ESSNET on International Profiling, Work Package B (partial):
Statistical units (Revised 12/2010)
0. Introduction 1
1. Economic environment and statistical initiatives 1
2. History and Background 2
3. Scope and organisation of the ESSnet on Profiling 3
4. Relationship between the Statistical Units and Feasibility study (WP A) 4
5. Relationship between the statistical units study and the EGR Project 4
6 Method used in the statistical units study 5
7 The impact on Statistical Units definitions 5
8. A proposal for some revision of the statistical units definitions 7
9. A modelisation of the SU system: the present state 8
10. A need for revision of the EU statistical units system 11
Conclusion 15
Annex : Business Register Regulation No 177/2008 of 20 February 2008 15
0. Introduction
As part of the ESSnet on profiling large and complex MNEs, this study examines the statistical units and their relationships in the context of globalisation. It makes recommendations for changes to the definitions of the Enterprise Group and Enterprise that are consistent with the outcome of the report on feasibility of the same ESSnet (Work Package A). The recommendations have implications for the Business Register and Statistical Units Regulations and provide input for the further development of the EuroGroups Register.
1. Economic environment and statistical initiatives
1.1 The process of internationalisation and globalisation has had enormous impact on the way many businesses operate. This regards not only foreign trade, but also and mainly the internal operational organisation of large businesses. Except for fiscal, legal and other administrative reasons, national boundaries are not longer a criterion for organising the business. Business operations are more and more organised at international or world level, while less notifying the national boundaries as this was the case previously.
1.2 Thus the growing globalisation of economies makes an accurate description of the national and international economy increasingly difficult: it causes some major distortions in the source data that support the compilation of national economic statistics. It also creates, or increases the size of, inconsistencies in the description of (the truncated parts of) the same Multinational Enterprise Groups (MNEs) in different countries.
1.3 In consequence, National Statistical Institutes (NSIs) consider the quality of the data on MNEs, and also all-resident enterprise groups (domestic), with a complex operating and/or financing structure as a cause for concern. If quality of the national economic statistics is affected, this also has a negative impact on the quality of European statistics compiled from these national statistics, resulting in concerns from Eurostat regarding quality.
1.4 It is the same for many international statistical agencies (OECD, UNECE, UN as well as Eurostat) which give more and more attention to the growing importance of MNEs in the world and in national economies.
1.5 In this context, the following two initiatives, completely relevant to this topic, are being undertaken by Eurostat to improve the data quality of MNEs in statistics:
- The Euro Groups Register (EGR) is designed to support compilers of statistics by providing access to an integrated and up-to-date delineation of MNE structures, and to keep information on their operational structure and on the derived structures (the statistical units) for the collection and compilation of statistical data.
- The ESSnet on profiling large and complex MNEs is designed to develop methods for and implementation of co-ordinated profiling process to fill the business register on a permanently updated basis.
1.5 As to profiling its accepted definition is:
“Profiling is a method to analyse the legal, operational and accounting structure of an enterprise group at national and world level, in order to establish the statistical units within that group, their links, and the most efficient structures for the collection of statistical data” (BR recommendations manual, paragraph 19.9).
1.6 The first action of the ESSnet on profiling was a feasibility study in combination with a study of the impact of globalisation and of MNEs on statistical units. This study is part of the broader defined work package B that relates to methodological issues regarding profiling. The results of the feasibility study appear in a separate closely related report.
1.7 The main objectives of the methodological study on statistical units are:
· To supplement the feasibility study with the methodological underpinning of the statistical units structure of enterprise groups to be used in business statistics and as an object of 'profiling'.
· To provide input for the development of the data model and the development of algorithms for delineation of statistical units within enterprise groups in the EGR and national statistical registers.
· To provide input for the process of adapting Council Regulation (EEC) No 696/93 of 15 March 1993 on the statistical units for the observation and analysis of the production system in the Community.
1.8 This report covers all three objectives.
2. History and Background
2.1 In the late 1980s there was general agreement among NSIs that Business Registers, as a cornerstone of the statistical value chain supporting the national economic statistics, required a formal basis.
2.2 The Business Registers (BR) Regulation of 1993 recognised the need for European standardisation of the business units held on national registers; the Statistical Units (SU) Regulation of March 1993 also provided the necessary framework. According to these regulations, both the “Enterprise” and the “local unit” were to be held as statistical units, with the basis of the enterprise being the smallest “autonomous” combination of legal units or a single legal unit if autonomous or independent of any group.
2.3 In more detail, the SU regulation was the result of a discussion of many years. Because the different countries practised the apparently similar definitions of the SU in various ways, it was the intention of the SU regulation to have more coordination in statistics by the use of the same statistical units system in the same way in the EU Member States. Eight units are defined in the statistical units regulation[1].
Two of these units are predominant in data collection and data processing for the European statistics, the Enterprise (ENT) and the Local Unit (LU), a third one having until now an undefined status, the Enterprise Group (EG). These three units are considered apprpriate for registering in the statistical business registers.
Five other units are for analytical purposes, for example for application in the National Accounts: the Institutional Unit (IU, specific to NA, mainly relevant for relations between business sector and other sectors of the economy – households, government etc- and for financial processes), the Kind of Activity Unit (KAU), the Unit of Homogeneous Production (UHP), the Local KAU (LKAU) and the Local UHP (LUHP).
The KAU and the LKAU, however, are more closely related to direct data collection and data processing than the UHP and LUHP.
The use of the units[2] is specified in separate regulations. Important ones are of course the Business Register (BR) regulation, but also the Structural Business Statistics regulation (SBS) and the Short Term Statistics regulation (STS). The ENT and the LU play a major role in the regulated applications of today. The ENT is the central unit in the BR and in the SBS. The KAU is the basic underlying unit in some other regulations like the STS.
2.4 When the SU Regulation was implemented, it adapted well to the situation at that time. Although there could be other reasons to have a critical review on statistical units, the present approach is restricted to the way globalisation of large and complex MNEs reveals the necessity to a revision of their definitions.
2.5 During the 1990s, the growing recognition of problems related to it suggested to Eurostat to “define globalisation as the existence of interactions between enterprises residing in different countries, which are related by links other than market trade, and their socio-economic consequences”. It then became apparent that for MNEs the enterprise group will take in the compilation of many economic statistics the cornerstone role that used to be this of the enterprise in its former use.
2.6 In practice the NSIs, which have to face issues relating to MNEs, referred informally as the problem of “measuring the elephant””. The MNE project (run in 2004-2006 for the purpose of seeing the elephant) showed that it was very difficult to get a good fit in national statistics of activities of large international operating business. It is even more difficult to have the data of these Multi National Enterprises (MNEs) consistent and comparable between the national statistics of different countries. The reasons can be in the statistical units definitions themselves, but also in their applications in statistics compilation in the various Member States.
2.7 Since 1993 many analytical studies have been carried out on the EU statistical units system with comparisons to other systems in the world. The use of the same term with different content leads to a lot of confusion, the Enterprise being the best example. There were also many initiatives to revise the SU definitions and regulation, but without a final conclusion until now. For example, in 2001 a proposal for recommendations on revision was discussed in the EU Business Statistics Directors Group. Another working group presented an operational definition of the enterprise for business statistics. But a decision has still not been made.
2.8 The new BR Regulation, replacing that of 1993, introduced the enterprise group and the enterprises themselves at the national level with the inclusion of foreign ownership data for MNEs, together with a framework for sharing data on structures of MNEs.
2.9 The way was thus open to a European operational programme whose object was to “fulfil the register framework” with MNE information and to clarify the use of the statistical units of the MNEs to gather the main economic statistics, including Structural Business Statistics and Foreign Affiliates Statistics (FATS) and to incorporate them in national or European syntheses such as National Accounts.
3. Scope and organisation of the ESSnet on Profiling
3.1 The scope and organisation is described in depth in the report of the feasibility study. Only a few important parts are repeated here
3.2 The statistical units study, part of Work Package B (WPB), which is the subject of this report, has Statistics Netherlands as the lead partner with DESTATIS (Germany), INSEE (France), ISTAT (Italy) and SFI (Finland) forming the remainder of the team. The other partners in UK (ONS) and Switzerland (FSO) contribute and review the outcomes.
3.3 The methodological and legal background for this work package is recorded in:
· Regulation 177/2008 on Business Registers
· Regulation 696/93 on Statistical Units
· Recommendations Manual on the Business Register (chapters 19, 21, 22):
o The Treatment of Large and Complex Businesses (chapter 19)
o Enterprise Groups (chapter 21)
o Demographic Changes and Continuity Rules for the Enterprise Group (chapter 22)
· Recommendations Manual on the Production of FATS
· International Financial Reporting Standards (IFRS)
3.4 All of the documents for the project are held on BRnet within CIRCA and those that are confidential to the partners are marked as such and have access restricted to the partners. This procedure satisfies the EU legal requirements but is not a solution in the longer term.
4. Relationship between the Statistical Units and Feasibility study (WP A)
4.1 WPB will in the coming years develop a common conceptual framework, methodology, rules and standards for profiling. WPB started with a study on statistical units, with the aim to make an input for an eventual revision of the SU Regulation. It was initially supposed to run in parallel with Work Package A (WPA) which deals with the feasibility of international profiling.
4.2 The two studies have appeared to be more interlinked than foreseen. The work on the feasibility study has informed the outcomes of WPB and vice versa. The proposed options from the feasibility study and the study on statistical units are totally consistent.
5. Relationship between the statistical units study and the EGR Project
5.1 The EGR project in its first stages has developed a model for managing MNEs in a global way. The EGR defines a global enterprise group through a cluster of control, direct or indirect. Operationally, it holds the relationship between each member of a group and its immediate controlling legal unit together with some information about the group as a whole. It currently holds legal units and global group clusters of legal units and can represent them as connected tree structures.
5.2 The development of the EGR requires the introduction of the “enterprise” (or an equivalent of this unit) as specified in the BR Regulation. The EGR holds only MNEs but many of the issues are also relevant to enterprise groups that operate only within one country. Moreover, many “domestic” groups, when they become large, extend progressively outside their home country. It would be paradoxical to totally change their representation at the time they create or buy a foreign affiliate.
5.3 In the present SU Regulation the enterprise is defined as the smallest combination of legal units that is an organisational unit producing goods and services (for the market), with a certain degree of autonomy in decision making especially for the allocation of its current resources. This definition is currently used only at national level, but the EGR requires coherence globally.
5.4 Among the characteristics defining an enterprise, “autonomy” appears essential and has found no fundamental critics, even if autonomy is supposed to be of “a certain degree”, and can be limited to the “allocation of its current resources”, which may be difficult to apply when (physical) investment is concerned.
5.5 The fact that the enterprise is an organisational unit is generally accepted, but two characteristics are not evident. The relationship between enterprises and legal units appear not to be straightforward, the legal units being often split between several operational or organisational units. The limitation of enterprises to national territories is doubtful. It is important to determine whether enterprises can cross national boundaries, an issue that is considered later on in this report.
5.6 Because countries maintain enterprises at the national level, it is possible that activity can be missed, double-counted or be misallocated where an enterprise group spans more than one country. Applying the same standards and definitions, at least for the EU, will improve coherence. This means that the EGR requires not only rules for profiling of MNEs, but also files of profiled “enterprises” within the EU of MNEs that operate at least partly in the EU, for the large and complex cases.