the GEN they belong to and/or according to their own activity. This § treats of the justifications to do so, the procedures we propose and the constraints we need to respect.

6.1.2 To begin the analysis, we need, from a conceptual point of view, to create the conditions for gathering significant European statistics, to find consistent solutions for classification:- in the framework of the presented SU model (see §2.1); - with reference to the NACE prescription (NACE rev2. Chapter 3.1), which states that every statistical unit is classified according to the following rule: “One NACE code is assigned to each unit recorded in statistical business re-gisters, according to its principal economic activity. The principal activity is the activity which contributes most to the value added of the unit.” - by using the method stated in the Business Registers Recommendations manual 2010 (also called “top-down” in the sense that a descending way of the NACE is used in the classification procedure from the sections (capital letters) to the classes (4 digits)).16

6.1.3 In consequence, the activity classification of GENs and TENs needs that we first interpret their definitions and delineation. Thus, the question of classific-ation is strongly related to the question of autonomy:

- as autonomy is compulsory for each GEN, its classification at its own prin-cipal activity is straightforward when delineated, but leaves open the relation with the criteria that are used to identify the autonomy (see the previous chapter 5).The criterion of autonomy is the (only) determing one, but the detail should be the result of analysis and tuning with the MNE.

- as autonomy is often absent in the case of the TENs, we have to treat separ-ately the question of their own classification. But we also have to treat the fol-lowing question: “how shall we classify TENs of the same GEN in different countries in order to add them up to European aggregates consistent with the

16 This does not seem to create problems and is explained at the right end of this paragraph 6.1.



document.doc

GENs they belong to”? The answer is in the dual coding approach, see 6.1.13 and 6.1.14.

6.1.4 According to the SU model that gives no place to homogeneity, GENs can be homogenous or heterogeneous global units in terms of activity. GENs should be considered to be real actors in the economy, not questioning the question of homogeneity, but only that on the autonomy regarding the current resources, (see the chapter 5)17. 6.1.5 The rules concerning the autonomy are, up to now, only related to the abil-ity to decide at least in production and commercial fields for a specific manage-ment that has a reporting system: they lead us to a solution with more heterogen-eous GENs, than previous ways of classification, be they national or not18. For example, if we follow the market orientation criterion that the autonomy implies usually, we cannot consider as a rule that head office activities, R&D activities and all other ancillary or auxiliary activities are autonomous GENs. It is the same for vertically integrated activities, even if they are operated in different countries. Thus, at the same time, those activities need to be included in the GEN or to be defined as separate quasi GENs.

6.1.6 As TENs are simply truncated parts (“local” units, if we state that a coun-try is “one” address) of global enterprises, this implies that we can have only one TEN per GEN and country. There is an analogy for what is currently done when local units are designed from the (national) enterprises. There can be only one local unit in the same address.

6.1.7 In practice, the availability of sufficient accounting-type or production-type data for surveying or gathering administrative files is a leading and substan-tial aspect of the enterprise definition in profiling GENs. Thus, it is to be sugges-ted utilising the IFRS when delineating GENs of MNEs, because IFRS implies a relationship with the management structure and includes basic information on operational divisions or segments. The operational divisions are seen as a good starting point for delineating GENs, even if in some case we end with GENs that are parts or sums of them.

6.1.8 In consequence, if we wanted the GENs to be as homogeneous as possible (close to global KAU), we would have to accept less information per GEN for

17 As many statisticians link the question of homogeneity to the kind of activity units (KAUs), we can state that the SU model itself does not indicate the relationship of the GEN and the KAU. Thus, from a theoretical point of view, GENs (and TENs) may consist of different KAUs. In pro-filing MNEs KAUs and LKAUs are considered to remain only national units.Several KAUs can exist in the same TEN. This issue is let open (not treated) in this report.18 It may occur that within group operational segments of a MNE more detailed autonomous parts can be distinguished. These should be taken under consideration then to be a GEN within the GOD. It can automatically lead to a more homgeneous GEN then.



document.doc

considerably more numerous GENs per MNE than when accepting more hetero-geneous GENs.

6.1.9 Another probable consequence is that, with heterogeneous GENs, the need to split legal units is less likely than with homogenous global GENs. This con-clusion is based e.g. on the assumption that many (but not all) MNEs organise their legal structure as much as possible in coherence with their management scheme.

6.1.10 As a summary about GENs, and to answer to the statisticians’ question about homogeneity, our conclusion is that there is no direct way to include more NACE homogeneity in the GEN definition that the homogeneity potentially exist-ing in the “real world”.The solution related to the question on homogeneity should be found in the ap-plication of other and nationally used statistical units which are in accordance and consistent with the GENs and TENs. The existence of KAUs and LKAUs among the statistical units creates the opportunity to gather data based on the wished (more) homogeneous units.

6.1.11 Up to now, the SU model does not give clear lines to determine the prin-cipal activity of the TENs included in one GEN:- of course it is possible that TENs are purely internal to a MNE (e.g.: the pro-duction plant, which production is sold exclusively abroad via foreign commer-cial affiliates; other e.g.: a production completely integrated in the MNE by for-eign affiliates etc) => their own activity will compulsorily be different of this of the GEN they belong to.- it is also possible that ancillary or auxiliary productions serve mainly affili-ates of the same MNEs in other countries; we have a turnover, which is not an internal flow inside the country of its production, but which is at the global level of the MNE => their own activity will be among auxiliary or ancillary activities (not displayed at GEN level); - it is frequent that a TEN involves itself in the (wholesale or retail) distribution of products of other TENs of the same MNE => their activity is absorbed by the manufacturing activity at GEN level.

6.1.12 so at the end, the key question remains about whether it possible to create a good system of classification both from theoretical and practical points of view ensuring sufficient consistency between GENs and TENs; this leads to the fol-lowing proposal: - if the GEN is present in only one country, the activity code of the TEN is equal to that of the GEN by definition. - in other cases, there is a necessary rule: that the activity code of the GEN is consistent with that of all the TENs together (general bottom up approach in activity classifying). For this purpose we need to define a decision tree; it could be the following:



document.doc

- apply the NACE rule directly to the GEN, with predominance in the analysis given to the marketed activities; 19

- define all the activities present in the GEN, including the non marketed ones (either because they are ancillary or auxiliary, either because their sales are internal to the group, or because they are vertically integrated even in the same local unit)

- apply the NACE rule to each TEN;

6.1.13 It is still to be decided which is the best proxy for the VA (Value Added), as well for GENs activities as for TENs: employment, an additive variable, would be better than the turn-over which arithmetic is very complex and not well documented.

- this procedure ensures that the own principal activity of each TEN is, in eco-nomic meaning, consistent with the GEN principal activity20;

- to ensure the European consistency of the statistics based on the GEN/TEN principal activity, each NSI should collect both classification: this of the TEN and this of the GEN it belongs to; and make a double set of statistics (a na-tional one based on TENs own principal activities; a European based on GENs’). This can be indicated as a dual or double coding approach. This ap-proach can be extended into multiple coding if we need statistical calcula-tions for more than 2 geographical areas, e.g. national, European, global.

6.1.14 Aggregation of national figures according the TENs own principal activit-ies would not fit with EU figures. The double coding intends to allow: - a purely national calculation, in which the main activity is fixed according to the national viewpoint (and consequently may be different from the TEN of a country to the TEN of another in the same GEN: e.g. manufacturing in one, wholesale in another, the GEN being classified as manufacturing); - a calculation according to the global coding of the GEN, in which all TENs are classified with this only NACE => the EU figure is exactly the sum of the TENs figures belonging to the same GEN, but only for the variables which can be summed up. In the case of variables which must be consolidated like turnover, the meaning of consistency is different. Consolidation has to be applied accord-ing to the geographical area for which statistics are being compiled.

6.1.15 The "Top down" (related to the different levels of the classification) method as stated in the Business Registers Recommendations manual 2010: we

19 In the case of GENs (and only in this case) if ancillary/auxiliary activities or integrated activit-ies have appeared in the process, we suggest that their specific data (on VA for example) will be reclassified with the activity they are used by or in which they are incorporated.20 If we found that this top-down procedure is too heavy, in particular for the country of the UCI, we could let each country classify its own TENs; but in this case we would have to build consist-ent activity tables, relating for example the frequent presence of NACE 2910 Motorcar industry with NACE 4510 Motorcar distribution or selling whether NACE 2410 Iron and steel is consist-ent with NACE 2550 Stamping



document.doc

suggest to use this approach to classify as well GENs as TENs, at their own main activity, in addition to the previous procedure determining their “basic” NACE activities: "The top-down method follows a hierarchical principle: the classification of a unit at the lowest level of the classification must be consistent with the classifica-tion of the unit at the higher levels".

6.1.16 Below are the steps that the hierarchical principle follows:1. Identify the section which has the relatively highest share of the value added.2. Within this section identify the division which has the relatively highest share of the value added within this section.3. Within this division identify the group which has the relatively highest share of the value added within this division.4. Within this group identify the class which has the relatively highest share of value added within this group.

The next example illustrates this approach.



document.doc

Example: a unit carries out the following activities (shares in terms of value added):

Section Division Group Class Description of the class ShareC 25 25.9 25.91 Manufacture of steel drums and similar con-

tainers10%

28 28.1 28.11 Manufacture of engines and turbines, except aircraft, vehicle and cycle engines

6%

28.2 28.24 Manufacture of power-driven hand tools 5%28.9 28.93 Manufacture of machinery for food, beverage

and tobacco processing23%

28.95 Manufacture of machinery for paper and pa-perboard production

8%

G 46 46.1 46.14 Agents involved in the sale of machinery, in-dustrial equipment, ships and aircraft

7%

46.6 46.61 Wholesale of agricultural machinery, equip-ment and supplies

28%

M 71 71.1 71.12 Engineering activities and related technical consultancy

13%

Identify the main Section among Section C - Manufacturing (52%), Section G - Wholesale and retail trade; repair of motor vehicles and motorcycles (35%)Section M - Professional, scientific and technical activities (13%).

Identify the main Division within main Section C: Division 25 Manufacture of fabricated metal products,

except machinery and equipment 10%Division 28 Manufacture of machinery and equipment n.e.c. 42%

Identify the main Group within the main Division 28:Group 28.1 Manufacture of general-purpose machinery 6%Group 28.2 Manufacture of other general-purpose machinery 5%Group 28.9 Manufacture of other special-purpose machinery 31%

Identify the main Class within the main Group 28.9:Class 28.93 Manufacture of machinery for food, beverage and tobacco processing 23%Class 28.95 Manufacture of machinery for paper and paperboard production 8%

Therefore the correct class is 28.93 Manufacture of machinery for food, bever-age and tobacco processing, although the class with the biggest share of value added is class: 46.61 Wholesale of agricultural machinery, equipment and sup-plies".



document.doc

6.2 The issue of institutional classification

The classification according the SNA and ESA institutional sectors is mainly ap-plied for the Enterprise Group unit.

6.2.1 The present context:

The profiling process described previously might be applied to all types of MNEs, including these which participate to the financial system (“financial in-termediation” e.g banks, insurance or financial auxiliaries). But the financial sys-tem is highly regulated, its main activities are submitted to the consent of regu-latory authorities prior to their beginning; they are also permanently supervised by authorities able to set management and structure conditions. In consequence the perimeter of the regulated activities is largely in the hands of specific gov-ernmental bodies, be they known as “independent agencies”. This perimeter does not evidently fit with the Enterprise Group structure: there is the need of a deep insight in how to define the perimeter of the cluster of control and to delin-eate the Global Enterprise Group. This insight is not yet sufficiently available..

According to the SNA and to the ESA, the activity classification of the financial intermediation is combined with their institutional sector classification, different from the institutional sector of other private businesses. Once more, the way the insitutional sector classification and the MNE analysis fit to each other is to be elaborated further in the future.

For all these reasons, in the initial stage of profiling, it was decided to concen-trate on the private non-financial enterprises sector. But this decision leaves open issues or questions such as what to do with financial parts of non-financial MNEs (or how to deal with significant non-financial parts of financial MNEs, where they are important). In a number of countries (e.g. The Netherlands) the compilation of statistics re-lated to financial institutions is the task of the Central Banks. The NSIs include these statistics as these are in their own statistical system, e.g. the National Ac-counts. On the contrary, over the past few years some NSIs have successfully profiled on a national basis (UK/ONS21), or have taken insight (INSEE22) in, a number of banks, real estate companies, investment companies and other busi-nesses classified in the financial sector. More close cooperation between NSIs and the Central Banks started for the building of the EGR.

21 including Royal Bank of Scotland, Lloyds TSB, Barclays Bank and Santander-Abbey Na-tional.22 Including BNP-Paribas, Société Générale



document.doc

This chapter will concentrate on concrete questions coming from this context, with the main aim to gather remarks or experience from NSIs.

6.2.2 Facts from experience:- large MNEs of the financial sector have some of the most complex structures. This results for example for the UK in the necessity to find several high level correspondents spread over the whole of the country to organize each profile; and in a specific group structure in France , in which the GGH is not a holding as usual, but the largest Legal Unit, simultaneously involved in operational activities of all the “” no “enterprise” the largest unit being split in different part belenging to the various “enterprises”of the group (the only exceptions of this general implication coming from regulatory forbidding). - The information coming from private providers is not accurate, as they may treat financial loans as ownership links.- The respective fields of structural surveys and of administrative (and super-visory) information do not easily fit to each-other; and if so the skills of the “in-dustrial” and of the “financial” staff in the NSIs are so different that dialogue proves to be difficult Nevertheless:- there are significant “banking” or “insurance” activities in non-financial groups: for financing purchases or for financial management of the the inde-pendant dealers in motorcar industry, distribution, for short term or revolving loans in retail networks, for long term financial loans in equipment industries etc. - there are also very significant non financial “enterprises” in MNEs that, as a whole, are clearly financial: operational leasing of motorcar, trucks and equip-ment (computers, machines) often including management, maintenance and staff and not only financing the purchase, real estate management and renting as well as development of building projects and building and construction23 - there could be very specific cases such as the “Caisse des Dépôts” in France, which is legally not financial but submits voluntarily to financial and banking regulations and which has also the largest French companies in venture capital (not classified financial by the supervisory regulations), in real estate (rental of offices and housing), “private public local and long distance transportation”.

6.2.3 proposals: - the list of MNEs to be profiled at European level in the initial stage will not include financial groups (eventual exceptions to be decided).- financial parts of non-financial MNEs will be treated at minimum. This in-cludesa. the delineation, in agreement with the concerned groups, of the “financial en-terprises” (GENs) and, if possible, their links with the Legal Units

23 in FR the largest operator in these fields is owned and controlled by a banking group



document.doc

b. the discussion with the concerned services of the NSI and of the National Bank on the financial clusterc. proposals on how to ensure consistency with NA on the Statistical units and on how to treat eventual secondary non-financial activities in financial enter-prises.

7 The core variables and the roles in the BR/EGR context

7A Introduction

7.1 We must know ourselves as statisticians: what do we want to collect and what statistical information do we want to compile and publish. We need well defined statistical methodology in terms of units, definitions and so on, but we must also speak the language of the business. We must have knowledge of their accounts, which many times is their source for reporting data for statistics. Core variables in the accounts do not coincide with the needed core variables in stat-istics, like the SBS. So we must be aware of possibilities of translation selected core variables from business accounts variables into statistical variables. Necessity and availability of needed core-variables is the subject of this chapter. A problem that may rise is the splitting of the global data available in the ac-counts of MNEs into national outcomes. Instead of working bottom up from na-tional data into global, it could be preferred to do the other way around, from global into national, and this all in relation to the data collection methods to be used (bottom up or top down).

7.2 In this section the issue is restricted to the core variables related to the En-terprise Group and Enterprise statistical units. The main focus will be on the En-terprise statistical unit (Global Enterprise and Truncated Enterprise), but here also attention will be given to their relationship with the Enterprise Groups unit and to the belonging core variables. No distinction is made in the present situ-ation of the EU statistics, because in the present ESS the Enterprise Group is not used in regulations for statistics, although the Enterprise Group is to be included as a (statistical) unit listed in the Business Register.

7.3 A distinction should be made between statistics describing the production processes (GDP related income generation) and statistics describing the financial processes (incl. income distribution or distribution of the income generated in the production processes). The distinction mainly relates the aspects of autonomy which can differ for the different mentioned processes. - Statistics on financial processes should be based on statistical units that are autonomous to the decisions related to the financial processes. These are the Global Enterprise Groups, with a differentiation to the Truncated Enterprise Groups.



document.doc

- The statistics on the production processes should be based on the statistical units that are autonomous to the decisions related to production processes. These are the Global Enterprise, with a differentiation to the Truncated Enterprises. In this approach, the P&L-accounts (Profit & Loss-accounts) are the main basis for the statistics on the production processes (mainly SBS), so the variables for the Global Enterprise must be collected or calculated statistically from these ac-counts, with a differentiation to the Truncated Enterprises.

7.4 The A&L-accounts (Assets&Liabilities-accounts) are the basis for the statist-ics on the financial processes, so the variables for the Global Enterprise Group must be collected or calculated statistically from these accounts, with a differen-tiation to the Truncated Enterprise Groups. Variables related to “physical” investment, GFCF, or tangible assets are specific: as mentionned previously they are mostly part of the autonomous decisions of the “enterprises”; but they belong to the balance-sheet accounts. So that we need at least to locate them in these accounts and to translate them from the account-ing language to the statistical one.

7.5 If no distinction is made between the different processes, as is not done in SBS a compromise must be found, but it is still useful to have the difference as a starting point.

7.6 For each statistical unit we should establish (see 7.14):- Variables for identification; - Statistical frame variables for underpinning data collection (e.g. for stratifica-tion purposes)- Variables, also important for statistics compilation (e.g. NACE-code, institu-tional sector code);- Variables for units’ continuity analysis- Information on mutual relationships of statistical units within the Enterprise group.These types of variables must be stored in the frame (EGR and/or national BRs). It is to be repeated here that there is a distinction between the core variables which are essential in profiling and the centralised data collection for statistics (which is not profilers’ responsibility).

7.7 In addition we have to check the availability, at the GEGs or at the GENs and/or TENs levels, of the necessary data for statistics compilation (for the spe-cified stakeholders) related to the adequate statistical unit. Availability can be in direct data collection or by statistical estimation from available records.The next steps of investigation could be: - to check these of the variables that we can find at the GEN or the TEN level

and also these which are not ready-made but could be calculated on request by the enterprises.



document.doc

- To discuss which is the better way(s?) to gather them for SBS and NA pur-poses.

7B Contexts: the BR/EGR and the statistical domains:

7.8 As suggested in the title of this §, core variables are implied in the profiling activities in which they play important roles. Two kinds of roles have to be considered:

- in the 1st role, some of the variables are usually present in registers; either as o identification or checking data, permitting

to validate the content of the register and/or to assess the continuity of its units and/or to manage the links between the different kind of units (e.g. enterprise

with legal units or with the enterprise group) or as o statistical frame variables, underpinning data collection (e.g. permitting to

split the populations in significant sub-populations - stratification purposes including NACE-codes and institutional sector codes - or as criteria for sampling purposes)

- the 2nd role is specific to the “enterprise” Statistical Unit; this SU is purely statistical, as it will be created to allow a better economic description of the “business world”, which is no more well described when relying on the Legal Units.

7.9 In this context, it is not longer possible to act as usual on a national basis: ask the reporting units only about the value of such and such variable about the SU they answer for; in this case, it is necessary to split the approach: - first to check if, or not, the variables are disposable - and, second, if or not, their value is significant from an economic point of

view.

7.10 Even more difficult, for the 1st time, on an “industrial” basis, the “enter-prise” SU is being tested for its capabilities from an European (and not longer a national) point of view: are we able to describe consistently, through the data we ask them for, MNEs in different countries and their global enterprises; are we also able to split these GENs in their national parts all over Europe, the TENs, and to gather consistent data on the TENs.The consistency is, of course very difficult to assess, and it is the topic of an-other ESSnet, whose results will be later incorporated in this report.Here, the scope is restricted to the domains of highest priorities, as detected and proposed in the “feasibility study”.



document.doc

7C The main statistical stakeholders24 of “profiling”

7.11 The prioritized list for the survey statisticians is quite simple: - the BR as a whole (including the EGR), as it includes an “enterprise register”- the EGR, as profiling is directly in charge of the “enterprise” register for the

“large and complex MNEs”, task including the checking of MNEs clusters, operating and economic structure, and the determination of their principal activities,

- OFATS (Outward Foreign Affiliate Statistics):o as they are in accordance with their own regulation providing information of affiliate “enterprises” outside Europe of European-centred MNEs; even if we know perfectly that the definitions might be more similar than identical, it would be paradoxical that the FATS statistics remain independent from the BR and EGR; o as they are the best candidates to provide basic significant information on the structure of the medium (and small?) Enterprise Groups for profiling pur-poses.

- SBS the Structural Business Statistics are, according to their own regulation, mainly based on “enterprises”; they constitute the domain which will get im-mediate improvement from the profiling, for use in o its own production of significant data and also as o a predominant provider of information for the National Accounts.

- Other domains are also interested in the profiling results, such as FDI (For-eign Direct Investment) and Balance of Payment statistics; their use is treated indirectly through their relations with the EGR system.

Even if one do not agree with the previous priorities and rank them differently, the results on core variables would not change much. If one thinks of omitted priorities, then the results could change more.

7.12 The place of National Accounts (NA) is predominant, but nevertheless dif-ficult to assess.- In reality, all links between NA and statistical units (SU) are indirect: o NA offer statistics of a 3rd type (the 1st is statistical, the 2nd is use of AD-

MIN data, the 3rd is a kind of synthetic statistics). o We propose SU for the 1st and 2nd type.

- In the world of concepts, the NA need a set of consistent concepts to make it possible to build their synthetic evaluations; of course, they tend to define themselves this set to avoid their coordination burden to be too important.

- Also in profiling large and complex MNEs the greatest consideration should be given to the manuals of NA; but weproposals for changes should be al-

24 Stakeholders can be defined in terms of Institutions or in terms of statistics. It is choosen for statistics here. National statistical offices (NSIs) and international statistical bodies like Eurostat, UNECE are of course stakeholders too.



document.doc

lowed if and when the present (use of the) NA concepts leads to significant inconsistencies.

- The relationship between National Accounts and Statistical Units, in particu-lar the Enterprise unit, should be assessed more clearly.

7.D Variables linked to each stakeholder:

7.13 The variables for the most important statistical use are dealt with under this heading, but not in an exhaustive way. Further elaboration should be done else-where and is beyond the scope here. 7.14 BR as an “enterprise” register, also including the Enterprise Group: - Usual identification characteristics, including a name (the enterprise (group)

being a pure statistical entity has no natural name; of course the MNE and the NSI can agree on the name generally used for the SU (established and/or agreed Enterprise and its truncated parts) on any other convenient naming)continuity criteria (see BR manual)

- relations with other SUs - the link and the composition of (legal) units to be made visible between the

global enterprise and its units on “lower” levels- including a dual notion of Enterprises (GENs & TENs)

7.15 EGR according to the role it gives to profiling: - the same variables apply on the three SUs (EG, GEN and TEN), previously

listed for BR, as it provides the “enterprise” register for large MNEs; - for checking MNEs cluster, structure & principal activities: o Delineation of the EG as combination of legal units under common con-

trolo Enterprise and its links with Legal Units (if possible as these links make it

possible to monitor the change of usual statistics and surveys based on LeU into the new system based on “enterprises” for those countries which used the legal unit as the basis for the SBS )

o information on size of the SU (to be precised: turnover, employment)- control variables necessary for IT.

7.16 FATS - the closeness between O-Fats and profiling is stressed by MNEs during the

contacts with them; this implies that => o OFATS could be replaced by quantified Profiles, in some kind of statistical

collection strategies o OFATS could provide the structure of profiles for semi-manual light pro-

filing procedures (if extended as well to EU and domestic parts as to for-eign parts of the same group) (so that OFATS would become a result of profiling for the largest MNEs and a source for EGR update for medium size MNEs)



document.doc

- the list of OFATS permanent variables will be soon completed, even if re-mains under exam in the appropriate committee of Eurostat.

7.17 SBS with 2 strategies: - profiling as a provider of a list of SUs for surveying, for using data of ad-

ministrative files (ADMIN if possible), or matching both kinds of data; with what kind of quality control?

- profiling in charge of mixed role (register and part of collection): with what kind of cost control?

- need to check the SBS list (the “regular” and the “specific” by industry: will be done soon)

Among the other stakeholders, the case of National Accounts needs a complete study, which is made in the paragraph 7.F. Thinking about consistency between truncated and global level, the attention should be paid to the global perspective!

7.E The position and needs of the MNE

7.18 The relations with the MNE, in the delineation of enterprises it comprises, have to be direct and confident.Nearly every time the organisation is complex, the MNE representatives ask about what kind of variables we would like to gather on the “enterprise”: their answer on the possibilities to do so might change, according to the type of vari-ables we need. So, it is to find a common European answer to this question, because this would lead to the least burden at MNEs from common European needs.

7.19 What could be the approach to this topic? - If we suppose that in the present situation, many countries rely on “tax office”

data, related to some kind of accounting data or system; - if the providing of (or enabling the SBS to provide ...) basic information to

build the National Accounts (which are balanced in the accounting way) is sup-posed to be the first priority;

- if we suppose that the SBS statistics (on turn-over, employment, etc) should rely to the same SUs as the SUs surveyed for NA purposes,

- priority is given to “accounting type” core variables and rely them first with the SBS core variables.

7.20 The Variables for reconciliation with SBS statistics, NA use and their links with accounting standards presented here are limited to the componentes of the Value Added They are elaborated more completely and in more details in the ANNEX III.

The needs of NA are usually presented along of 3 different lines: - the components of value added as a source for sectorial accounts (as well by

Institutional Sectors as by industry);



document.doc

- the basic data for evaluation of fixed capital (elements coming from the bal-ance-sheets or the so-called “financial statement”)

- bases for wealth estimates (coming from the rest of the balance-sheets); these data could also be useful for checking the fixed capital elements as part of the balanced items of a global account.

Here only the first line is developped.

7.21 The IFRS accounting standard (International Financial Reporting Standard) provides a template of Profit & Losses (P&L) accounts, also for operational seg-ments of a MNE. This is a tentative version of a very simplified data set that try to include enough variables to produce the main aggregates of sectorial accounts for NA flows; it includes nearly all variables that appear in SBS requirements (missing at this moment change in merchandise inventories) plus as few vari-ables as necessary to have complete accounts. These accounts and the data of these seem very difficult to obtain when the En-terprise statistical unit has insufficient autonomy regarding its production pro-cesses from a financial (production process related) point of view, but, in this case, the existence of a specific “enterprise” including holdings, managing activ-ities (and other residual ( mainly auxiliary)activities of the MNE) is very prob-able; this residual “enterprise” would also include most of the “production’ ori-ented financial-type variables.

7.22 Very important: it works from a by-activity version of the P&L ( in which purchases, wages and salaries, depreciation of fixed capital, etc are displayed explicitly) and not from the frequently displayed version, which is by- function (in which the main variables are “cost of sales, distribution costs, administrative costs, etc”). So that the existence (and MNE the agreement to display) of this by-nature version of the accounts is crucial.The beneath table with variables reflect the Enterprise and Truncated Enterprise statistical units, resulting in the operational (truncated) results (remark: in some cases the information of this table is less detailed that requested by the present SBS questionnaire e.g. on splitting between sales of own production, sales of merchandise etc; the same for inventories):



document.doc

P & L by nature (not by function)

(consolidated for the GEN or the TE)

Content of the boxes

1st evaluation of VA( by the production

side )=

“Production” = Proxy of the “Revenue” variable?

+ pure T-O

+ other activity products

+- produced inventories

T-O (sales to 3rd parties) net of granted dis-counts & rebates (IAS18) if possible split between sales of merchandise and other sales. Newly produced inventories and fixed as-sets (incl grants on fixed assets)

+- Other products and expenses Patent, franchise & brand royalties etc; more useful if split between products and charges.

- Used purchases

- Other external expenses

Purchases of goods (raw material, equip-ments and parts etc), services and mer-chandises, net of in inventories of re-ceived discounts and rebates +ancillary costs on purchases (CIF etc)

2nd evaluation of VA( by the components) =

+ Employees wages and salaries Including bonuses and share of results+ Taxes on products and indirect taxes

+Depreciations Physical and value change = Current operating result

(optional)Close to net operating result of the NA

+- Other current products and charge Other current non recurring products & charges (e.g. sellings of assets, depreciations

of goodwill)Operational result

What is very important here, is that the translation process has begun: the variables in the center column give proxies to the wished NA-required variables, the left column giving the clue for the calculation of the VA as well by the production side (something close to production less intermediate consumption) as by its components (wages and salaries including social contributions paid on salaries plus “indirect taxes” plus current operating surplus)

8 Classification of changes: which, why and how? Continuity rules.

To be drafted in accordance with EGR methodology and in accordance with the BR manual 2010.

9 Analysis of usefulness of units of the new model for SBS, STS, FATS, Prod-com and NA

First proposal to be drafted in 2012



document.doc

10 The concept of autonomy related to main information and variables: mean-ingfulness and availability of necessary information/data. How relates global autonomy to geographical autonomy at MNEs? What are the con-sequences?

First proposal to be drafted in 2012

II Consistency and breaks

Aim: To enlarge this assessment of the present theoretical or used statistical units (e.g. present enterprise, KAU and eventually LKAU) through a contribu-tion to the ESSnet on Consistency, in accordance with the domains it will choose (SBS, STS etc)

11 Relationships with used statistical units (Ent/KAU/LKAU) in present stat-istics (SBS, STS, FATS, Prodcom, NA) and related to Consistency ESSnet

In agreement with the ESSnet Consistency colleagues the treatment of the item is postponed to 2012

12 Analysis of potential breaks in time series in changing the used units, as-pects of continuity and consistency; implications of changing

12.1 Profiling activity, national or multinational, meets the need to treat the com-plex situations in which the statistical unit “enterprise” is built as a combination of several (more than one, numerous) legal units. The result is the creation of a unit closer to the economic/organizational world rather than to the administrative one. For MNEs acting in a global perspective with hundreds (or thousands) of Legal Units (LeUs) often created for fiscal/administrative reasons, profiling is the proper tool to make a better description of the activities carried out.

12.2 The top-down approach followed for the MNEs profiling activity intro-duced a double view of the statistical unit “enterprise”: the “global enterprise” (GEN) derived from the operational segments disclosed by the group itself and the “truncated (national) enterprises” (TEN) which are the truncated national parts of a global enterprise. Operationally, the first step is the delineation of the GEN and the second is the identification of the TENs as the national parts of the



document.doc

GENs both options recommended in the Feasibility Study (Options D and E WPA – SGA 2010)25.

12.3 Following the global profiling approach adopted in the ESSnet it is neces-sary to take into account three elements that can produce substantial effects: 1) the global view; 2) the possibility to split legal units into several enterprises; 3) the accounting data consolidation to prevent double-counting.

- The first effect is due to the fact that starting from the group perspective, even if it is a combination of single legal units, the global view is conceptually irre-spective of national boundaries. - Furthermore the single legal units may contribute to more than one enterprise (second effect: one to many relationships). In that case the links are not referred to financial or controls ties but they are only related to the actual economic structure of the enterprises within the group. - The application of a top-down method leads to a data consolidation of the in-tra-group flows (or better intra-GEN flows). If IFRS standards are followed they represent a very helpful standard. Considering the business operating divisions introduced in IFRS8 principle26 the perspective changes in favour of the group view in comparison with the legal or administrative criteria, until now mainly adopted for the enterprise delineation. Anyway profiling activity is independent of the application of accounting standards (IFRS, GAAP) and it could be pos-sible to obtain the same results also with the “traditional” financial statements or adopting another (e.g. bottom-up) approach in close cooperation with the MNE.

12.4 The application of the global profiling produces results differing from those obtained with the classic and frequent present situation in which each legal unit is entirely comprised in a single enterprise. Such approach determines con-sequences in terms of the treatment of legal units (especially for units perform-ing “ancillary” activities): - the delineation of statistical units (impact on the structure of the BR); - stratification characteristics (for example NACE codes, size class); - values of variables (for example, number of persons employed, turnover).

12.5 Profilers from UCI Countries produce the first “global” view. Considering the situation in which the financial statements are useful, the NSI responsible 25 Option D - Profile at the global group level and derive and classify GENs at the global, and TENs at the national, level centrally by the NSI of the UCI. Pass the TENs and the list of legal units for each country to the relevant NSI for use as supplied in the national business register and for data collection.Option E - Create GENs and collect data centrally and apportion to the national level, where a GE includes activity for more than one country. There would be no link to the legal units other than at the global group cluster level.26 The core of the principle is “disclosing information to enable users of the financial statements to evaluate the nature and financial effects of the business activities in which the groups are engaged and the economic environments in which they operates”



document.doc

collects the data for the Global Enterprises (GENs) starting from the consolid-ated balance sheets (that means eliminating the intra-group flows) not consider-ing the national boundaries. The result is that the pure sum of the national results (for example for the EU Countries) for some variables can never be equal to the values obtained considering the GENs. So if the minimum requirement for GEN was a principal activity code, an estimate of the employment and the turnover, the differences with the aggregation of the National data would be the following: “global” NACE code could be different from the national (change of view);

there could be two extreme situations: a group operating in the same way in all Countries (=> global=local NACE code for all the Countries and for thee GEN); a group with the principal activity only in one Country (global=local only for one Country27). In practice, MNEs are within the extremes and it is advised to adopt a dual-coding system especially necessary in the application of Option D to permit calculation of European aggregates (see paragraph 6.1). But, apportioning data for all the Countries in the option E should also be supported by a careful evaluation of the MNEs organisations.

the employment values are the same if we consider the entire group: sum-ming national data or GENs values must produce the same result. But if we consider the variable “employment for each activity” the results could be dif-ferent because GENs are created in a different manner.

“global” turnover is necessarily different from the pure sum of national data both for the entire group and the enterprises: in contrast to what happens for the employment the accounting consolidation produces lower values than the summary; considering the single GEN the effect could be different because the reduction due to the consolidation could be balanced from the legal units’ aggregation.

12.6 The second task of the profiling team of the UCI Country is deriving the TENs both for Option D and E. Ideally the TENs would be the unit for statistical surveys but they need to be able to provide the full set of accounts. This is pos-sible if the TENs comprise whole Legal Units that have the ability to provide ac-counting data for statistical purposes28.

12.7. Going down one step more towards the national level, the effects are both in the UCI Country of the MNE and in the other Countries. In the majority of the cases (when TEN = sum of LeUs) the profiling activity brings to the aggregation of different legal units. Only in some cases, also a split of single legal units will be needed. Taking into account the national dimension, the values of the economic variables are produced after consolidating the intra-group flows within each different

27 in very extreme, but exceptional cases the GEN NACE code could be different of the NACE codes of all the TENs it comprises28 Until now, with the exception of a few “European companies”, LeU are nationally defined



document.doc

Country. The potential breaks in changing the used units should then be the fol-lowing: a reduction of the total number of “enterprises” in the BR. changes in the structure of the BR in terms of size classes (increase of the

share of large enterprises) a simplification in terms of NACE codes scattering: if the statistical units are

considered as complex units performing a principal and some secondary activities: 1) the TENs are generally29 larger than the “traditional” units (in relation to their principal NACE code); 2) “support” units or ancillary units, often of small size, disappear.

a break for the turnover time-series due to two main effects: 1) an increase due to the aggregation of legal units previously coded in a dif-ferent way (or aggregated in a different way from national profiling activity); 2) a reduction due to the intra-group flows detection.

For turnover, the final result is globally lower than the turnover coming from bottom up procedures; but it may be higher or lower on each NACE code de-pending on the impact of detailed intra-group flows.

12.8 For the other Countries global profiling needs also a preliminary evaluation of the TENs. Broadly, the steps for non-UCI partners after receiving their own truncated parts are: evaluating the delineation structure; could be the TENs the unit for statistical

surveys? Is it possible to collect data in that way? comparing national and global NACE code; is a dual-coding system neces-

sary (or sufficient) a to guarantee coherence at the national level? comparing the structure and the values (12.7 para): different number of enter-

prises (and LeUs); NACE changes; different values.

12.9 It deserves a speech in part the evaluation of the value added (VA) as result of the profiling activity. In the 6th chapter of this report (The core variables and the roles in the BR/EGR context) and in the more comprehensive annex III the core variables implied and their roles in the profiling activities are described. Data consolidation (global or national) means deleting the intra-group flows for some variables and summing for others. In theory VA does not change. In fact, if we consider the simple case in which only two LeUs constitute an enterprise the situation may be described as follows:

LeU A (NACE Section C, manufacturing) sells all its output to LeU B (NACE division 46, wholesale);

Purchases of goods (costs) for A:100;

Production of A= Purchases of goods (costs) of B: 120;29 Generally for the TEN’s employment the final result is an increase but the effect is uncertain when we compare the previous situation (LeU=Ent or national profiling with LeUs’ combina-tion) with the global profiling results in which the LeUs are split into more enterprises.



document.doc

Production of B: 160;

Evaluation of VA by the production side (simple hypothesis Production-Costs) for the aggregation of A and B: (120-100)+(160-120)=20+40=60.

The GEN obtained from the global profiling is classified in section C and delet-ing the intra-group flows VA=160-100=60: the value added does not change. But which is the situation in terms of NACE codes? In the “traditional” view 20 is for Section C and 40 for Division 46; after global profiling 60 is for Section C. So if the total (for a Country) value added is built for aggregation weighting in a different way the values in terms of NACE codes, global profiling also determ-ines a change for the variable “value added”.

III Consolidation aspects

Aim: The consistency of information collected at the different levels of the groups should be considered and the potential sources of inconsistencies between the data produced at the different levels should be explored (this case includes a possible insight into External Trade in goods and services).

13 Description of basic organisational principles of large MNEs and the ap-pearances of these

13.1 The questions concerning basic organisational principles of large and com-plex MNEs refer more to the economic or the management science than to the statistical studies. But, as previously stressed, our aim is to describe the “real economic” world, and not exclusively the legal organisation. Knowledge on or-ganisational principles and the reasons for creation of a legal structure within a group give first holds for the statistical approaches.

In this context, we could easily conclude that, according to the large variability of managerial structure, each MNE case is specific.

So, prior to the pure statistical work, let us go for a very short while into an in-sight of the “reasons” for an EG to become “large and complex”. There are many books on organisational issues of companies, also in relationship with ac-counting, but for the purposes of profiling it would be interesting to have means and criterions for groupings according different motivations.

13.2 The enterprise group can be motivated by the following looked for effects: 1 industrial leveraging effect of scaleThese effects tend to promote a world-type organisation, in which each country plays a role very specific according to its economic advantages; the management is usually either centred on “business lines”, the production of each country be-ing integrated in a global process or of the “functional type”, with stress put on



document.doc

the “markets” and on the supply chain, which leads to great difficulties to de-scribe the production and to apportion it country by country. They can operate: - on procurement and/or logistics - on manufacturing - on back-office and ancillary services - on use of R&D results (patents, process, models)2 « brand » effect of scale These effects lead to a “distribution” model, reproducible in each country, not compulsorily related to a unique or united production model; it can be also used in a franchise type of development, in which the “autonomy of the franchisees can be interrogated.They can operate: - on advertising (world brands) - on relations with other MNEs (global contracts) - on Human Resources (attraction effect)3 financial effect of scale Their priority is not the costs or the efficiency of the MNE (industrial) or its dis-tribution (“brand”) but the financing of its activity.They can operate: - lending and borrowing (leveraging?) - using fiscal benefits and/or (in combination with) relocation.

13.3 Reversely, the enterprise group can meet conditions that lead to a “regional and/or local type of organisation”:1 the nature of goods and services traded - small unitary price of goods (forbidding long distance transport and leading

to local production; but the production processes can be world-wide) - maintenance and repairs (similar to previous point) - services to persons (as well personal as relative to persons) (having also to

be produced locally and commonly also related to local habits or culture for-bidding integration of process management)

2 necessary local distribution (can supplement the alinea 1, or be specific) - of perishables (as varied as fruits, cement …) - with advice or fitting (clothes) - etc3 specificity or constraints of (local) regulations - regulated activities (e.g. fix or mobile phone, broadcasting etc) - services of most professionals - etc

13.4 The previous effects can combine; after combination, they produce typical appearances of MNEs that we will try and describe more precisely through our testing experiences. The result of testing will be included in the “manual” of pro-filing as guidelines to profile in concrete situations.



document.doc

At this stage, we can list examples of what we think to be in force in some in-dustries: 1 – pure industrial effect:Motorcar construction industry / High value industrial edibles (chocolate, sol-uble coffee, etc) / Iron and steel producers2 – industrial effect combined with nature of products:High unit value construction materials producers (even in this case they remain “heavy” products / Escalator constructors (to be locally installed according to local regulations)3– industrial leveraging effect combined with « local » regulations (networks, defense) Defence industry / Pharmaceutical distribution / Fix and mobile phone (seem to be all organised on regional basis even if all the techniques are global) 4 – Pure financial:This effect results in conglomerates (highly diversified MNEs), that have no descripted cut-offs with investment funds. Tests will try to reveal significant limits (if no combined system of reporting, no synergies between the different activities, no common general staff, no share of treasury => the different seg-ments will be considered as independent “EGs”.

14 How relates consolidated information at the (global) group-level to re-quired geographical statistics? How to decompose and how to deal with in-ternal flows of goods and services within global groups, including the aspect of transportation?

The treatment of the item is postponed to 2012 to better evaluate the possibility for the ESSnet members to collect from the MNEs and manage intra-group flows

IV Modes in profiling

Aim: To elaborate further the different approaches to profiling (mainly through description of the use of top-down and/or of bottom-up procedures)

15 The standard and basic steps in profiling large and complex MNEs15.1 Methodologically the basic steps in profiling large and complex enterprise groups (MNEs) are:- Selection of the MNEs to be profiled- Defining the UCI, that determines the responsibility for the profiling of the

MNE- Desk preparation:

- collection of information on the legal structure, using EGR data



document.doc

- determination of the countries in which the MNE operates, and so determ-ination of the NSIs involved- collection of information on the operational structure of the MNE, using the annual reports and information at MNE websites- collection of the MNE information that is already available at the NSIs (e.g. from surveys)- analysis of the relationships between the legal structure and the operational structure. If these are not clear, priority should be given to the operational structure of the MNE

- Analysis of the role and place of the MNE in statistical outputs (uses)- Informing the MNE management at the highest level as possible by the NSI

at highest level as possible on the plan for profiling the MNE and explaining the goals and procedures of it.

- Drafting a first proposal for the profile in terms of statistical units (global Enterprises)

- Informing for comments the NSIs of EU-countries in which the MNE oper-ates on the draft profile

- Discussing the proposed profile with the MNE- proposal is in accordance with reality - autonomy of the proposed units regarding production processes (allocating current resources)- the availability of the core variable data items (see chapter 4)- possibilities of the MNE to split global enterprise data according the trun-cated enterprises, what means according the countries in which the global enterprises are operating - discussing the reporting and observation structure- discussing the way of maintaining the profile- discussing the management of relationship between the MNE and the NSI

- Discussion and final agreement with the relevant NSIs as to the structure of the global and truncated enterprises for profiling

- Formal agreement between NSI and MNE on the statistical units’ structure, the reporting structure and the structure for observation. This agreement should be formalised and approved by a contract or a formal letter, to be drafted and sent by the NSI

- Information on the profile and contract to the users in the NSIs (all NSIs of countries in which the MNE operates)

- Evaluation and regularly check of the profile, both with the MNE and the users

- Update of the profile if necessary. Update also includes confirmation of ex-isting profile.

The presented steps are according the Deming quality circle: plan, do, check, and act.

Procedures and guidelines for this will be prepared in the work package C.



document.doc

16 Description of the general principle approach

16.1 The basic principle should be that used statistical units represent the real statistical economic reality. The units (global) EG and (global) ENT (GEN) must be considered units that autonomously generate economic processes or are in-volved in these processes. This means that the units are not represented by pure legal, fiscal or administrative constructions.

16.2 As to small enterprises, the operational structure and the group structure will not differ very much. Enterprises can, as in many other cases, be defined in terms of (combination of) legal units within the Enterprise group. But enterprise groups and enterprises mostly will be equal, which means that all legal units constituting the group will be combined in only one enterprise.

16.3 The operational structure of domestic large and complex enterprise groups (all resident enterprise groups) will in most cases differ from the legal structure. Then it mostly will not be possible to have a good delineation of enterprises along the lines of the legal units. A tailor-made analysis using the requested cri-terions must result in the enterprises to be used in statistics.

16.4 At multinational groups the international dimensions must be taken into consideration additionally. The only level of a real economic analysis is the global level. From the results of the global analysis the national truncated parts of the globally established enterprises are derived as truncated enterprises.

16.5 There is a huge gap between the small units and the large and complex ones. In between a big amount of medium sized groups exists, of which a num-ber can be complex too. How to deal with these “in- between” units needs fur-ther attention. The solution for these must be found in a combination or com-promise of the approaches used for the smallest and the largest units.

16.6 Different intensities of profiling will be described in the chapter 18.

17 The difference between top-down and bottom up profiling and how to com-bine these modes?

17.1 In top down profiling firstly the full group as a set of all legal units under common control is to be established. After this the group must be analysed into its group operating divisions (GODs) and from there in the global enterprises for statistical uses. Having done this, the profiler may in some cases be able to link the legal units to the enterprises. In other cases, the units must be identified through discussion with the business. Enterprises are to be established according the main criterion of autonomy regarding production processes. Market-orienta-tion and data availability are components in the autonomy check.



document.doc

Diagram representing top-down profiling (see report WPA 2010).

An advantage of this way of profiling may be that daily or frequent changes in the legal organisation do not immediately affect the structure of the statistical units, although these changes could indicate the need for an update or a partial update of the profile. Another advantage is that profiling in this way leads to an exhaustive profile that means that all legal units are included in the profile, are incorporated in one or more enterprises or are separate individual quasi-enterprises (including SPEs and inactive or dormant legal units).

17.2 For very large and complex groups, it would be very exceptional if the profile analysis would result in an enterprise structure as presented in the diagram related to the paragraph 21.75 of the BR recommendations manual, see below. Here the underlying principles are: “ Once the enterprise group is reconstructed from the legal units, the enterprises must be delineated within the set of legal units that are the basis of the enterprise group: The enterprise consists either of a single legal unit (‘simple enterprise’) or of a combination of two or more legal units (‘complex enterprise’). It follows that an enterprise group must consist only of complete enterprises, not parts of them.” (21.75 quoted)


GlobalGroup

GlobalGroup

TEN1 TEN2 TEN3 TEN4

GOD1 GOD2

GEN1 GEN3GEN2 Statistical environ-ment at global (GEN) and na-tional (TEN) levels

Legal or opera-tional(non statistical) environment


document.doc

Diagram: Legal units as a common basis of the enterprise group and its constituent enterprises

17.3 In the bottom-up approach, a legal unit is identified together with its relationship with its immediate parent, which is the direct controlling legal unit. This creates tree structures as the basis for creating enterprises. It deals effectively with time-lags and other deficiencies in administrative data. While it results in an accurate reflection of the structures and activity of a group, it is a costly way to deal with groups with hundreds of affiliates. The legal units’ diagram, including the resulting enterprises structure, could be the result of a top down analysis or the result of a bottom up analysis. The bottom up procedure starts with the individual legal units and it is checked each time if the legal units fulfil the requirements to be an enterprise. If not, it is checked if it fulfils if it is combined with its direct controlling legal unit or with the units it is direct controlling and so on.

17.4 From the above presented descriptions and diagrams it may become clear that the use of a mixture of the approaches can be very useful. In practice we will have an iterative process in which the requirements are checked for the found possibilities of combinations. If no combinations are possible at all we can not longer rely on the legal units structure and must use the approach as presented in the above top down profiling diagram. The mixed approach mostly


Enterprise groupEnterprise groupEnterprise 1

Enterprise 4

Enterprise 2

Enterprise 3

Enterprise 5

Head of groupHead of groupAA

Legal unitLegal unitEE

Legal unitLegal unitDD

Legal unitLegal unitCC

Legal unitLegal unitBB

Legal unitLegal unitFF

Legal unitLegal unitGG

Legal unitLegal unitII

Legal unitLegal unitHH

Legal unitLegal unitJJ

Legal unitLegal unitKK


document.doc

will apply for the medium sized groups (e.g. more than … persons employed) and for the very large units that are not selected for global profiling.

18 The different intensities of profiling (Intensive, light, manual, automatic)18.1 From the definition of profiling all kind of organisational units can be sub-ject of it for the establishment of the enterprise structure, except for single units carrying out economic activities in their own and for their own account. It is an exception if a large economic activity would be carried out in a single legal unit not belonging to a group. It still occurs in governmental institutions, but these are out of scope here.

18.2 Available means and resources should optimally be assigned. Most atten-tion is needed where most effective results can be reached in terms of costs and quality. Response burden is included in these criterions.

18.3 Automatic procedures for the constitution of enterprises based on legal units are suitable and sufficient for small units. It could be agreed to automatic-ally combine all legal units under common control if the size of the combination does not exceed a certain level.

18.4 Automatic generated proposals for the constitution of the statistical units structure for groups exceeding the criterions mentioned before, must maybe manually checked according quality signals from the automatic procedures. Here we come to the procedure of passive profiling. Manual action is done if the auto-matic business registers procedures generate signals, or if there is a need to do so based on feed back information of users of the register. This kind of manual in-teraction can be seen as a kind of very light profiling.

18.5 Intensive profiling is to be done for the very large and complex groups which have very high impact on the statistical outcomes. So the top xxxx enter-prise groups should be statistically structured as actual as possible. It is recom-mendable to assign a profiler (sometimes called an account manager as in private business or a coordinator) to these groups. This manager, well known with the group, can judge the statistical impact of changes within the group. As compared with usual domestic profiling activities of the (truncated) groups the methodology and procedures to start global profiling will have consequences for the selection of the groups to be profiled according the approach of intensive profiling.

18.6 There are a large number of groups between the groups selected for intens-ive profiling and units which are constituted according automatic procedures. Depending on available means and resources and statistical input (such as en-larged OFATS surveys in some countries) the number of units under active pro-filing can be defined. Active profiling can additionally be done for quality checking for a defined group of units, although these units do not belong to the population for the intensive profiling, or signal based.



document.doc

19 The selection of groups to be distinguished (criteria and numbers), see also WP C

First proposal to be drafted in relationship with the WPC in 2012

20 Priorities and follow up strategies: which group in which intensity with which frequency.

First proposal to be drafted in relationship with the WPC in 2012

21 Relationship with updating strategy of the EGR

First proposal to be drafted in relationship with the WPC and EGR which, if pos-sible, identifies links between legal units and enterprises for identification of the enterprises. Will be done in 2012/2013

V EU-parts of non-EU MNEsAim: To define profiling for the European part of non European groups: desk profiling of a few groups could be undertaken to define potential different mod-els (of profiling as well as of statistical units etc) and to propose a way of treat-ing them.

22 Description of the problem

Profile the EU-parts of a non-EU MNE necessitates to know, guess or estimate:

- the GGH of the MNE (we will often precisely know it from the commercial providers, with more difficulties when individuals are concerned)

- its UCI (we will most of the time know it easily for the MNEs whose UCI is located in the most developed countries; it can be more difficult to locate it with certitude when either the owner(s), the capital or the management comes from developing countries; not because it is theoretically impossible but be-cause the present commercial data providers, centred on developed countries, have less information of satisfactory quality. We probably will have more problems when the scheme changes rapidly than in more traditional struc-tures. A question to treat is the eventual link between the rate of growth of an economy and the rapidity of change of its productive structures: if this link does exist then we will have more problems with MNEs whose UCI is in fast developing countries.



document.doc

- the hierarchical scheme of the overall control (because of lack of information, the case is more difficult when developing countries are involved; but it is frequently missing whatever the country of origin is. What we know quite well is that there is a control, direct or indirect, without being able to draw the tree).

As a consequence it is difficult to be sure of the global size of the MNE, the EU being often insufficient to be taken as a significant characteristic, as well on turn-over as on employment.

We probably, in numerous cases, will not be able to find correspondents in charge of the MNE in Europe; correspondents that we usually look for in order to discuss the profiling proposals.

It is also difficult to know the operational segments that do exist in the MNE and to be sure of their impact on the “truncated enterprises” - truncated meaning in this case as well restricted to EU as restricted to an individual MS.

What might often happen is the existence of apparently independent sets of affil-iates of the group, each set operating in a bundle of MS, with great difficulties to build relationships between what is on in the sets of affiliates and the bundle of countries.

Nevertheless, only testing can show us if our fright is justified. The 2 first ex-amples (one in France with a very large global and diversified non-EU group, one in UK with a non-EU group specialised in all electronic and computerised apparels) do not justify the difficulties described previously, but this “small” testing is not yet able to permit significant conclusions.

23 The impact of partly profiling on the statistical units model

What is our real aim when profiling a non-EU MNE? Our answers can be of two types:- we would like to do exactly the same process as for EU-MNEs:o determination of the exhaustive global cluster of affiliates and/or con-

trolled legal units;o delineation of the global enterprises through the analysis of the operating

segments (but with difficulties to meet global MNE officers if required), then the delineation of the TENs;

o checking the ability of the TENs to provide the required statistical data.- we would want to profile only the group defined by the EU affiliates and their

daughters, grand daughters etc (exactly as if the upside part of the MNE was limited to one unique GGH with direct links to the EU parts, rubbing the in-termediate structures). A possible problem can be the delineation of the right group cluster (perimeter) because of the possibility that control relationships only can be detected outside the EU.



document.doc

The advantages and disadvantages of the 2 solutions have been slightly ex-plored:- we are not sure to be able to get the upside information from non EU MNEs;

on the contrary all the downside data are legally available in Europe; - if we are sure that we want to describe Nokia as belonging globally to the

manufacturing industry, we see no real advantage to describe Samsung in the same way (for the EU, Samsung is a pure distributive group)

- we feel comfortable when the group has created a European holding structure; we found it uncertain when each affiliate report to non EU owners. In the last case it could be impossible to do more than a legal unit investigation or to survey a large bunch of very small subgroups, having no direct relations with each other even if all related to the same but paractically unknown global en-terprises.

We have nevertheless made a provisionnal choice and decided to test the profil-ing restricted to the EU and downstream.

VI SPEs and other special cases

Aim: To describe the implications of changing from current enterprise to GEN/TEN, in-cluding the treatment of some specific cases among: financial vehicles, holdings, auxili-ary activities and units, R&D, SPEs.

Introduction

VI.1 The term Special Purpose Entity, SPE, is used for legal units that are cre-ated to fulfil narrow, specific or temporary objectives and/or to gain fiscal ad-vantages30. Other special business structures frequently encountered in MNEs such as legal units consisting purely of centralized R&D, headquarter or auxili-ary activities are labelled as other special cases.

VI.2 SPEs and other special cases have an increasing importance in the global organisation of the MNEs and in their extension to new geographic areas.

VI.3 SPEs and other special cases affect business statistics and national account statistics through altering structures of enterprise population, GDP in relation to GNI and international trade in services. For example, SPEs often locate in coun-tries offering tax advantages, and one of their main purposes is to generate cross transactions between subsidiaries located in different countries. Therefore, SPEs directly affect FDI creating artificial cross-border flows.

VI.4 The recent international statistical regulations and manuals recognize the impact of SPEs and other special cases. However, there are few detailed guidelines on how to handle these entities in various statistics, and countries are 30 The guide “Impact of Globalisation on National Accounts”, p. 39 Item 4.5.



document.doc

in many cases left to choose their own treatment. National experience, task forces and working groups show that existing definitions and recommendations for treatment are specific to the very statistical domain in question – reflecting its objectives and characteristics. Moreover, decision trees and other solutions are typically described for legal units within national boundaries.

VI.5 MNE profiling may take a crucial role in assuring coherent statistical treat-ment of SPEs and other special cases. As profiling will be realised in close co-operation with the EGR and as SPEs are MNE-specific phenomena, the NSI-profiler of the MNE may provide the EGR with updated information on SPEs and other special cases. The EGR could then flag SPEs and other special cases accordingly, and the information would be centrally available for production of various statistics. As a consequence, individual statistics need no longer to identify SPEs and other special cases but may take their status as given, and ap-ply the appropriate statistics-specific treatment of these entities.

VI.6 In this Section VI Chapter 24 summarizes the identification criteria of SPEs as described in recent international regulations and manuals. Chapter 25 pro-poses treatment of SPEs in MNE profiling. Chapter 26 summarizes ancillary activities and R&D activity as described in recent international regulations and manuals. Chapter 27 proposes treatment of R&D and auxiliary activities in MNE profiling. Chapter 28 introduces the issue of branches and “VAT positions”.

24 SNA principles and requirements of Special Purpose Entities (SPEs)24.1 MNEs often have several legal units that are insignificant from economic31

point of view. In different statistical manuals, these types of units are referred as ancillary or auxiliary units or artificial subsidiaries. Moreover, terms SPE and SPV32 are used for certain types of units. To a large extent, the complexity of a MNE arises from the amount or variety of these special types of units.

24.2 SPEs Concerning the SPEs, the new international manuals (the 2008 SNA, BD4 and BPM6) include general criteria which may help statisticians to identify them. In addition, many different task forces/working groups in Eurostat have attempted to facilitate the identification of SPEs.

24.2.1 SPEs according to the SNA 2008The SNA 2008 Chapter 4 Institutional units and sectors defines institutional units and their use. The chapter deals with different types of corporations includ-ing not only legally constituted corporations but also cooperatives, limited liabil-ity partnerships, national resident units and quasi-corporations. Moreover, spe-

31 more precisely, the point of view of production and distribution of goods and non-financial services32 special purpose vehicle



document.doc

cial focus is given to special cases such as groups of corporations, head offices and holding companies and special purpose entities.33

The main principle in the SNA 2008 is that each individual corporation should be treated as a separate institutional unit, whether or not it forms part of a group. The SNA 2008 mentions some special cases (mainly related to large groups of corporations) and includes some paragraphs on SPEs:

“Such units often have no employees and no non-financial assets. They may have little physical presence beyond a “brass plate” confirming their place of registration. They are always related to another corporation, often as a subsidi-ary, and SPEs in particular are often resident in a territory other than the territ-ory of residence of the related corporations. In the absence of any physical di-mension to an enterprise, its residence is determined according to the economic territory under whose laws the enterprise is incorporated or registered. For more detail on problematical cases see BPM6.

Entities of this type are commonly managed by employees of another corpora-tion which may or may not be a related one. The unit pays fees for services rendered to it and in turn charges its parent or other related corporation a fee to cover these costs. This is the only production the unit is involved in though it will often incur liabilities on behalf of its owner and will usually receive investment income and holding gains on the assets it holds.”34

According to the SNA 2008 only following categories of SPEs need to be treated different from other resident entities:

– Captive financial institutions

– Artificial subsidiaries of corporations

– Special purpose units of general government

An example of a captive financial institution is a holding company that simply owns the assets of subsidiaries.

Artificial subsidiaries of corporations are created to provide services to the par-ent corporation or other corporations in the same group. For example, the parent may create a subsidiary to which ownership of its building or equipment is trans-ferred and whose sole function is to lease them back again to the parent corpora-tion.

Special purpose units of general government are excluded here, because the fo-cus is on the corporate sector.

The SNA 2008 states that there is no common definition of a SPE. Maybe be-cause of that the term SPE is not always logically used in the SNA 2008. How-ever, there is other research on the subject done at least in two different Eurostat Task Forces. 33 Part B of the chapter 4 (SNA2008).34 Paragraphs 4.56 and 4.57 of the SNA2008.



document.doc

24.2.2 Other work on identification of SPEs

In addition to SNA2008 the 4th edition of the OECD Benchmark Definition of Foreign Direct Investment (BD4) provides guidelines on how to identify SPEs. The OECD35 criteria are:

”i)The enterprise is a legal entity, a) formally registered with a national authority; and b) subject to fiscal and other legal obligations of the economy in which it is resident. ii) The enterprise is ultimately controlled by a non-resident parent, directly or indirectly.iii) The enterprise has no or few employees, little or no production in the host economy and little or no physical presence.iv) Almost all the assets and liabilities of the enterprise represent investments in or from other countries.v) The core business of the enterprise consist of group financing or holding activities, that is – viewed from the perspective of the compiler in a given coun-try – the channelling of funds from non-residents to other non-residents. How-ever, in its daily activities, managing and directing plays only a minor role.”

In 2010, the Eurostat Balance of Payments Working Group and the ECB Work-ing Group on External Statistics36 agreed on definition of SPE for the compila-tion of FDI statistics in the EU. The definition is in line with the BD4 guidelines.

The Eurostat Task Force on FDI has recommended to use the definition agreed by the BoP WG and the WG-External Statistics as a basis to identify SPEs in the EGR. The Task Force also recommends that SPEs have to be flagged in the EGR for FDI and FATS purposes and it lists criteria that could be used to produce a list of potential candidates for national SPEs using the available information in the EGR. This work is closely related to that of profiling of MNEs. On the one hand, profilers can provide input on the SPEs and other special cases to be flagged in the EGR. On the other hand, profilers could also use the list of poten-tial candidates for SPEs and other special cases generated by the EGR.

Also another Eurostat Task Force has dealt with the identification and the treat-ment of SPEs. The Task Force on recording of certain activities of multination-als in national accounts recommends that the treatment of SPEs in EU member states should conform to the classification table reproduced in Annex IV of this report. The Task Force provides also a decision tree for the identification of SPEs (see also the AnnexIV).

All the guidance on SPEs in current manuals and other research on the subject is taken into account in a recent guide called “Impact of Globalisation on Na-

35 The OECD Benchmark Definition of Foreign Direct Investment, 4th edition, p. 102, Box 6.2.36 External Statistics refer to all statistics that focus on cross border transactions (such as FDI, foreign trade statistics)



document.doc

tional Accounts”37. Chapter 4 of the guide focuses on SPEs. However, the manual still states that a clear definition of SPEs is not available and the work on common conceptual framework on SPEs is one of the priority items on research agenda.

25 Treatment of SPEs in profiling; the global and national perspective

25.1 Profiling raises expectations for a coherent and extensive identification of units generally defined as SPEs. An important outcome from profiling could be flagging SPEs in EGR where different statistical domains such as FDI, NA and SBS could obtain the SPE status in a coherent way.

25.2 The core of the profiling methodology consists of the global perspective, the top-down approach and the use of IFRS standards. This has implications for the enterprise delineation, as well as for recognition and treatment of SPEs.

25.3 The global perspective and the top-down approach imply a precise order in the enterprise delineation: GENs first and then TENs as national parts of them. The IFRS standards are used to delineate GENs: the financial statement dis-closes segments each of which usually consists of several LeUs or parts of them, and a GEN is generally derived from a segment or a GEN closely corresponds to a segment. The IFRS standards allow MNE holdings, financial units, headquar-ters and SPEs being reported in segments. Thus a GEN also may include these types of activities.

25.4 According to the model adopted for profiling (Chapter 2) SPEs are con-sidered as quasi enterprises to have a full coverage of the enterprise group. More specifically, legal units that cannot be included in operational GENs or TENs are considered as SPEs.

25.5 The problem of SPEs has also emerged in the accounting sector. The IFRS 10 standard has been effective from the 1st of January 2010. The principle SIC-1238 deals with SPEs. The IFRS 10 standard introduces a single consolidation 37 See http://live.unece.org/fileadmin/DAM/stats/groups/wggna/Guide_on_Impact_of_globalization_on_na-tional_accounts_FINAL.pdf38 Under SIC 12, an entity must consolidate a special purpose entity ("SPE") when, in substance, the entity controls the SPE. The control of an SPE by an entity may be indicated if: The SPE conducts its activities to meet the entity's specific needs The entity has decision-making powers to obtain the majority of the benefits of the

SPE's activities The entity is able to obtain the majority of the benefits of the SPE's activities through an

'auto-pilot' mechanism By having a right to the majority of the SPE's benefits, the entity is exposed to the SPE's

business risks The entity has the majority of residual interest in the SPE Examples of SPEs include entities set up to effect a lease, a securitisation of financial assets, or R&D activities. The concept of control used in IAS 27 requires having the ability to direct or dominate decision making accompanied by the objective of obtaining benefits from the SPE's


http://live.unece.org/fileadmin/DAM/stats/groups/wggna/Guide_on_Impact_of_globalization_on_national_accounts_FINAL.pdf

http://live.unece.org/fileadmin/DAM/stats/groups/wggna/Guide_on_Impact_of_globalization_on_national_accounts_FINAL.pdf


document.doc

model for all entities based on control, irrespective of the nature of the investee (i.e., whether an entity is controlled through voting rights of investors or through other contractual arrangements as is common in special purpose entities)39.

25.6 There are two alternative sources available for SPE identification: one based on the IFRS standards, and the other based on NACE codes recorded in EGR.

25.7 The IFRS standards suggest two types of legal units to be investigated as SPE candidates: a) the LeUs that are not consolidated in the financial statements of the group, and b) the LeUs that are consolidated in the residual segment of the financial statement typically named as “Other”. For the first type the solution is to flag all LeUs as potential SPEs. For the second type it is necessary to under-stand whether the LeUs within the “Other” segment can be organized in autonomous and meaningful GENs. The LeUs that fall outside the GEN corres-ponding to the “Other” segment should be flagged as potential SPEs. It is note-worthy that the IFRS 10 instructs most SPEs to be consolidated in the financial statement. As a consequence, on one hand most LeUs outside the scope of con-solidation are SPEs, and on the other hand more LeUs consolidated in the resid-ual “Other” segment are in fact SPEs.

25.8 The use of EGR data including NACE code in the identification of SPEs is described in the report by the FDI Task Force40. Concerning the types introduced

activities.An SPE shall be consolidated when the substance of the relationship between an entity and the SPE indicates that the SPE is controlled by that entity. In the context of an SPE, control may arise through the predetermination of the activities of the SPE (operating on 'autopilot') or otherwise. IAS 27.13 indicates several circumstances which res-ult in control even in cases where an entity owns one half or less of the voting power of another entity. Similarly, control may exist even in cases where an entity owns little or none of the SPE's equity. The application of the control concept requires, in each case, judgement in the context of all relevant factors. In addition to the situations described in IAS 27.13, the following circumstances, for example, may indicate a relationship in which an entity controls an SPE and consequently should consol-idate the SPE: (a) in substance, the activities of the SPE are being conducted on behalf of the entity according to its specific business needs so that the entity obtains benefits from the SPE's operation; (b) in substance, the entity has the decision-making powers to obtain the majority of the benefits of the activities of the SPE or, by setting up an 'autopilot' mechanism, the entity has delegated these decision-making powers;(c) in substance, the entity has rights to obtain the majority of the benefits of the SPE and there-fore may be exposed to risks incident to the activities of the SPE; or(d) in substance, the entity retains the majority of the residual or ownership risks related to the SPE or its assets in order to obtain benefits from its activities.

39 Under IFRS 10, control is based on whether an investor has 1) power over the investee; 2) exposure, or rights, to variable returns from its involvement with the investee; and 3) the ability to use its power over the investee to affect the amount of the returns. 40 Final report of the Task Force ”FDI Requirements on the EGR”, p.22 Box 6: Recommendations for the identification of SPEs.



document.doc

by the GGNA manual (Financing and holding companies; Royalty and licence companies; Factoring companies; Leasing companies) profiling could help in-tegrating EGR and desk results, especially based on the financial statements. Good results could also be obtained in delineating empty shells during the UCI evaluation. In general, at the global level flagging the financial sector should be easier.

25.9 Based on either IFRS standards, or the EGR data, or a combination of the two sources, the UCI profiler first flags the suspected SPEs. Next, the partnering country will either confirm or disapprove the suggested SPE status based on the national first-hand sources such as registers and surveys. The SPE delineation in profiling is realised in a close co-operation between the UCI profilers and the partnering countries. Coherence with the GEN implies that for example the de-cision tree for the identification of SPEs of “Eurostat Task Force on the record-ing of certain activities of multinationals in national accounts (TF MUNA)” is not completely applicable. In fact, an entity controlled by another resident insti-tutional unit could be part of another GEN and data consolidation following only the territorial aspects should not be the proper solution. Anyway, specialists in global profiling members will be aware that in terms of automatic procedures and for situations in which a complete international profiling cannot be under-taken the decision trees developed in different Task Forces will lead to good solutions.

25.10 The operational proposal of the FDI task force

Here some important proposals of the FDI task force are copied. The FDI task force report presents the most recent and the most operational definition of SPEs; it is also nearly described as a process to delineate them. So the content of the Box 6 of the report of the FDI task force leads us to issues such as: can it either be used as a starting point in our profiles or as a set of requirements to be fulfilled for further use by the FDI statisticians?

Box 6: Recommendations for the identification of SPEs6.1. SPEs have to be flagged in the EGR for FDI and FATS purposes. Three categories should be identified:

(i) empty group heads, (ii) SPEs used for “passing through capital” and possibly (iii) other SPEs.

6.2. The European SPEs, which are a subgroup of the national SPEs, should be also identified in the EGR.6.3. The definition of SPE (national and European concept) as agreed by the BoP Working Group and the WG-External Statistics should be used as a basis to identify SPEs in the EGR.6.4. The following criteria included in the definition of SPE could be used to produce a list of potential candidates for national SPEs using the available information in the EGR:



document.doc

(i) The enterprise is a legal entity incorporated in a Member State or a branch re -gistered in the country → in the EGR, it should be a legal unit resident in a Member State.(ii) The enterprise is ultimately controlled by a parent not resident in the Member State, directly or indirectly41 → in the EGR, the UCP/UCI should be resident in an-other Member State.(iii) The enterprise has no or few employees → in the EGR, the number of employees should be 0, 1 or blank.(iv) The core business of the unit frequently consists of group financing or holding activities → in the EGR:– units under NACE codes 6420, 6430 or 7010 could perform activities potentially related to SPEs (codes 6492, 6499 and 6619 could be also considered) and,– a legal unit will not very likely be a SPE if there is another entity in the group per-forming a production activity in the same country of residence.

6.5. The information on SPEs could be fed into the EGR in two possible ways:(i) National FDI compilers provide their own list of SPEs to be incorporated and flagged in the EGR.(ii) Alternatively, the EGR produces a list of potential candidates for being SPEs (at national and EU/EA level) using the criteria described in Recommendation 6.5. This list is checked by the national institutional contacts and the validated entities are flagged as SPEs in the EGR.

25 Auxiliary activities and R&D according to SNA

26.1 Introduction

National account statistics are identified as important end-users of data compiled for new profiled statistical units, and hence discrepancies with SNA are to be avoided. BPM6 is considered here next to SNA, since it focuses on cross-border transactions which play a significant role in MNE activities. Therefore, both SNA and BPM6 are to be acknowledged when deciding on how to deal with MNEs’ auxiliary activities and R&D in profiling.

SNA and BPM6 apply the term institutional unit for a unit, supposed to be close to TEN in the profiling terminology, and establishment for a unit supposed to be close to LKAU42.

26.2.1 Ancillary activities and artificial subsidiaries

The SNA 2008 makes a distinction between artificial subsidiaries and a unit un-dertaking only ancillary activities.

The SNA 2008 Chapter 5 discusses the role of ancillary activities. When the ba-sic routine services43 that are required by all enterprises are provided in-house, 41 we should add “or controles affiliates that are all non residents –case of the empty group head”; but the conclusion stays right: the UCI/UCP is resident in another member state;42 Relationships between TENs and statistical units in present statistics (including national ac-counts) will be discussed in more detail in chapter 11.43 Such as keeping accounts; providing communication facilities; purchasing materials and equip-ment; hiring, training, managing and paying employees; storing materials or equipment: ware-housing; transporting goods or persons inside or outside the producer unit; promoting sales;



document.doc

they are called ancillary activities. According to SNA, an ancillary activity is a supporting activity undertaken within an enterprise in order to create the condi-tions within which the principal or secondary activities can be carried out.

According to SNA, ancillary activities are not separately identified for small en-terprises with single location but for larger enterprises with multiple locations it may be useful to treat them as secondary or even as principal activities.

SNA specifies that activities, such as producing own machinery and equipment for the enterprise or carrying out own R&D, are not to be treated as ancillary.

Concerning the treatment of ancillary activities, SNA states that if an establish-ment undertaking ancillary activities is statistically observable or it is in a differ-ent location from the establishments it serves, it may be useful to treat it as a separate establishment unit.44 However, ancillary activities are not satisfying the conditions of being separate institutional units in the SNA framework.

Within SNA, the artificial subsidiary is a subsidiary corporation wholly owned by a parent corporation and created only to provide services to the parent corpor-ation or other corporations in the same group. Artificial subsidiaries are not re-garded as separate institutional units in SNA but as an integral part of the parent, and their accounts are consolidated with those of the parent (unless they are res-ident in an economy different from that of their parent and treated as SPEs).45

However, SNA does not discuss whether activities of artificial subsidiaries may be treated as separate establishments when possible.

In conclusion, attempting to translate the SNA46 guidelines into the profiling ap-proach, SNA allows ancillary activity units to be treated as nationally truncated parts of GENs where GEN represents the principal activity and ancillary activity units form a TEN that is dual coded to represent the national activity which is different from the GEN activity. SNA indicates the similar treatment for artifi-cial subsidiaries. Thus, remaining in line with SNA, ancillary activities or artifi-cial subsidiaries may not constitute a GEN but be dual coded TENs.

26.2.2 Research & Development

According to SNA, research and development is not an ancillary activity, and it is recommended that a separate establishment should be distinguished for it when possible. The output of R&D should be valued at market prices or at the sum of total production costs plus an appropriate mark-up. Moreover, the output of R&D should be capitalized as intellectual property products.47

However, SNA does not discuss directly whether a subsidiary carrying out R&D activity is to be interpreted as an artificial subsidiary or as a conventional subsi-diary which is treated as a separate institutional unit.cleaning and maintenance of buildings and other structures; repairing and servicing machinery and equipment; and providing security and surveillance. Paragraph 5.35 of the SNA2008.44 Paragraph 5.41 of the SNA2008.45 Paragraphs 4.62-4.64 of the SNA2008.46 It is important to have in mind that the SNA is written from a national perspective47 Paragraph 6.207 and 10.104 of the SNA2008.



document.doc

In conclusion, attempting to translate the SNA guidelines into the profiling ap-proach, SNA bypasses altogether the question whether R&D subsidiaries should form a GEN. As SNA recommends R&D to be interpreted as LKAU, brought to the profiling terminology, a national R&D unit should be presented as a TEN with R&D activity code also in case the GEN does not present R&D as its prin-cipal activity. The introduction of dual coding (see chapter 6) will contribute to a solution here.

26 The treatment of R&D and auxiliary activities, with focus on crossing bor-der services

27.1 Introduction

The problems of R&D and auxiliary activities in the frame of the multinational groups are related to different kind of issues.

First, we need to decide whether the R&D or auxiliary activities of the enter-prise group can form a global enterprise (GEN “other” or GEN “auxiliary”). Es-pecially this question arises in cases where the R&D and/or auxiliary activities are organised and reported as own operational segments according to IFRS rules.

Second question is what is the impact of GENs and TENs on treating the finan-cial flows of such activities between countries. Do the GENs and TENs help the identification and measurement of cross-border services.

Auxiliary activities are by nature similar to those of R&D. The only difference is that auxiliary activities are limited in scope to the type of service functions that virtually all enterprises need to some extent. R&D, instead is not necessary to all enterprises.

27.2 Problem statement examples and recommendations

There are several ways to organize the R&D or auxiliary activities inside the group. Three different examples are put in the annex to illustrate the issue. Further examination of the issues and experiences from testing must lead to more concrete recommendations.

27 Branches and the question of VAT positions

28.1 Some specific situations as for example VAT positions need to be further investigated. Here follows some first insights of the phenomena.

28.2 MNEs often open VAT positions in different countries mainly for fiscal reasons. In some cases high levels of turnover and/or exportations are recorded for these, mostly empty, shells. In these cases it may not be clear whether the values are related to other resident subsidiaries or to foreign units of the same group. A practical solution often adopted in a national perspective is to look for units of the same MNEs and consolidate the data. Profiling methodology could



document.doc

be a valid tool also in such situations in defining the activities carried out by the units and finding the proper solutions especially evaluating the intra-group flows (cross-border) or the presence of particular agreements (for example, contract manufacturing) within the MNE.

To be elaborated further



document.doc

List of abbreviations

A&L Assets and LiabilitiesBR Business RegisterBD4 The OECD Benchmark Definition of Foreign Direct Investment, 4th

editionBoP Balance of PaymentsBPM6 Balance of Payments and International Investment Position Manual,

sixth editionBR Business RegisterCBS Centraal Bureau voor de Statistiek (Statistics Netherlands)D&B Dun & BradstreetECB European Central BankEG Enterprise GroupEGR European Groups RegisterENT EnterpriseESS European Statistical SystemEU European UnionFATS Foreign Affiliates StatisticsFDI Foreign Direct InvestmentGAAP (United States) Generally Accepted Accounting PrinciplesGDC Global Decision CentreGDP Gross Domestic ProductGEG Global Enterprise GroupGEN Global EnterpriseGGH Global Group HeadGGNA Expert Group on the impact of Globalization on the National Ac-

countsGNI Gross National IncomeGNP Gross National ProductGOD Group Operating DivisionIDBR UK Interdepartmental Business RegisterIFATS Inward Foreign Affiliates StatisticsIFRS International Financial Reporting StandardsINSEE Institut National de la Statistique et des études économiques (French

National Institute for Statistics and Economic Studies)ISTAT Istituto nazionale de statistica (Italian National Statistical Institute)KAU Kind of Activity Unit LKAU Local Kind of Activity Unit MNE Multinational Enterprise GroupMUNA Multinationals in National AccountsNA National AccountsNACE Nomenclature statistique des Activités économiques dans la Com-

munauté Européenne



document.doc

NSI National Statistical InstituteOECD Organisation for Economic Co-operation and DevelopmentOFATS Outward Foreign Affiliates StatisticsONS Office for National Statistics (UK)P&L Profit and LossesProdcom Production CommunautaireR&D Research and DevelopmentSBS Structural Business StatisticsSNA System of National Accounts (UN)SPE Special Purpose EntitySTATFIN Statistics FinlandSTS Short Term StatisticsSU Regulation Statistical Unit RegulationTEN Truncated EnterpriseUCI Ultimate Controlling InstituteUK United KingdomUN United NationsUNECE United Nations Economic Commission for EuropeVA Value AddedVAT Value Added TaxWPA Work Package A of the ESSnet profiling



document.doc

ANNEX I: BR recommendation manual 2010 chapter 21, the points 21.31 to 21.46

“21.31 For the definition of control in the BR Regulation, the definition given in point 2.26 of Annex A of Council Regulation (EC) No 2223/96 (the European System of Accounts ESA 1995) shall be used:

“Control over a corporation is defined as the ability to determine general corporate policy by choosing appropriate directors, if necessary. “

A single institutional unit (another corporation, a household or a government unit) secures control over a corporation by owning more than half the voting shares or otherwise controlling more than half the shareholders’ voting power. In addition, government secures control over a corporation as a result of a special legislation decree or regulation which empowers the government to determine corporate policy or to appoint the directors.

In order to control more than half the shareholders’ voting power, an institutional unit need not own any of the voting shares itself. A corporation C could be a subsidiary of another corporation B in which a third corporation A owns a majority of the voting shares.

Corporation C is said to be a subsidiary of corporation B when either corporation B controls more than half of the shareholders’ voting power in corporation C or corporation B is a shareholder in C with the right to appoint or remove a majority of the directors of C.

21.32 The definition states that control may be exercised in different ways. The acquisition of an absolute majority (50 % +1) of shareholdings with voting rights is the main instrument used to take control over a legal unit and in the absence of other information it is generally used as a proxy to control. On the other hand, the absolute majority of ownership of the capital share ownership is not always necessary or a sufficient condition to have control.

21.33 It may not be a necessary condition because there may be situations in which a large relative shareholding with voting rights but without absolute majority is enough to take control. This can be due to:

(a) Legislation, contracts or agreements affecting control;

(b) Absenteeism in meetings on the part of other shareholders. This is more a de facto situation and difficult to prove in practice.

21.34 It may not be a sufficient condition because the ability to effectively exercise control depends on the ability to actively participate in the decision-making process. This may be limited by:

(a) Shareholdings with limited voting rights;



document.doc

(b) Statutory provisions that limit the transferability of shares;

(c) Temporary suspension of voting rights.

21.35 Effective minority control means having effective control of a unit without holding the majority of voting stock. It does not include indirect control via a majority-controlled subsidiary. The most common case is a minority but large shareholder and a very large number of dispersed small shareholders, none of whom hold a significant share of the capital. The minority shareholder can thus exercise effective control insofar as no majority of shareholders is really able to oppose it. However, it is possible that the small shareholders could join forces in order to have more influence over strategic decisions. Effective minority control is, in general, difficult to prove in practice and a shareholding between 10 and 50 percent is generally regarded as influence, not control.

21.36 Situations vary considerably from country to country and depend on the legal framework concerning corporate governance, i.e. the legislation that regulates the allocation of property rights and control of enterprises in the economy. In particular, the principles vary sometimes notably between civil law systems and common law ones.

21.37 Control can be a de facto situation without legal basis or other proof and in such a case it is not to be stored in the business register. Strategies of outsourcing, aimed at reducing production costs and increasing productivity, such as exclusive sales or supply contracts, may generate dependency of one legal unit on another without any direct participation on the part of the latter in the capital share of the former. A legal unit can thus be ‘captured’ by another unit without being owned by it. The link may be a commercial contract, which ensures the ‘de facto controlling’ legal unit the exclusive rights to the work of the ‘subordinate’ unit. In the case of natural persons, such subordinate units have often been referred to as false self-employed in business demography. At least in theory such a subordinate can cease this control by closing down their legal unit. This is not possible for a subsidiary that is owned by the parent.

21.38 According to the SU Regulation (Annex III C, Explanatory notes 4), ”The subsidiary enterprises of a subsidiary enterprise are considered to be subsidiaries of the parent enterprise. (…)”. This means that a parent unit may have indirect control over a legal unit (sub-subsidiary) through one or many other subsidiaries. Indirect control does not require the parent unit to own a majority of an integrated shareholding in the capital share of the sub-subsidiaries. The difference between control and ownership is shown in Figure 3.

21.39 The example in figure 3 shows that X has indirect control over unit C, even though it owns indirectly (60 % * 30 %) + (60 % * 30 %) = 36 % of its capital share, X controls C through its two subsidiaries A and B by combining their voting rights in the meeting of C (30 % + 30 %) = 60 %. On the other hand, Y owns (40 % * 30 %) + (100 % * 40 %) = 52 % of C, but has no power



document.doc

to control it, since the voting rights that it has in its meeting of C amount only to 40 % of its capital share. In other words, the voting rights resulting from paths X-A-C and X-B-C have to be cumulated to determine actual control (sometimes this is called ‘cumulative control’).

Figure 3 : How control can differ from ownership

21.40 Control is a (direct or indirect) relationship between legal units such that either one legal unit is controlled by exactly one other legal unit or it is not controlled by any other legal unit. The ownership of a unit or a group of units is related to the holding of its assets, and determines the distribution of financial flows and income. If a unit or group of units is owned by shareholders, its ownership is vested in the shareholders collectively and can be seen as diffused among the legal units that own its shares in proportion to their shareholdings and independently of voting rights.

21.41 In fact, deriving control links from the ownership structure between legal units defines an operational hierarchical structure of the enterprise group with one legal unit at the top (the global group head) which is not controlled by any other legal unit and which controls all other legal units in the hierarchy. Therefore it is necessary to also record minority intermediate shareholdings into the business register, in case there is indirect control where the links can only be derived from the complete ownership structure.

21.42 The control hierarchies in Figure 4 are derived from the example in Figure 3 and should be recorded in the business register.

Figure 4 : How to record control links from ownership structures



document.doc

21.43 The group head can be either resident in the country that compiles the business register, if the group is domestically controlled, or abroad. Unless otherwise noted, ‘group head’ refers to the ultimate or global group head, not to a national (or European) group head, which has a foreign parent.

21.44 If the group head is a resident legal unit, it must be recorded in the national business register as a single legal unit, which may possibly form an enterprise in combination with other legal units, according to the principles stated in Chapter 7.

21.45 The statistical concept of the enterprise group is different from the accounting concept, as can be derived from the Seventh Council Directive (see the Annex). In fact, as it is stated in explanatory note 3 of section III C of the Annex to the SU Regulation “this definition (of accounting groups …) is not suitable for statistical analysis because they do not constitute mutually exclusive, additive groups of enterprises. A statistical unit known as ‘enterprise group’ based on the ‘accounting group’ concept must be defined by applying the following four amendments:

- consider accounting group at the highest consolidation level (group head);

- include in enterprise group units whose accounts are entirely integrated in those of the consolidating company;

- add majority-controlled units whose accounts are not included in the overall consolidating by virtue of application of one of the criteria allowed by the Seventh Directive, i.e. difference in the type of activity or small relative size;

- discount temporary links of less than a year.”

Operational rules for identifying control links

21.46 Control is a complex concept in economic terms. Statistical operational rules need to rely on observable criteria: Proof of control. Therefore it is sufficient that at least one of the following applies, in order to identify a link of direct or indirect control between two legal units:

1) A legal unit directly owns more than 50 % of the voting rights of another legal unit (direct control);

2) A legal unit indirectly owns more than 50 % of the voting rights of another legal unit, through subsidiaries (indirect control);



document.doc

3) Existence of special legislation decree or regulation, which empowers the government to determine corporate policy or to appoint the directors of the legal unit;

4) A legal unit fully consolidates the accounts of another legal unit, according to the criteria of the Seventh Directive, and no other legal unit consolidates the same legal unit (control by virtue of full consolidation);

5) Administrative sources, collecting declarations in application of specific laws for market regulation, provide the information that a legal unit controls one or a set of legal units, even though it owns less or 50 % of its voting rights (effective minority control) and no other legal unit owns more.

Note — It may be possible that two rules, e.g. both cases 1 and 4, could apply simultaneously. As one unit cannot be controlled by two different units, the de facto controlling unit should then be chosen. Case 4 can in general be considered as weaker than 1, because there can be consolidation situations with less than 50% ownership and situations with over 50% ownership without consolidation.”

End of quotation out of the BR recommendations manual.



document.doc

ANNEX II: Background information related to autonomy

The concept of autonomy in the present European Statistical System (ESS) and in the SNA

II.1 The concept of autonomy in the Statistical Units Regulation

The Council Regulation (EEC) N. 696/93 “on the statistical units for the obser-vation and analysis of the production system in the Community” does not clearly define the concept of autonomy for a given statistical unit. It mentions this concept in different areas of the Annex:1. Section II/A. legal, accounting and organisational criteria: “In order to define

units that are recognisable and identifiable in the economy, legal or institu-tional criteria must be applied. In some case, legally separate units must be grouped together as they are not sufficiently autonomous in their organisa-tion.”

2. Section III/A Enterprise: “The enterprise is…….. which benefits from a cer-tain degree of autonomy in decision making, especially for the allocation of its current resources”.

3. Section III/B Institutional unit: “The institutional unit is an elementary de-cision-making centre characterized by uniformity of behaviour and decision-making autonomy in the exercise of its principal function. ….. In order to be said to have autonomy of decision in respect of its principal function, a unit must be responsible and accountable for the decisions and actions it takes”

4. Section III/C Enterprise Group: “An enterprise group…… . … can have more than one decision-making centre, especially for policy on production, sales and profits. It may centralize certain aspects on financial management and taxation”.

Concluding remarks on SU Regulation: differences in the concept of autonomy between enterprise and enterprise group are due to different types of “decision making processes”: allocation of current resources for the enterprise (but what are they?) and financial management and taxation for the group. Somehow it seems that time is the discriminant: for the enterprise, autonomy is related to de-cisions taken in the short and medium term; for the enterprise group, autonomy is related to strategic, long term decisions.

II.2 The concept of autonomy in the new SNA

The SNA is “the internationally agreed standard set of recommendations on how to compile measures of economic activity” (chapter 1, A.1.1), at the same time “the SNA …..serves as a coordinating framework for economic statistics” be-cause it “is seen as the conceptual framework for ensuring the consistency of the definitions and classifications used in different, but related, statistics” and it



document.doc

“acts as an accounting framework to ensure the numerical consistency of data drawn from different sources…” (chapter 1, E.1.57).According to this point of view, the SNA should not only be the point of arrival, but mostly the starting conceptual basis, on which to build up - or base (?) adapt (?) - the concept of autonomy.

II.3 The main conceptual element of the SNA is the following:

The Institutional Unit (IU) is the fundamental unit identified in the SNA. It is defined as: “…. an economic entity that is capable, in its own right, of owning assets, incurring liabilities and engaging in economic activities and in transac-tions with other entities. The main attributes …. are:a. an IU is entitled to own goods or assets in its own right: … able to exchange

the ownership of goods and assets in transactions with other IU.b. it is able to take economic decisions…..for which it is itself held be directly

responsible and accountable at lawc. it is able to incur liabilities on its behalf…..d. either a complete set of account…..exists for the unit or it would be possible

and meaningful, from economic viewpoint, to compile a complete sets of accounts if they were to be required” (chapter 4, A.4.2).

From this SNA approach it can be concluded that the focus is on the economic meaningfulness of the basic statistical unit; from there on the ability to exchange ownership (buy and sale), to take decisions and to incur liabilities; and not on the legal basis. The words “meaningfulness” and ability need further explanation. First we have to assess the meaningfulness and from there the ability.

II.4 Other fundamental points are:

The two different types of ownership: “The legal owner of entities such as goods and services, natural resources, financial assets and liabilities is the IU en-titled in law and sustainable under the law to claim the benefits associated with the entities”….”The economic owner as goods and services, natural resources, financial assets and liabilities is the IU entitled to claim the benefits associated with the use of the entity in the course of an economic activity by virtue of ac-cepting the associated risks” (chapter 3, B.3,21, B.3.26). “…when the expres-sion of ownership or owner is used and the legal and economic owner are dif-ferent, the reference should be understood to be the economic owner” (chapter 3, B.3.27)

Valuation of transactions: “Market prices of transactions as defined as amount on money that willing buyers pay to acquire something from willing sellers; the exchanges are made between independent parties and on the basis of commer-cial consideration only, sometimes called “at arm’s length” (chapter 3, E.3.119).



document.doc

National boundaries: “The total economy….consists of all the IUs which are resident in the economic territory of a country. … The concept of residence in the SNA is not based on national or legal criteria. An IU is said to be resident in a country when it has a centre of predominant economic interest in the economic territory of that country” (chapter 2, B.2.19).

In all of these statements the economic aspects are said to be predominant as compared to the legal ones (see what has been shown in bold).

II.5 At the same time the SNA contradicts these “predominant economic ele-ments” when it says: In practice some IU are controlled by others and thus in such cases autonomy of decision is not total and may vary over time. Legally in-dependent holding of assets and liabilities and autonomous behaviour do not al-ways coincide. In the SNA preference is generally given to the first case aspect because it provides a better way to organise the collection and presentation of statistics….” (chapter 2, B.2.16). And“…. However, each individual corporation should be treated as separate IU, whether or not it forms part of group……Although the management of a subsi-diary corporation may be subject to the control of another corporation, it remains responsible and accountable for the conduct of its own production activities…….Another reason for not treating groups of corporations as single IU is that groups are not always well defined, stable or easily identified in prac-tice” (chapter 4, B.4.51-4.52).

In these points the SNA seems to follow a pragmatic policy because even if “its usefulness is limited in some cases” the autonomy is considered in terms of leg-ally independent holding of assets and liabilities.

Concluding remarks on SNA concepts: even if the SNA prefers economic as-pects (see transaction evaluation at arm’s length and the concept of economic ownership), at the same time it suggests that these items are preferably identified under "legal" viewpoint and inside the National boundaries (in order to simplify the statistical data acquisition?). The SNA does not want to consider the existence of complex organisational structures (among which enterprise groups at national or global level are the main and the most clear) and the dependence links that characterize the relation-ships between the different legal units of these structures.

The fact that the SNA does not consider the existence of the complex organisa-tional structures (at national or global level) and the dependence links that char-acterize the relationships between the different legal units of these structures must lead to a reconsideration of or addition to what is described here in the SNA. The issue of the data collection is to be considered after initial establish-ment of the most adequate statistical unit. Adaptation of the statistical unit to



document.doc

solve data collection problems should only be done by exception in the cases where it has appeared to be impossible to collect according to the usual SUs.

As we can see in practice, operational economic structures of an Enterprise Group can vary from the formal legal structure. Here we can not find a general appearance. Some groups operate economically according the legal structure, al-though combinations of legal units can be necessary. With other groups the oper-ational economic structure has no or only a few correlation with the legal struc-ture. Following the legal structure in the last case leads to meaningless economic figures/data in statistics. On the contrary, the “enterprises” that we propose are more realistic (and also more pragmatic) that the present IUs; they can be made consistent with the other concepts of the SNA.



document.doc

ANNEX III: Variables for reconciliation with SBS statistics, NA use and their links with accounting standards:

III.1 The needs of NA48 are presented along of 3 different lines: - the components of value added as a source for sectorial accounts (as well by

Institutional Sectors as by industry);- the basic data for evaluation of fixed capital (elements coming from the bal-

ance-sheets or the so-called “financial statement”)- bases for wealth estimates (coming from the rest of the balance-sheets); these

data could also be useful for checking the fixed capital elements as part of the balanced items of a global account.

III.2 The IFRS accounting standard (International Financial Reporting Standard) provides a template of Profit &Losses accounts, also for operational segments of a MNE. This is a tentative version of a very simplified data set that try to include enough variables to produce the main aggregates of sectorial accounts for NA flows; it includes nearly all variables that appear in SBS requirements (missing at this moment change in merchandise inventories) plus as few variables as ne-cessary to have complete accounts. These accounts and the data of these seem very difficult to obtain when the En-terprise statistical unit has insufficient autonomy regarding its production pro-cesses from a financial (production process related) point of view, but, in this case, the existence of a specific “enterprise” including holdings, managing activ-ities (and other residual ( mainly auxiliary)activities of the MNE) is very prob-able; this residual “enterprise” would also include most of the “production’ ori-ented financial-type variables.

III.3 Very important: it works from a by-activity version of the P&L ( in which purchases, wages and salaries, depreciation of fixed capital, etc are displayed explicitly) and not from the frequently displayed version, which is by- function (in which the main variables are “cost of sales, distribution costs, administrative costs, etc”). So that the existence (and MNE the agreement to display) of this by-nature version of the accounts is crucial.The beneath table with variables reflect the Enterprise and Truncated Enterprise statistical units, resulting in the operational (truncated) results (remark: in some cases the information of this table is less detailed that requested by the present SBS questionnaire e.g. on splitting between sales of own production, sales of merchandise etc; the same for inventories):

48 See also the UNECE Handbook on the impact of globalisation on National Accounts, 2011. In this handbook also attention is given to topics like Special Purpose Entities (SPEs). (http://live.unece.org/fileadmin/DAM/stats/groups/wggna/GuideNA_April.pdf)


http://live.unece.org/fileadmin/DAM/stats/groups/wggna/GuideNA_April.pdf


document.doc

P & L by nature (not by function)

(consolidated for the GEN or the TE)

Content of the boxes

1st evaluation of VA( by the production

side )=

“Production” = Proxy of the “Revenue” variable?

+ pure T-O

+ other activity products

+- produced inventories

T-O (sales to 3rd parties) net of granted dis-counts & rebates (IAS18) if possible split between sales of merchandise and other sales. Newly produced inventories and fixed as-sets (incl grants on fixed assets)

+- Other products and expenses Patent, franchise & brand royalties etc; more useful if split between products and charges.

- Used purchases

- Other external expenses

Purchases of goods (raw material, equip-ments and parts etc), services and mer-chandises, net of in inventories of re-ceived discounts and rebates +ancillary costs on purchases (CIF etc)

2nd evaluation of VA( by the components) =

+ Employees wages and salaries Including bonuses and share of results+ Taxes on products and indirect taxes

+Depreciations Physical and value change = Current operating result

(optional)Close to net operating result of the NA

+- Other current products and charge Other current non recurring products & charges (e.g. sellings of assets, depreciations

of goodwill)Operational result

III.4 Depending on the size of autonomy some more variables can be associated to the enterprise unit, but this will not be a general situation. So these variables can be considered as in between belonging to the Enterprise unit and the Enter-prise Group unit. The next table gives some insights in these variables. Also large investments can belong to this category. The investment variable needs more attention because decision making on investments can done on different levels depending on the size of the investments and/or the kind of the invest-ments (strategic or operational).



document.doc

Financial resultOperational result+ Treasury

products Net costs of net financial debt- Cost of debt

+- Other financial products and costs- Taxes costs Taxes on result + deferred taxes + participa-

tions of employees in the results+- Net result quota-share of minority owning of non-consolidated interests

When the MNE has a real influence on de-cisions

+- Pre-tax net result of:. non-continued activities

. ceased or being sold activities IFRS 5Net result

- Minority interestsPart of the Net result accruing to minority holders, for globally consolidated affiliates

Net result, share of the group

III.5 The IFRS also defines templates for A&L (asset & liabilities) tables. As argued in the introduction (see 7.3) the variables presented in the A&L-accounts are applicable to the Enterprise Groups units’ types and less for the Enterprise units’ types. A very simple table for assets and liability is shown in the tables be-low.

ASSETS Content of the boxes

Non floating assets

Intangible asset GoodwillOther intangible (patents, brands, etc)

Tangible assets

Equipment, machines etcActivity-used buildingsBuildings for hire Biological non current assets (forest, orchard, cattle, etc)

Financial assetsEquity share in associate enterprises (Equity share in TENs of other countries) Other disposable (sellable) financial assets

Non current assets, disposable to be sold ??Deferred taxes (asset type)

Floating assetsStocks Goods and merchandise

Biological current assets (crops, vegetable and fruits, animal products, etc)

Financial floating assets Credits to clients Other financial current assets Cash and similar

Total Assets



document.doc

In order to be able to give sufficient information to the stakeholders (mainly the NA) this table has to be completed by specific data that are to be found in the Notes and annexes to the Annual Report, at least: Tangible assets ( Gross value 1st January + Increase – acquisition, creation, etc - less Decrease – sales, withdrawal of tangible assets, etc = gross value at the 31st

December)Amortisation accounts (Total value 1st Jan + Increase of the year less Withdraw-als (e.g. when selling) = Total value 31st December) The same for depreciations incl in P&L accounts.

Some extra, nationally specific, information can be necessary.

LIABILITIES Content of the boxes

Shareholders funds Shareholders funds

+ Equities + Other non distributed reserve funds+ Non distributed net results + Value of non current assets disposable for sale (directly accounted for in the shareholders funds)

= Total of shareholders funds

- Minority shareholding

= Shareholders funds, part of the group

Non current liabilitiesLong term debtDeferred taxes Long term reserves

Current liabilities

Suppliers credit Short term financial debtShort term part f long term debtEligible taxesShort term reserves

Liabilities directly linked to non current assets disposable for sale

TOTAL LIABILITIES



document.doc

ANNEX IV: SPEs: classification tables

Special purpose entities, financial sector: classification table by Eurostat



document.doc

Special purpose entities, non-financial sector: classification table by Eurostat



document.doc

Decision tree for the identification of SPEs - Eurostat Task Force on the recording of certain activities of multinationals in national accounts (TF MUNA)



document.doc

ANNEX V: R&D problem statement examples

Example 1 : R&D activities of the MNE is centralized in one LeU in one country and sold as a R&D service

In this example, R&D services are separate items over which ownership right can be established.

A MNE has 2 GENs and operates in 2 countries France and Italy. It also has, in France1 a R&D centre, which has a role of supporting the operational entities worldwide. It has financial flows with all the legal units in both GEN1 and GEN2 and in both countries.

To keep the example “simple”, we assume that all other LeUs except that of R&D activities are operating in the market. Thus, there are no other intra-group flows expect those of the R&D LeU4.

The money flows from LeUs 1, 2 and 3 to R&D LeU4 are drawn into the figure.

GEN1 GEN2

Nace: 152 Nace: 141

Empl: 830 Empl: 520

Operating revenue: 2500 Operating revenue: 2900



document.doc

Different options:

a- Consider LeU4 as a separate GEN3 (and TEN3 in France) with an activity of R&D (NACE:721)

b- Split LeU4 into GENs and TENs according to value added (employment?)

c- Allocate LeU4 to GEN1 (NACE: 152) which generates the highest value added (employment?) and split it then into TEN1_IT and TEN1_FR.

d- Allocate LeU4 to GEN1 (NACE: 152) which generates the highest value added (employment?) and then into TEN1_IT, which generates the highest value added inside the GEN1

f- Other solution ………………………………………………………

Purely from the conceptual and theoretical point of view, and ignoring all possible practical or technical problems, the recommended option is b.

However, taken into account the practical constraints, several options remain open. One possible pragmatic solution is to follow the idea in the identification rule 8, introduced in paragraph 5.21. This would mean choosing option a.

Option a could be pragmatic choice also in case of ancillary activities (instead of R&D) even though there the option b is even more recommended as in case of R&D.

Example 2 : R&D of the MNE is included in each LeU in each country

In this example, R&D services are not sold outside a legal unit.

Otherwise, the example is the same: a MNE has 2 GENs and operates in 2 coun-tries France and Italy.

We assume that all the LeUs are operating in the market. Thus, there are no in-tra-group flows.

Compared to example 1, the idea of this example is to illustrate a situation, where similar R&D activities exist but from point of view of the legal structure they are organised differently.

GEN1 GEN2



document.doc

Nace: 152 Nace: 141

Empl: 830 Empl: 520

Operating revenue: 2500 Operating revenue: 2900

Different options:

a- Consider each Leu as a TEN and use the activity codes of LeUs, too.

b- Consider R&D activities as a separate GEN3 (and TEN3 in France and TEN3 in Italy) with an activity of R&D (NACE:721)

c- Other solution …………………………………………………………

………………………………………………


FRIT

ITIT

Leu1 = TEN1_FRNace: 152Empl: 400of which R&D people 50OR: 2000


Leu2 = TEN1_ITNace: 152Empl: 430of which R&D people 50OR: 2500

FRIT

ITIT



Leu2 = TEN1_ITNace: 152Empl: 430of which R&D people 50OR: 2500


document.doc

Example 3 : R&D of the MNE is centralized in one LeU in one country but not sold as such (global production)

Services, such as R&D, are not generally separate items over which ownership right can be established and cannot generally be separated from their production.

For example, merchanting arrangements may be used for the management and financing of global manufacturing processes. It is said in BPM6 paragraph 10.42 that ”In cases where the merchant is the organizer of a global manufacturing process, the selling price may also cover elements such as providing planning, management, patents and other know-how, marketing, and financing. Particu-larly for high-technology goods, these nonphysical contributions may be large in relation to the value of materials and assembly.”

Thus, in such cases, the merchant is involved in activities such as research and development but that activity is not directly recorded as balance of payments transaction of R&D services.


FR

IT

Leu1 Nace: 152Empl: 300OR: 500

Leu3 Nace: 141Empl: 420OR: 500

Leu2 =Nace: 152Empl: 330OR: 500

LeU4: R&D and sales officeNace: ?Empl: 450OR: 4500

500

500

500


document.doc

In this example LeU4 in France is the R&D centre of the group and it also sells all the final products to customers. The responsibility for the production process is in France. The other three legal units are producing the products based on cost-plus contract. Their operating revenue is costs plus fixed fee.

Different options:

a- Consider LeU1 and LeU 2 as GEN1, Leu3 as GEN2 and LeU4 as GEN3 with an activity of R&D (NACE:721)

b- Consider LeU1 and LeU 2 as GEN1, Leu3 as GEN2 and LeU4 as GEN3 with an activity of wholesale (NACE:46)

c- Consider LeU1 and LeU 2 as GEN1, Leu3 as GEN2 and LeU4 as GEN3 with an activity of manufacturing (NACE:152)

d- There is only one GEN…

e- There are two GENs… Allocate LeU4 into GEN1 (NACE: 152) and GEN2 (NACE: 141) based on…?

f-

g- Other solution …………………………………………………………



document.doc

ANNEX VI: Example of profiled GEG and the impact on business statistics

The goal of the following example is to show the impact on the business statistics of taking into consideration the global dimension of the statistical units in the case of 2 countries C1 and C2. It also shows that several options are possible to delineate the GENs and the TENs among which one needs to choose.

The impact will be measured in country C1. In this country, there are 8 legal units among which 6 belong to a multinational enterprise group and 2 are independent legal units (LeU 10 and LeU11).

Description of the situation and the relationships between the units without taking into consideration the multinational enterprise group:

LeU9Nace: 7010

Empl:50OR:1500

C1

C2

LeU1Nace:151

Empl:0OR:2000

LeU3Nace:141

Empl:0OR:1800

LeU2Nace:152

Empl:0OR:4000

LeU6Nace:741Empl:150OR:500


LeU4Nace:4642

Empl:0OR:3000

LeU8Nace:4772

Empl:50OR:6000




3000

6000

800

500

500

5002000 800

12002800

1000

500

6000

380

300

300

20

1000

The above schema presents the legal units in the two countries C1 and C2, their activity and their relationships in terms of turnover. The blue arrows show intra group turnover inside a countryThe green arrows show export turnoverThe purple arrows show turnover outside the group

The description of the overall activities based on legal units' observation leads to the following statistics:



document.doc

Country Activity NACE code

Employ-ment

Operating revenue

Export revenue

intra-group revenu

esC1 Manufacture of wearing apparel 141 0 1800 0 800C1 Tanning and dressing of leather 151 0 2000 0 2000C1 Manufacture of footwear 152 200 5000 500 1200C1 Wholesale of clothing and footwear 4642 0 3000 3000 3000C1 Retail sale of footwear and leather goods 4772 100 6000 0 0C1 Activities of head offices 7010 50 1500 1000 1500C1 Other human resources provision 7830 1000 1000 0 1000C1 Total in country C1 1350 16300 4500 9500

C2 Manufacture of wearing apparel 141 150 800 0 800C2 Specialised design activities 741 150 500 500 500C2 Retail sale of footwear and leather goods 4772 50 6000 0 0C2 Total in country C2 200 6500 500 500

From now, we take into consideration the existence of the multinational group of enterprises and the relationships between its legal units.

The multinational EG has two global activities: one of manufacturing of wearing apparel (Nace: 141) and one of manufacturing of footwear (Nace:152). It is composed of 9 subsidiaries and is acting in two countries: C1 and C2.

The headquarters (LeU9) are in C1.

In C1, the 2 global activities are represented into 2 truncated enterprises. TEN1 performs the activity of manufacturing of footwear and TEN2 performs the activity of manufacturing of wearing apparel. TEN1 sells a part of its production outside the group and the rest to a wholesale company in the group. TEN2 sells all its production to the same wholesale company. This company exports all the production to a retail trade company in C2. This retail trade company belongs to the group.Moreover, in C1, there is one legal unit which manages the employees for all the legal units.

In C2, the two global activities are represented into 2 truncated enterprises. TEN3 performs the activity of specialised design and TEN4 mainly performs an activity of retail trade.



document.doc

Legal units' structure of the group

GEN1Nace:152Empl:842


LeU9Nace: 7010

Empl:50OR:1500

C1

C2

Leu1Nace:151

Empl:0OR:2000

LeU3Nace:141

Empl:0OR:1800

LeU2Nace:152

Empl:0OR:4000



LeU8Nace:4772

Empl:50OR:6000


LeU4Nace:4642

Empl:0OR:3000

It is assumed that:LeU1, LeU3, LeU4, LeU5, LeU6 and LeU7 have no access to the market LeU1 accesses the market with operating revenue of 600LeU9 (the Headquarters) has global revenue from TEN1 and TEN4, but it is not possible to split the revenue by TEN. It also has outside revenue from royalties for licensing. LeU9 is also funding the activity of TEN3.

The proposed profiling takes into consideration 3 GENs: GEN1 is performing the manufacturing of footwear (Nace:152), GEN2 is performing the manufacturing of wearing apparel (Nace: 141) and GEN3 is the constituted of the transversal activities. In that case, it is constituted only of the HQ. In that case, the activities of LeU4 and LeU5 are split into GEN1 and GEN2 and more precisely between TEN1 and TEN2 in country C1.



document.doc

Other possibilities to constitute the GENs:

1) Create only 2 GENs (GEN1 and GEN2) and split the HQ activity into the 2 GENs. This option needs to have information on the criteria to apply to split the activity.

2) Create 3 GENs, but arrange the LeUs differently: Include LeU5 and LeU6 in GEN3 LeU5 is working for both GEN1 and GEN2. However, it is very important to

associate the employees to the production in order to calculate productivity ratio like VA/employees. Moreover, LeU5 has relations only with other subsidiaries in the same country and ancillary information to split the activity is available (number of employees by subsidiary).

LeU6 is performing a design activity for GEN1. However, it has financial relations with GEN3 which is funding it. The choice of including LeU6 in GEN1 or GEN3 is made balancing between the functional structure (LeU6 activity is used by GEN1) and the financial structure (GEN3 is funding its activity).

3) The treatment of trade activities can be made on several ways: In C1, LeU4 sells production from both TEN1 and TEN2, and its activity is split

into the two TENs In C2, LeU8 sells production from both GEN1 and GEN2 and it is integrated into

TEN4 that belongs to GEN2. The choice between these two solutions depends on the information transmitted

by the group itself and the ancillary information available to split the activity of one legal unit into two different TENs.

Observation by activity:

One important question is how to treat the part of the GENs outside the country of the UCI.In the example, GEN1 (Nace:152) can be split into one TEN in C1 (Nace:152) and one TEN in C2. It has been chosen to attribute to TEN3 in C2 the Nace corresponding to the design activity, and not the Nace of the GENs it belongs to since no production takes place in the country. Proceeding that way does not make any difference between including TEN3 in GEN1 or including TEN3 in GEN3.

In C2, again for TEN4, the Nace 4772 corresponding to the retail trade activity has been chosen rather than the Nace code of the GEN (manufacturing of wearing apparel). This decision is based on the fact that the most part of the revenue is generated by the retail trade activity (the production activity is marginal in TEN4). The treatment of LeU8 in country 2 differs from the one of LeU4 in country1. Indeed, the retail trade activity report both to GEN1 and GEN2, but in country 2, GEN1 does not have any production activity (but only a service activity).

Treatment of intra-group flows

Several options are possible: Remove the intra-TEN flows Remove the flows inside a country, even extra-TEN Remove all intra-group flows, including between 2 countries

In the example, the intra-country flows have been removed, except intra-flows between TEN1 and TEN5 but intra-group flows between 2 countries have been kept in the operating revenue.



document.doc

Economic structure of the group



TEN5Nace: 7010

Empl:50OR:1500

C1

C2 TEN3Nace:741Empl:150OR:500

GEN3Nace:7010

Empl:50

TEN1Nace:152Empl:692OR:4600



Taking into consideration this new economic structure leads to the following schema of the economies and the flows between the new economic entities. TEN1 to TEN5 belong to the group and TEN6 and TEN7 are respectively the same than LeU9 and LeU10.



document.doc

Description of the situation and the relationships between the units taking into consideration the global and truncated enterprise:



TEN5Nace: 7010

Empl:50OR:1500

C1

C2 TEN3Nace:741Empl:150OR:500




500

5006000

6000

500

1000

500 28001200

1800

1000

The description of the activities based on truncated enterprises' observation leads to the following statistics:

countryNACE code

Employ-ment

consolidated operating revenue

Export revenue

C1 Manufacture of wearing apparel 141 308 2200 1200C1 Manufacture of footwear 152 892 5600 1800C1 Retail sale of footwear and leather goods 4772 100 6000 0C1 Activities of head offices 7010 50 1500 1000

Total in country C1 1350 15300? 4000

C2 Specialised design activities 741 150 500 500C2 Retail sale of footwear and leather goods 4772 200 6000 0

Total in country C2 350 6500 500

The changes in the figures between the statistics based on legal units and statistics based on enterprises are, in this particular example, the following:

The activity of "tanning and dressing of leather" disappears, since it was an integrated activity of the manufacture of footwear.

The support activities in country C1 disappear as well ("retail sale of footwear and leather goods" and "other human resources provision") since they were serving only economic entities in the group.



document.doc

The comparison between the figures by TEN and the figures at the global level of the group are not consistent on the level of the revenue and on its split by activity codes. Indeed, the group presents its results according to 3 segments:Manufacture of wearing apparelManufacture of footwearAdministrative services (completely internal to the group)

According to the group, no operating revenue is dedicated to an activity of sales, but the total revenue is split between the manufacture of wearing apparel (4100) and the manufacture of footwear (5700). Moreover, the group removes all intra-group flows, so the global revenue (9800) is lower than the sum of the countries revenue (14800).

This problem of inconsistencies by activities can be solved by the double codification. Depending on the geographical level the figures are relating to, the Nace code of the TEN or the Nace code of the GEN can be used.

The following graphics allow comparing the situations by using legal units or truncated enterprises as statistical units:

Figures on employment by activity according to the statistical unit

0

200

400

600

800

1000

1200

1400

1600

Manufactureof wearing

apparel

Tanning anddressing of

leather

Manufactureof footwear

Wholesale ofclothing and

footwear

Retail sale offootwear and

leathergoods

Activities ofhead offices

Other humanresourcesprovision

total countryC1

Legal unitsTruncated enterprises

The use of the enterprise rather than the legal units allows giving a picture of the employment more in line with the economic activity.



document.doc

Figures on operating revenue by activity according to the statistical unit

0

5000

10000

15000

20000

25000

Manufactureof wearing

apparel


leather

Manufactureof footwear

Wholesaleof clothing

andfootwear

Retail saleof footwearand leather

goods


Otherhuman

resourcesprovision

total countryC1


The use of enterprises instead of legal units reduces the operating revenue by removing intra-group flows between the operating units in country C1. The revenue in the support and integrated activities disappears since it was internal revenue inside the group.

Figures on exportation by activity according to the statistical unit

0

500

1000

1500

2000

2500

3000

3500

Manufacture ofwearingapparel


leather

Manufacture offootwear

Wholesale ofclothing and

footwear

Retail sale offootwear and

leather goods


Other humanresourcesprovision


The figures on exportation do not change in total, but the share between activities changes, the exportations are affected to the production activities rather than to a wholesale activity.


Work package B: preliminary report 2011 ESSnet … · Web viewCountry Activity NACE code...

Documents

Transcript of Work package B: preliminary report 2011 ESSnet … · Web viewCountry Activity NACE code...