23Rd Meeting of the Wiesbaden Group on Business Registers

1

23rd Meeting of the Wiesbaden Group on Business Registers
International Roundtable on Business Survey Frames

Washington, D.C. 17 – 20 September 2012

Jean Ritzen
Statistics Netherlands
Session No.3

Business Register Profiling of Multinational Enterprises

International vs. National profiling: what makes the differences?

1. Introduction

There is no longer discussion that very large companies need special attention in business statistics because they have a large contribution to the statistical outcomes. The complexity of large companies can vary according several criteria. They can be fully domestic operating according a simple or a complex structure, legally and/or operationally. But they can also operate internationally, again in a simple or complex way. Complex international structures occur more and more in a globalising world. This makes that we have to judge the meaningfulness of national figures, collected nationally at large multinational companies. Some countries already introduced the method of profiling for large and complex companies to establish the statistical units’ structure of these companies or Enterprise Groups. The profiling is restricted to the national parts of the Groups,named truncated Enterprise Groups, however. If the national parts of the same Groups are nationally profiled in more countries, this will lead to inconsistencies in statistics between these countries. The sum of the national figures will not be the real totals for the EU or globally. Large and complex Multinational Groups should be profiled from an international perspective. This paper will deal with the differences between national and international profiling and also with the main consequences. International profiling is being developed and tested in the ESSnet on profiling Large and Complex International Enterprise Groups. The ESSnet is part of the MEETS program of the European Union.

2. Statistical units and profiling

2.1 Units, classifications and definitions are basic elements in statistics. It is in the definition of statistics that unit are classified according a classification system and that variables belonging to units are well defined. The delineation of units in the business area is much more complex than in area of persons. A lot of attention has been given and is still being given to business related statistical and other related units. The core of the European system of statistical units is the enterprise. This unit plays a basic role in the statistics which describe the economic production process.

2.2 Statistical units are units to which statistical data (output) relate. Other types of units exist for data collection like reporting units and observation units. The observation unit reflects the unit on which data are collected. These can be parts of the desired statistical units and the collected information should be transformed into statistical units’ information. The reporting unit is the unit which provides the data in data collection.

2.3 The essential criterion that has to be fulfilled to be an enterprise as a statistical unit is the autonomy in decision making related to the current resources in the production process. This is according the definition in the EU statistical units regulation. The present definition implies a kind of bottom up approach in the delineation of the Enterprise, because it is defined as the smallest combination of legal units under common control.

2.4 The use of legal units (including natural persons who carry out economic production activities) is essential for the identification of the enterprise. Enterprises are equal to legal units in many times. But also in many times, legal units can not be considered as enterprises in their own. They only can be considered as enterprises in combination with other legal units, which are under the same control. All legal units under the common control together are called an Enterprise Group. This group presents itself in reality.

2.5 It is proposed to change from the bottom up approach in establishing the enterprise into a top down approach, what means the establishment of the enterprise from an analysis of the group. This analysis is called profiling.Definition of profiling:

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.

2.6 The next diagram illustrates the result of profiling. Five autonomous enterprises can be distinguished within the Enterprise group, which has eleven legal units under common control. Some legal are to be considered as enterprises in their own, other must be combined into one enterprise. The identification of the enterprise follows from this combination.

Diagram 1

3. Reasons for profiling and the variation(s) in it

3.1 The operational economic structure of a business can differ from the legal or administrative structure of it. This is the main reason for profiling. The need for a profiling action can come from:
a. compilers of statistics: the big impact of large units (groups) on statistical output requires a solid analysis of a statistical structure of a group in order to compile high quality meaningful figures;
b. (International) (in)consistency in statistics
c. an enterprise group itself because of desired agreements for data reporting, also for reduction of response burden;
d. business register rules.

3.2 Profiling can vary from very light to very intensive, depending on the complexity of the Group. In principle profiling implies staff involvement for analysis. But it could be concluded to use automated procedures in the case of simple situations of e.g. a group consisting of a small number of legal units. If this group is also operating fully domestically (so in only one country) it could be decided to establish the enterprise as equal to the Group according automated procedures.

3.3Analysis by staff will be necessary in more complex situations or if a group is operating as an international or globalised organisation. (Very) active involvement of the group itself is necessary in profiling the most complex and very large groups. The profiling is defined as light ff this involvement is not or only marginal necessary. International operating Groups with parts in more than one EU member state are known from the European Enterprise Groups Register (EGR). The results of profiling will also be put in this register at the end of the profiling process.

3.4 In general there are four standard stages in profiling:
- The preparation
- The definition of the units' structure
- The implementation
- The regular update of the profile.
The content of each stage will vary depending on the kind of the profiling and the complexity of the group.

4. The definitions and models developed for profiling

4.1 Profiling should be done according agreed definitions and models. The bottom up approach and thetop down approach will not lead to very significant differences for small and simple groups. For large and complex groups the different approaches can result in big differences, if the bottom up approach is possible here at all. The more is this the case with large international globalised groups.

4.2 The proposals for adapted definitions, restricted to enterprise groups, are:
Enterprise Group: the combination of all legal units under common control.
Enterprise: a part of an enterprise group that produces goods or services and that is autonomous regarding decision making on the current resources.
The proposed formal definition also includes the units which are active, but not operating as a group or as part of a group.[1]

4.3 An important distinction is to be made between fully domestic operating groups and international/global operating groups. Full domestic groups only operate in one country, but also these can have a complex structure that needs intensive profiling. In both cases the groups are mentioned Global Enterprise Groups (GEG) and the enterprise(s) within them Global Enterprise(s) (GEN). The national parts of the GEG are called Truncated Enterprise Group (TEG) and the national part of the GEN is called Truncated Enterprise (TEN). It may be clear that in the case of an all domestic group the TEG is equal to the GEG and the TEN(s) are equal to the GEN(s). It can be concluded that one general units model is sufficient now as is shown in the next diagram.

Diagram 2: Generalised statistical units model

Special purpose Entities are in separate boxes because of the special goals and functions of these units.

5. National profiling

5.1 Only national statistical objectives are reasons for national profiling. Large groups which can be fully domestic or large national parts of multinational groups are analysed into the units’ structure to be used in national statistics. In the process national users are involved as are representatives of the (national part of) the group. In the pure national profiling control along abroad is not being recognised. This can be done however in profiling according the improved national profiling in which control along abroad is or can be taken under consideration, see next diagram 3.

Diagram 3: (Improved) national profiling

The results can beshown as is done in the diagram 1 presented in 2.6 for a fully domestic group or as is shown below in the diagram 4. In national profiling the term truncated has no special meaning but is used here for uniformity.

Diagram 4: Example of result of national profiling a MNE

In this case all NSIs of countries in which the group operates are profiling the national parts independently.

6. International or global profiling

6.1 In international or global profiling the units’ structure is being established by a top down analysis of the whole group, regardless the borders. This results in the establishment of global enterprises (GENs). It can be shown as is done in the next diagram 5.

Diagram 5: The top down analysis in profiling

(IFRS: International Financial Reporting System; OS: Operational Segment).

Annual reports of the group can give a good first indication of segments which can be distinguished and which can be a good starting point in the analysis for the establishment of global enterprises (GENs). The final establishment is a result of a statistical analysis, because it must fit the statistical needs and requirements. The statistical users have to be consulted for this. After the establishment of the GENs, the TENs can be defined as national parts of the GENs.

6.2. The general units’ model is being applied different now as is shown in the next diagram 6.

Diagram 6: Model of International (global) profiling

The process starts with the constitution of the global group that consists of all legal units under common control. The group is analysed into GENs and after this the TENs are defined. So the process is top down and goes from the left to the right in the diagram 6. The top down approach can also be used in the national profiling, starting from the group that is under domestic control, but mostly this is done bottom up here.

6.3 The results of the global profiling are now different as is shown in the next diagram 7

Diagram 7:: Example of result of international profiling a MNE

The term “truncated” has a real meaning now, because the truncated enterprise must be seen as a national part and not longer as an independent unit.

7. Differences between national and international (global) profiling

7.1 There are big differences between the approaches of global profiling and national profiling. These differences appear in all stages of the profiling process. The differences regard content, process, benefits and costs.

7.2 Differences in content:

• In global profiling the units’ profile is fully similar and comparable for all countries in which the MNE operates, where this can (will) differ in national profiling;

• Statistics based on the units’ structure from global profiling will be consistent, within the countries and along the countries;

• Information collected at the MNEs in global profiling is uniform and comparable for all countries, where this can/will differ in national profiling in each country for the same MNE;

• The defined core variables are much more harmonised, because these are discussed with the MNE in global profiling. These can/will differ from country to country in national profiling. Also this will lead to better harmonised statistical output;

• All countries involved should analyse effects on time series;

Conclusion: global profiling will lead to improved quality in statistics in international context.

7.3 There are big differences in processes and procedures:

• One country (NSI), the country of the UCI will have the lead in profiling the whole MNE (GEG). It will take the initiative to contact the MNE, will do the analysis of the MNE structure and will do an initial proposal for the global units’ structure, what means regardless the national borders;

• The country of the NSI has to do consultation of the statistical users on needs and consequences;

• The national parts (TENs) are direct consequences of the global units (GENs);

• The NSI of the UCI will inform and consult colleague NSIs in which the MNE has substantial activities on the proposed structure, eventually for consultation statistical users;

• The NSI of the UCI will centrally discuss proposals with the MNE for the whole group to get agreement on the structure for statistics and for data reporting;

• The NSI of the UCI will inform NSIs of all countries in which the MNE operates on the agreements with the MNE and on the consequences for the countries and will ask for comments and/or agreement of the NSIs. This is done using developed procedures and tools;

• After agreement the profile report will be established and the results are to be registered in the EGR (and in the national business registers);

• An agreement is to be made for a coordinated and harmonised implementation of the new structure in the statistics;

Where in national profiling a NSI is only responsible for the national part of an MNE, the MNE is contacted by the same number of NSIs as it has countries in which it operates. This can be done according differentnational practises. In global profiling, one NSI is representing “the statistical world” for the whole MNE

Important issues for successful global profiling are:

• Countries must be willing to agree on central global profiling (subsidiarity?)

• The big importance of communication and co-operation between NSIs

• The co-operation of the MNEs to provide information for central (global) profiling

7.4 Differences in benefits and costs:

A. Benefits of global profiling in relation to national profiling:

• Improvement of international overall statistical quality. This is to be seen as a net result. One or some NSIs may have an opposite opinion for their national statistics; important improvement is in the international comparability and in the increased meaningfulness of statistical data.

• Improved relationship with MNE. The relationship is more efficient and effective. The units’ structure will have a better fit with the operational structure of the MNE and the MNE will recognise itself in a better way. This will also lead to less (perceived) burden for statistical reporting.

B. Costs of global profiling in relation to national profiling:

•The costs for the NSI of the country of the UCI will increase, but if all countries participate there will be some balancing. Costs will especially increase for countries in which many large and complex MNEs have their UCI. These must also coordinate activities of “partnering countries” (Non-UCI countries in which the MNE operates). Models are needed to analyse.

• Costs for the MNE itself will probably decrease after some initial investments

• The elapse time to finish a profile action is quite long because of high requirements in communication, with the MNE and with the colleague NSIs.

8. Concluding remarks

8.1 Some main conclusions are:

•Globalisation leads to the need of adaptation of systems (frames and processes);

•There will be more and more mutual interdependencies between countries. The consequence will be that countries themselves will have less independency;

•Globalisation goes beyond European boundaries, which leads to a need for tuning different statistical systems (e.g. EU and UN);

•Profiling MNEs requires national and international communication and co-operation;

•Statistical units delineation is a statistical activity and needs to fulfil statistical requirements. The statistical users of the profiling results are important stakeholders and should be partners in the process;

•Implementation using prototyping approach and Revision SU-regulation is to be considered;

•More MNE commitment because of recognition;

•Lower response burden at MNEs (after initial investments);

•Double coding can contribute in bringing together (sub)global and national interests.

•The Euro Groups Register will have a crucial role as a future instrument for international coordination. Results of profiling must be registered in it.

8.2 The ESSnet on profiling large and complex Enterprise Groups, that is part of the EU MEETS program, has developed methodological background and practical guidelines for global profiling. These are being tested by the partnering NSIs and a number of additional statistical institutes. The ESSnet will be concluded at the end of 2013.

8.3 A model for analysing the costs of global profiling is being developed and collection of data to test it has started.

8.4 Some challenges for the future can be mentioned:

•The approach is to be proved by testing followed by adoption and implementation. The proof of the pudding is in the eating. Implementation is to be done according an agreed strategy (gradually of big bang?

• Further development of methodology and guidelines

• Invitation to all EU-countries to join in testing the new approach

• Communication with non-EU areas to discover and discuss possible partnering in the approach (nevertheless different units systems)