IV. Foreign affiliates trade in services statistics
- International Monetary Fund
- Published Date:
- March 2003
4.1. For both goods and services, international sales may be effected not only through the transactions between residents and non-residents that are recorded in balance of payments accounts as specified by BPM5 (and in the 1993 SNA’s external account for goods and services), but also through direct investment enterprises or affiliates established in the countries of foreign customers.59 For services, this method of serving foreign markets is particularly important because it is often the only method that permits the close and continuing contact between service providers and their customers necessary to compete effectively with indigenous firms.
4.2. In the present Manual, statistics describing the overall operations of affiliates are termed “foreign affiliates trade in services statistics”, or “FATS statistics”. Consonant with the Manual’s theme and purpose, its recommendations for compiling these statistics have been designed and presented with services in mind. However, except for the particular activity and product breakdowns suggested, most of the recommendations are equally applicable to goods or services and may be considered in developing a generalized framework for statistics on affiliate operations.
4.3. The most pertinent information on the operations of affiliates may be considered to be that on their sales. Services delivered through transactions between residents and non-residents are measured in terms of sales (apart from any services that may have been donated), and a comparable measure must be available for affiliates in order to measure services delivered through them on a parallel basis. Although, as discussed below, the present Manual recommends a broader programme of data collection, it recognizes that some countries may, at least initially, limit themselves to the collection of statistics on sales because these most directly support the monitoring of commitments under GATS.
4.4. While sales may thus be considered to be the most important information to collect for FATS, additional information is generally required for an adequate assessment of the economic effects of affiliate operations and of measures to liberalize the delivery of services through the commercial presence mode of supply. For example, information on value added allows output originating within the affiliate to be distinguished from output originating in the firms that supply it with intermediate inputs. Similarly, information on employment is required to assess the impact of affiliates on labour markets. Accordingly, the Manual recommends multiple indicators, or variables, for FATS, rather than only sales.
4.5. FATS statistics may be developed for both foreign-owned affiliates in the compiling economy (inward FATS) and foreign affiliates of the compiling economy (outward FATS). Because, under GATS, countries make commitments with respect to the supply of services in their own economies rather than services they supply abroad, the most directly related data with respect to commercial presence may be those on the activities of foreign-owned affiliates in the domestic economy. Nonetheless, the reason countries make these commitments is to secure commitments on the part of other countries, with a view to enhancing the ability of their firms to supply services in those countries. For commercial presence, this supply is tracked by data on outward FATS, which therefore must also be considered relevant.
4.6. In addition to being more directly related to the compiling country’s own commitments under GATS, data on inward FATS are often easier to collect than data on outward FATS. The entities covered are located in the compiling country, and data for them would ordinarily already be included in the country’s domestic enterprise statistics. Thus, compiling data for them may involve only identifying the foreign-owned subset of domestically located firms and tabulating existing data for them. For outward FATS, in contrast, the entities covered are located outside the compiling economy and generally would not be covered by existing data. Furthermore, there may be legal or practical obstacles to surveying them directly; generally, the data would have to be collected from resident direct investors rather than from the foreign affiliates themselves. The present Manual recognizes that for these reasons, many countries may initially limit their FATS statistics to those related to inward investment. Nonetheless, it notes that some countries have successfully compiled data for outward FATS as well.
4.7. Because one country’s inward FATS statistics provide information on the outward FATS of partner countries, exchanges of information among partner countries have the potential to provide countries that do not collect data on outward FATS with information on the overseas activities of their own multinational companies. For such data to be useful, it is important that they be compiled using standardized definitions and methodologies, and in this regard the Manual can play an important role in promoting comparability. In addition, international organizations can, by republishing member country data, serve as clearing houses for such information. The value of such clearing houses can be considerable, inasmuch as they can help to achieve consistency in presentation and greatly reduce the number of contacts required to assemble the data.
4.8. Foreign direct investment financial transactions and related investment position (stock) and income measures are not, strictly speaking, FATS variables because they do not pertain to the overall operations of foreign affiliates but relate only to transactions between and positions with direct investors and their foreign affiliates. In addition, FDI measures are ordinarily compiled with respect to transactions and positions with all foreign affiliates, whereas FATS variables are, as discussed in paragraphs 4.17-4.24 below, to be compiled only with respect to affiliates in which the direct investor holds a majority interest.
4.9. Notwithstanding those differences, FDI statistics should be considered an important adjunct to FATS statistics. Countries that cannot implement the compilation of FATS statistics immediately may find that FDI statistics can provide an alternative interim indicator of commercial presence. In addition, FDI statistics can be used in conjunction with FATS statistics to indicate the extent to which the operations of affiliates were financed with funds from direct investors, as well as the extent to which the income generated by affiliates accrues to direct investors. The present Manual recommends that FDI statistics be compiled as specified by BPM5 and BD3. For convenience, those guidelines are summarized in box 7.
Box 7.Measuring foreign direct investment
Following BPM5 and BD3, direct investment is the category of international investment that reflects the objective, by a resident entity in one economy, of obtaining a lasting interest in an enterprise resident in another economy (the resident entity is the direct investor and the enterprise is the direct investment enterprise). The lasting interest implies the existence of a long-term relationship between the direct investor and the enterprise and a significant degree of influence on the management of the enterprise. Direct investment comprises not only the initial transaction between the investor and the enterprise but also all subsequent transactions between them and among affiliated enterprises.
For statistical purposes, a direct investment enterprise is defined as an incorporated or unincorporated enterprise in which a direct investor, who is resident in another economy, owns 10 per cent or more of the ordinary shares or voting power (for an incorporated enterprise) or the equivalent (for an unincorporated enterprise). Direct investment enterprises comprise those entities that are subsidiaries (a non-resident investor owns more than 50 per cent), associates (a non-resident investor owns from 10 to 50 per cent), and branches (wholly or jointly owned unincorporated enterprises) either directly or indirectly owned by the direct investor. By encompassing enterprises that are not majority owned by the direct investor, the concept of direct investment is broader than the concept of majority-owned affiliate used by the present Manual in defining the universe of firms covered by FATS statistics.
Direct investors may be individuals; incorporated or unincorporated private or public enterprises; associated groups of individuals or enterprises; Governments or government agencies; or other organizations that own direct investment enterprises in economies other than those in which the direct investors reside. The members of an associated group of individuals or enterprises are, through their combined ownership of 10 per cent or more, deemed to have an influence on management that is similar to the influence of a single individual or enterprise with the same degree of ownership.
Compilation of statistics on direct investment will entail collection or estimation of three broad types of data -direct investment income, direct investment financial transactions and direct investment position (stock).
Direct investment income consists of income on equity and income on debt, and covers income accruing to a direct investor in one economy from the ownership of direct investment capital (see below) in an enterprise in another economy. Income on direct investment is presented on a net basis for both direct investment abroad and direct investment in the reporting economy (i.e., receipts of income on equity and income on debt less payments of income on equity and income on debt for each). Income on equity is subdivided into (a) distributed income (dividends and distributed branch profits) and (b) reinvested earnings and undistributed branch profits. Income on debt consists of interest payable (i.e., accrued) between direct investors and direct investment enterprises; both interest payable by the direct investor to the direct investment enterprise and interest payable by the direct investment enterprise to the direct investor are included.
Direct investment financial transactions consist of (a) capital provided (either directly or through other related enterprises) by a direct investor to a direct investment enterprise or (b) capital received from a direct investment enterprise by a direct investor. For the economy in which the investment is located, such capital includes funds provided directly by the direct investor and funds provided by other direct investment enterprises associated with the same direct investor. For the economy of the direct investor, such capital includes only funds provided by the resident investor. Direct investment financial transactions are recorded primarily on a directional basis—resident direct investment abroad and non-resident direct investment in the reporting economy—and are broken down into equity capital, reinvested earnings and other capital, the last category being associated with various inter-company debt transactions.
The direct investment position measures the value of the stock of direct investment. In principle, the position should be measured at current market prices as of the dates involved (that is, beginnings or ends of reference periods). In practice, however, there may be some departures from the market price principle. In many cases, book values from the balance sheets of direct investment enterprises (or of direct investors) will be used to determine the value of the stock of direct investments.
Following BD3, in both inward and outward direct investment statistics, the direct investment enterprise should, if feasible, be analyzed both by its own industrial activity and by the industrial activity of the direct investor. The minimum level of detail recommended by BD3 is that of nine major divisions of ISIC.a For the purposes of consistency with FATS statistics compiled as recommended by the present Manual, a further breakdown to the level of the Manual’s ISIC Categories for Foreign Affiliates (see paras. 4.41-4.43 below) would be desirable.
BPM5 and BD3 provide additional details on the treatment of direct investment, including special instructions for the treatment of intercompany transactions between affiliated banks (depository institutions) and between affiliated financial intermediaries (e.g., security dealers). Compilation hints can be found in the IMF Balance of Payments Compilation Guide (Washington, D.C., 1995).aSee BD3, para. 49 and ISIC, Rev.3, para. 133.
4.10. Interest in FATS statistics has arisen from two primary sources. The first is GATS. By including commercial presence and presence of natural persons as modes of supply, it has created a new need for information describing the activities of foreign-owned or -controlled firms in host economies. The information will mainly relate to commercial presence. However, partial information on presence of natural persons may also be available from this source, if employment by foreign affiliates is among the variables collected and if their foreign employees, who have moved temporarily to the country of location of the foreign affiliate, can be separately identified.
4.11. A second source of interest in FATS statistics has been the growing integration or globalization of the world economy. For a variety of motivations—such as to achieve benefits from geographic diversification, circumvent trade barriers, increase proximity to markets, or reduce costs of labour, transportation, or other inputs--an increasing number of firms have expanded their operations beyond the countries of their owners. The need to understand this phenomenon of international operations, and to monitor the performance of the foreign affiliates through which they are conducted, exists quite separately and apart from any trade agreements, and in this regard FATS statistics are an important analytical tool.
4.12. For both of these purposes, FATS statistics are of interest in their own right, but it will often be possible to discern their full significance only when they are viewed in conjunction with other information, such as comparable information on total home- or host-country economic activity or on services supplied through modes other than commercial presence. For example, while the number of employees of foreign-owned affiliates in the domestic economy is useful information in and of itself, a fuller understanding of its significance could be obtained if the percentage of domestic employment accounted for by these affiliates could be computed. For such computations to be possible, compilers will have to devote attention to issues of comparability between FATS variables and statistics on the domestic economy covering the same items.
4.13. To promote such comparability, the present Manual’s recommendations for FATS statistics draw to a very considerable extent on concepts and definitions found in the 1993 SNA—in terms of both the entities to be covered and the selection and definition of variables for measuring their operations and performance. This approach also allows FATS statistics to be related to or integrated with not only statistics for the domestic economy but also similar statistics on the activities of foreign affiliates in the production and distribution of goods. Concerning the latter, the concepts, definitions and recommendations of the present Manual are also, to the extent possible and appropriate, harmonized with those expected, at the time of writing, to be used in the forthcoming OECD manual of economic globalization indicators, which will deal with foreign affiliates as suppliers of both goods and services.
4.14. The present Manual’s recommendations for FATS statistics centre around the answers to four questions. (a) What is the universe of firms to be covered? (b) How are the data to be grouped: by country, by industrial activity, or by product? (c) What variables should be covered? (d) How might the statistics be developed?
Outline of recommendations
4.15. The four questions raised above are addressed in detail below. However, it will be helpful to outline the key statistical recommendations. As to firms covered, the Manual recommends that FATS statistics cover all firms that are majority-owned by a foreign direct investor. Data on such firms in the compiling economy should be grouped geographically, primarily on the basis of the country of ultimate beneficial owner but, if possible, with some data also being provided on the basis of the country of the immediate investor (first foreign parent). Data on foreign affiliates of the compiling economy should be grouped according to the country of location of the firm whose operations are being described. As to grouping by industry or product, the first priority is for a grouping by industry into categories - termed ISIC Categories for Foreign Affiliates - drawn from the ISIC, Rev. 3. Not all FATS variables lend themselves to a product breakdown, but for those that do the breakdown for services should be on a basis compatible with EBOPS. Several specific variables are suggested for collection, to include at least the following basic measures of foreign affiliate activity: (a) safes (turnover) and/or output, (b) employment, (c) value added, (d) exports and imports of goods and services, and (e) number of enterprises.
B. Universe to be covered
4.16. Methodological antecedents for FATS statistics are much less well developed than those for trade between residents and non-residents, where the transactions to be covered are clearly indicated by BPM5. The conventions with respect to residence found in BPM5 and the 1993 SNA can provide clear guidance on the residence of enterprises and the present Manual recommends that they be followed without exception, but no internationally agreed ownership concept exists that was designed specifically with FATS statistics in mind. Nor is there agreement on the types of firms to be covered and, specifically, on whether the statistics should cover all producers or only producers of services. These issues are addressed in turn below.
1. Ownership criteria
4.17. Although GATS does not provide statistical definitions, it does give some indication of the kind of ownership criteria that would be useful in support of the agreement. The agreement refers to concepts of “ownership”, “control”, and “affiliation”. Under GATS, a juridical person (such as a business enterprise) is “owned” by persons of a WTO member country if more than 50 percent of the equity interest in it is beneficially owned by persons of that member country; “controlled” by persons of a member country if such persons have the power to name a majority of its directors or otherwise to legally direct its actions; and “affiliated” with another person when it controls or is controlled by that other person, or when it and the other person are both controlled by the same person (article XXVIII, section [n]). Thus, GATS would appear concerned with cases of majority ownership—where, by that very fact, control typically could be assumed to exist—as well as by cases in which control can be demonstrated to have been achieved with a smaller ownership share.
4.18. Among statistical guidelines, rules relating to ownership appear in BPM5, in BD3 and in the 1993 SNA. All of these use ownership of 10 per cent of the ordinary shares or voting power (for an incorporated enterprise) or the equivalent (for an unincorporated enterprise) as a lower threshold for direct investment, but they also provide rules that are more closely aligned with the ownership and control concepts found in GATS. All three of these harmonized standards define “subsidiaries” as enterprises in which the direct investor owns more than 50 per cent, “associates” as enterprises in which the direct investor owns between 10 and 50 per cent, and “branches” as wholly or jointly owned unincorporated enterprises. In the 1993 SNA, subsidiaries and branches are considered to be “foreign-controlled enterprises”; associates may be included in, or excluded from, this category by individual countries according to their qualitative assessment of foreign control.
4.19. The present Manual considers that it is relevant to consider criteria used or recommended at national and regional levels. Here majority ownership—that is, cases in which the direct investor owns more than 50 per cent of the ordinary shares or voting power in the direct investment enterprise—is seen to have played a key role in determining the subset of foreign affiliates that is covered. A report of a Eurostat task force on foreign affiliates trade indicated agreement by its members that the criterion of majority owned would be used because the concept is very clear and in this way very operational.60 While the report indicated other criteria that could be used in identifying foreign-controlled firms, actual collection of FATS statistics by Eurostat and OECD has been based on majority ownership; in particular, the Joint OECD-Eurostat survey on the activity of domestic firms and of foreign affiliates in the service sector requests that the data reported should relate to majority-owned affiliates. Among the few national presentations, the United States annual presentation of data on sales of services by affiliates covers data for only those affiliates that are majority-owned by direct investors.61
4.20. Although it may in some respects have been conceptually more appealing to categorize firms on the basis of the actual presence or absence of foreign control, majority ownership of the ordinary shares or voting power by a single direct investor or by an associated group of investors acting in concert has been selected as the recommended ownership criterion for FATS statistics in the present Manual.62 Unlike control, its implementation does not require the use of subjective criteria, nor does it require that compilers examine the nature of investments on a case-by-case basis. The absence of a subjective factor has the added benefit of eliminating a potential source of bilateral asymmetry -namely, the possibility that the issue of control will be assessed differently by home- and host-country compilers. Finally, majority-owned foreign affiliates do fall within the scope of GATS definitions of ownership and - all but the rarest instances, typically involving highly regulated economies - control.
4.21. Although the measure of majority ownership recommended by the present Manual differs from the 1993 SNA concept of a foreign-controlled enterprise, the two concepts are alike in that both refer to ownership by a single investor (or investor group). This approach is followed in the present Manual not only for consistency with the SNA (and with BPM5 and BD3, both of which use ownership by a single investor or investor group in defining direct investment), but also because it is only through a single investor or associated investor group that control can be systematically exercised. However, the relevance—both for GATS purposes and for globalization analysis—of other criteria for selection is acknowledged, and the Manual encourages countries that can do so to provide supplemental statistics covering cases in which foreign control may be deemed to be present, even if no single foreign direct investor holds a majority stake.
4.22. Examples of investments that might be covered on a supplemental basis are majority ownership by multiple foreign direct investors, ownership of exactly 50 per cent by a foreign direct investor, and cases in which a qualitative assessment has been made that effective control has been achieved through a minority stake in an enterprise.63 These should be labelled in a way that indicates the nature of the investments in the firms covered by the statistics. Similarly, countries for which any of these groups are embedded in the basic statistics should provide explanatory notes indicating their coverage and, if possible, the aggregate amounts accounted for by them. Such supplemental statistics on affiliates that are not majority owned may be particularly relevant where majority ownership by foreigners is restricted.
4.23. The present Manual is consistent with BPM5 and BD3 in including in FATS statistics all types of firms, including the so-called special purpose entities (SPEs), if they otherwise meet the criteria for selection. Whatever their structure (e.g., holding company, base company, regional headquarters) or purpose (e.g., administration, management of foreign exchange risk, facilitation of financing of investment), SPEs are an integral part of the structure of the direct investment network, and their activities and products should be included, as appropriate, in the compilation of FATS statistics.64
4.24. Statistics on foreign affiliates owned by residents of the compiling economy should include all majority-owned foreign affiliates, irrespective of whether the affiliate is held directly or indirectly and irrespective of whether the direct investor in the compiling economy is the ultimate owner or is, instead, an intermediate owner in an ownership chain.65 However, because the activities of an affiliate held through an ownership chain could be recorded in the outward FATS statistics of the countries of both the ultimate and intermediate owners, countries are encouraged to indicate the aggregate share of FATS variables accounted for by foreign affiliates for which the compiling country is an intermediate rather than final owner.
2. Types of producers
4.25. Because FATS as used in the present Manual stands for foreign affiliates trade in services, it might be expected that FATS statistics should cover producers of services only. However, the Manual recommends that producers of goods be covered as well. Some firms produce both goods and services, and it is only by coverage of all producers that the activities of producers for which provision of services is a secondary activity would be reflected in the statistics. In addition, coverage of all producers allows for the activities of services producers to be examined in the context of statistics covering all firms. As explained in paragraph 4.39 below, the method for attributing and presenting FATS variables by activity and, to the extent possible, selected ones by product, rather than the limitation of the universe of producers to service producers alone, is the mechanism recommended for isolating services from goods.
C. Time of recording
4.26. Following BPM5 and the 1993 SNA, FATS variables should, in principle, be measured and recorded on an accrual basis. Flow variables, such as output and value added, should be for the reference year. Stock variables, such as assets and net worth, should be as of the end of the reference year. Reference years should be defined on a calendar year basis, if possible. Countries that can collect and compile data only on a fiscal year or accounting year basis should provide explanatory notes indicating this practice, as well as information on the extent to which fiscal and calendar years deviate in their FATS universe.
D. Statistical units
4.27. In principle, FATS statistics could be collected at either the enterprise (company) level or the level of individual business locations or establishments. Neither basis of collection is unequivocally superior to the other; rather, each has its own strengths and weaknesses. For example, some financial indicators, such as total assets, are more naturally collected from enterprises than from establishments. However, because enterprises are more likely than establishments to have activities in multiple industries, data that are classified on the basis of primary activity can sometimes be more difficult to interpret for enterprises than for establishments. Although there may thus be advantages and disadvantages associated with each basis of collection, the present Manual makes no recommendation as to statistical units. As discussed in paragraphs 4.69-4.75 below, FATS statistics often will be developed in the context of existing statistical systems, in which the statistical units are already defined, and in these cases there may be little choice in the units that are to be used for FATS.
4.28. Because the statistical units used can have an important bearing on how the statistics should be interpreted, the present Manual recommends that information on statistical units be disclosed in explanatory notes.
E. Attribution of FATS variables
4.29. FATS variables may be attributed or classified in a variety of ways. One way is geographic - that is, in what country did the production take place, and what country is to be regarded as the country of the owner of the producing affiliate? Another way is on the basis of the primary industrial activity of the producer. Some variables may, in addition, be classified by product -according to the types of goods or services produced. Recommendations on each of these bases of attribution are provided below.
1. By country
4.30. The issues to be addressed in attributing variables by country differ as between inward FATS and outward FATS. For inward FATS, a choice must be made between attribution to the immediate investing country and attribution to the ultimate investing country. For outward FATS, the issue is whether to attribute variables to the immediate host country or to the ultimate host country. (As noted in para. 4.24 above, both directly and indirectly held affiliates are to be included in the FATS universe.)
4.31. For FATS statistics, the need to follow investments to their ultimate origins or destinations reflects both the nature and uses of the statistics. These issues are discussed below with respect to each type of investment, and recommended bases of attribution are given.
(a) Inward FATS
4.32. For foreign-owned affiliates in the compiling economy, the question is whether to attribute FATS variables to the country of the immediate investor (first foreign parent) or to that of the ultimate investor (ultimate beneficial owner (UBO)) (for a discussion of these two ownership concepts, see box 8). Commonly, the first foreign parent and the UBO are one and the same, but in many cases they differ. Thus, it is necessary to determine the primary principle to be applied. Some perspective on the matter may be gleaned from a review of relevant information concerning recommendations for FDI statistics, the Eurostat task force report mentioned above and an OECD-Eurostat questionnaire on the subject.
4.33. Both BPM5 and BD3 recommend that FDI balance of payments flows be compiled only in respect of the immediate investing country, a practice followed by virtually all countries. With respect to direct investment positions, BPM5 recommends that these should be allocated to the country of the immediate host or investing country. BD3 suggests that positions should also be compiled separately with respect to the ultimate host or controlling country.
4.34. The Eurostat FATS Task Force report (see footnote 60 above) recommended that the UBO be considered as the first priority to define the level of control because it is the most relevant. It observed that this principle can be very complex to follow in practice, and thus it proposed that initially the country of the first foreign parent could be used as a substitute, with the possibility of later obtaining data on the basis of UBO to be explored. However, in a pilot study, conducted in the framework of structural business statistics, data were successfully collected on the UBO basis by most participants. Also, the United States has collected data on this basis for some time. In the light of this successful experience, the UBO basis is to be recommended in the proposed OECD manual of economic globalization indicators.
Box 8.Immediate and ultimate investors
Foreign-owned affiliates in the compiling economy may be grouped geographically based on the country of the immediate investor (first foreign parent) or the country of the ultimate investor (ultimate beneficial owner). The first foreign parent is the first foreign person in the chain of ownership of the affiliate. The UBO is the first person in the chain - beginning with and including the first foreign parent - that is not majority owned by another person. The following examples illustrate how these entities may be identified in particular cases. In each case, the chain of ownership runs from top to bottom, with the company at the bottom being the foreign-owned affiliate whose ownership is at issue.
|Company A||Company C||Company F||Company I||Company L||Company O|
|Company B||Company D||Company 0||Company J||Company M||Company P|
|Company E||Company H||Company K||Company N|
|Case I.||Company A is both the foreign parent and the UBO of company B.|
|Case II.||Company D is the foreign parent of company E. Because company D is, in turn, majority owned by company C, company C is the UBO of company E; its country would be considered the country of ownership of company Ε in FATS statistics.|
|Case III.||Following the same reasoning as in Case II, company G is the foreign parent of company H, while company F is its UBO. Note that company F’s indirectly held ownership in company H is only 42 per cent—the product of its 70 per cent share of company G and company G’s 60 per cent share in company H. Company H is nonetheless included in FATS statistics as a foreign-owned affiliate because the majority ownership criterion relates to the ownership interest of the first foreign parent (which is the direct investor), rather than to that of the UBO. Despite company F’s lack of majority ownership, it can be presumed to control company H because each entity in a chain of majority ownerships can control the entity below it, including that entity’s actions with respect to the entities that are, in turn, below it.|
|Case IV.||Company J is both the foreign parent and the UBO of company K. Company I is not the UBO because it is not the majority owner of company J.|
|Case V.||Company M is the foreign parent of company N. Because company M is, in turn, majority owned by company L, company L is the UBO of company N. However, company Ν is not covered by FATS statistics because it is not majority owned by its foreign parent.|
|Case VI.||Company O is both the foreign parent and the UBO of company P. Company Ρ is not covered by FATS statistics because it is not majority owned by its foreign parent, but it nonetheless represents a case that may be deemed relevant for the purposes of GATS or globalization analysis. Thus, the compiling country may wish to show data for company Ρ (and other cases of interest) on a supplemental basis (see paras. 4.21 and 4.22 above).|
4.35. Abstracting from practical considerations, the country of ultimate beneficial ownership is conceptually preferable for the attribution of variables concerning industrial activity because that is the country that ultimately owns or controls - and therefore derives the benefits from owning or controlling - the direct investment enterprise. In the light of the relevance of the UBO basis and the demonstration by a number of countries that compilation on this basis is feasible, the present Manual recommends the UBO basis as the first priority for compilation of FATS statistics and the basis on which estimates should be prepared in the greatest detail. However, considering that information on first foreign parents may be available as a by-product of linkages to FDI data and to facilitate comparisons with these data, countries are encouraged to make available some data in which variables are attributed according to the country of the first foreign parent.
(b) Outward FATS
4.36. For affiliates owned by residents of the compiling economy, two options for attribution of FATS variables by country are possible. The variables could be attributed to the country of location of the affiliate or - if the ownership is through a directly held affiliate located in another country - to the country of that affiliate. The present Manual recommends that attribution be to the country of the affiliate whose operations are described by the variables, for that is the country in which the foreign direct investor’s commercial presence exists, and it is the country where the various activities (sales, employment etc.) tracked by the statistics are carried out. This recommendation is consistent with the treatment of foreign-controlled enterprises in the 1993 SNA, in that the value added in production by the enterprise is attributed in both cases to (i.e., is included in the gross domestic product of) the economy of location of the enterprise. To the extent that the statistics may be used in conjunction with statistics on resident/non-resident foreign direct investment transactions recorded in balance of payments accounts, it should be noted that the latter are attributed, following BPM5 and BD3, to the immediate host country, as is appropriate for tracking financial flows and positions.66
(c) Note on the equal ownership of shares by residents of more than one country
4.37. Ordinarily, FATS variables for a given foreign affiliate are attributed in their entirety to a single country of owner. As descriptors of the operations of affiliates, they should not be factored down by ownership shares. Nor should the values of the variables be apportioned between the majority owner and any foreign minority owners. However, where supplemental statistics are provided covering cases in which foreign control has been achieved other than through majority ownership by a single investor, classification dilemmas may arise where direct investors of different countries have collectively achieved majority ownership through holdings of equal shares. Because the ownership is evenly split, the determination of the country of owner has to be made using criteria other than ownership percentages.
4.38. Although it is sometimes difficult to reach a decision in such cases, there is often some factor that would lead to the selection of one country rather than the other. For example, if one owner’s interest in the affiliate is held directly and the other owner’s interest is held indirectly, the affiliate generally would be classified in the country of the owner holding the direct interest. As another example, if one of the foreign owners is a government entity, then the country of that Government would probably be considered the country of owner. Finally, if one of the foreign owners is a holding company or is located or incorporated in a tax haven country, then the country of the other owner would probably be considered the country of owner. In the absence of any such factor that could be used as a basis of attribution, the value of FATS variables may be allocated evenly among the foreign countries of ownership. However, data so allocated may pose problems of interpretation, and efforts should first be made to determine a basis for allocation to a single country.
2. By activity and by product
4.39. Ideally, it would be possible to attribute all FATS variables on the basis of the industrial activities of producers and, in addition, particular variables, such as sales or output, exports and imports by the types of services products produced and sold. Data on a product basis would identify the specific types of services delivered through the commercial presence mode of supply and could most readily be compared with data on services delivered through trade between residents and non-residents. However, some FATS variables, such as value added and employment (discussed below), do not lend themselves to a product classification. Also, for some countries, FATS statistics may be developed as a subset of domestic enterprise or other statistics that are classified only on an activity basis. On this basis, all of the data for a given enterprise are classified in the single activity—often termed the “primary” activity—that, based on some key variable (such as employment or sales), is the largest. Finally, for some purposes, the data may need to be viewed in conjunction with data on stocks and flows of foreign direct investment, which normally would be classified by activity but not by product.
4.40. Taking these factors into account, an activity basis is recommended as the first priority for FATS statistics. However, as a longer-term goal, countries are encouraged to work toward providing product detail for the items that can be classified on this basis. Countries that are building on existing data systems that already include product detail will probably wish to use this detail from the outset in their tabulation and presentation of FATS statistics, because it could help them in monitoring commitments under GATS that are specified in terms of services products. Similarly, countries that are building their FATS data systems from the ground up should consider the feasibility of providing for a product dimension.
(a) By activity
4.41. The present Manual recommends that, for reporting to international organizations, FATS variables be classified by activity according to ISIC, Rev.3 and grouped according to the ISIC Categories for Foreign Affiliates (ICFA), which have been derived from ISIC. These categories, shown in table 3, cover all activities, but with more detail provided for services than for goods.67 This all-inclusive basis of presentation allows activities of services enterprises to be viewed in the context of the activities of all enterprises. In addition, it provides a framework for displaying services produced as a secondary activity by enterprises classified as goods producers. Finally, this all-inclusive approach has been adopted by the countries and international organizations that are actively developing FATS statistics.68
|ICFA headings / elements|
|1.||Agriculture, hunting, forestry, and fishing||01,02,05|
|1.1. Agriculture, hunting, and related service activities||01|
|1.2. Forestry, logging, and related service activities||02|
|1.3. Fishing, operation of fish hatcheries and fish farms; services activities incidental to fishing||05|
|2.||Mining and quarrying||10,11,12,13,14|
|Of which: Services activities incidental to oil and gas extraction excluding surveying||112|
|3.||Manufacturing||15 to 37|
|4.||Electricity, gas, and water supply||40,41|
|6.||Trade and repair||50,51,52|
|6.1. Sale, maintenance, and repair of motor vehicles and motorcycles; retail sale of automotive fuel||50|
|6.2. Wholesale trade and commission trade, except of motor vehicles and motorcycles||51|
|6.3. Retail trade, except of motor vehicles and motorcycles; repair of personal and household goods||52|
|7.||Hotels and restaurants||55|
|8.||Transport, storage, and communications||60,61,62,63,64|
|8.1. Transport and storage||60,61,62, 63|
|8.1.1. Land transport; transport via pipelines||60|
|184.108.40.206. Transport via railways||601|
|220.127.116.11. Other land transport||602|
|18.104.22.168. Transport via pipelines||603|
|8.1.2. Water transport||61|
|22.214.171.124. Sea and coastal water transport||611|
|126.96.36.199. Inland water transport||612|
|8.1.3. Air transport||62|
|188.8.131.52. Scheduled air transport||621|
|184.108.40.206. Non-scheduled air transport||622|
|8.1.4. Supporting and auxiliary transport activities; activities of travel agencies||63|
|220.127.116.11. Supporting and auxiliary transport activities||6301,6302,6303,6309|
|18.104.22.168. Activities of travel agencies and tour operators; tourist assistance activities, not elsewhere classified (n.e.c).||6304|
|8.2. Post and telecommunications||64|
|8.2.1. Post and courier activities||641|
|9.1. Financial intermediation, except insurance and pension funding||65|
|9.2. Insurance and pension funding, except compulsory social security||66|
|9.2.1. Life insurance||6601|
|9.2.2. Pension funding||6602|
|9.2.3. Non-life insurance||6603|
|9.3. Activities auxiliary to financial intermediation||67|
|9.3.1. Activities auxiliary to financial intermediation, except insurance and pension funding||671|
|9.3.2. Activities auxiliary to insurance and pension funding||672|
|10.||Real estate activities||70|
|11.||Renting of machinery and equipment without operator and of personal and household goods||71|
|12.||Computer and related activities||72|
|13.||Research and development||73|
|14.||Other business activities||74|
|14.1. Legal, accounting, market research, and consultancy||741|
|14.1.1. Legal activities||7411|
|14.1.2. Accounting, bookkeeping, and auditing activities; tax consultancy||7412|
|14.1.3. Market research and public opinion polling||7413|
|14.1.4. Business and management consultancy activities||7414|
|14.2. Architectural, engineering, and other technical activities||742|
|14.4. Business activities, n.e.c.||749|
|16.||Health and social work||85|
|17.||Sewage and refuse disposal, sanitation, and similar activities||90|
|18.||Activities of membership organisations, n.e.c.||91|
|19.||Recreational, cultural, and sporting activities||92|
|19.1. Motion picture, radio, television, and other entertainment activities||921|
|19.1.1. Motion picture and video production and distribution; motion picture projection||9211,9212|
|19.1.2. Radio and television activities||9213|
|19.1.3. Other arts and entertainment activities||9214,9219|
|19.2. News agency activities||922|
|19.3. Library, archives, museums, and other cultural activities||923|
|19.4. Sporting and other recreational activities||924|
|20.||Other service activities||93|
4.42. Primarily because the activities carried out by a given firm are usually not limited to its activity of classification, the data recorded against any given activity must be interpreted as an indication of total activity of firms for which the given activity is the most important, or primary, activity, rather than as a precise measure of the activity itself.69 For the same reason, as well as because of differences in the classifications themselves, the extent to which data on resident/nonresident trade classified according to EBOPS can be aligned with data on FATS variables classified according to ICFA is inherently limited. Nonetheless, a correspondence between the two bases of classification may be useful for some purposes, mainly involving activities that tend to be carried out only by firms that are specialized in the activity and tend not to have significant secondary activities.70 For such purposes, the EBOPS categories corresponding most closely to the ICFA categories for services activities have been indicated in annex IV, table A.IV.1. The reverse correspondence is shown in annex IVtable A.IV.2.
4.43. The present Manual recognizes that data for particular ICFA categories may sometimes have to be suppressed (that is, not separately shown) to preserve the confidentiality of data for individual companies. This will most often happen at the most detailed level of the classification, or in the case of smaller countries, or where the data are cross-classified by country or area.
(b) By product
4.44. As a longer-term goal, countries are encouraged to work toward disaggregating by product some or all of the variables—which include sales (turnover), output, exports and imports—that lend themselves to this basis of attribution. Product-based statistics are free of problems of interpretation related to secondary activities, are consistent with the basis on which GATS commitments are made, and are consistent with the basis of classification used for trade between residents and non-residents.
4.45. To the extent possible, the breakdown should be on a basis compatible with EBOPS for services and according to the Harmonized System for trade in goods, to facilitate comparisons with resident/non-resident trade classified on this basis. If this level of specificity cannot be achieved, countries may wish to disaggregate sales in each industry as between sales of goods and sales of services as a first step toward a product basis (see paras. 4.52 and 4.53 below on the sales variable for a discussion of this option).
F. Economic variables for FATS
4.46. A wide range of economic data or variables -operational and financial - in regard to FATS may be pertinent for analytical and policy purposes. The selection of the variables to be collected should be based primarily on their usefulness in implementing GATS and in analysing globalization phenomena. The practicalities of data availability also must be considered. With such considerations in mind, the present Manual recommends that the FATS variables to be collected include at least the following basic measures of foreign affiliate activity: (a) sales (turnover) and/or output, (b) employment, (c) value added, (d) exports and imports of goods and services, and (e) number of enterprises. Although these variables constitute a basic set that can provide answers to a variety of questions, additional measures of foreign affiliate activities may prove useful in addressing specific issues. The Manual suggests several measures that might be considered for collection by countries that are able to compile such additional information. Most of both the “basic” and the “additional” variables were drawn from the 1993 SNA, as were their definitions.
4.47. To provide a concrete illustration of how these variables might be presented, table 4 presents a sample table format in which the basic variables could be presented, classified by activity on the basis of ICFA. Additional tables might be constructed to view the variables from different perspectives. For example, time series or geographic detail could be shown for a single variable by placing time periods or country names, rather than the names of the variables, in the table heading.
|Industry of affiliate||Sales/turnover or output||Employment||Value added||Exports of goods and services||Imports of goods and services||Number of enterprises|
|1. Agriculture, hunting, forestry, and fishing|
|1.1. Agriculture, hunting, and related services|
|1.2. Forestry, logging, and related services|
|1.3. Fishing, operation of fish farms; services incidental to fishing|
|2. Mining and quarrying|
|Of which: Services activities incidental to oil and gas extraction, excluding surveying|
|3. Et cetera|
1. Sales (turnover) and/or output
4.48. Sales and turnover are used here interchangeably to mean the same thing. Following the 1993 SNA (which may be consulted for additional details and examples), output differs from sales because it includes changes in stocks of finished goods and work in progress and because of differences in measurement applicable to activities involving trade or financial intermediation. Output is a superior and more refined measure of activity for most purposes and is recommended as the preferred variable for compilation. However, sales data are easier to collect and may present more options for disaggregation. Thus, there may be a continuing role in FATS statistics for both measures.
4.49. Services activities do not involve stocks of finished goods, and changes in work-in-progress will usually be impossible to measure. In practice, therefore, measured output will be identical to sales for most service activities. For wholesale and retail distribution, although the sales are of goods, the output is defined as a service, equal not to the total value of sales but to the trade margins realised on goods purchased for resale. For financial intermediaries, output is equal to service charges actually levied, plus financial intermediation services indirectly measured, the values of which are estimated from the difference between the property incomes received by financial intermediaries from the investment of borrowed funds, and the interest they themselves pay on such funds (see box 5). For insurance, output is measured not by total premiums earned but by a service charge that takes into account income on technical reserves and also the actual or expected value of claims. In all these cases, output will generally be considerably lower than sales because it, that pass through the enterprise without being considered unlike sales, excludes the amounts - which may a part of its intermediate consumption, constitute a large portion of total operating revenues - that pass through the enterprise without being considered a part of its intermediate consumption.
4.50. Sales measures gross operating revenues, less rebates, discounts and returns. Sales should be measured exclusive of consumption and sales taxes on consumers and value-added taxes. Although lacking the duplication-free quality of value added, the sales variable generally presents fewer collection difficulties and thus is likely to be more widely available than value added. Also unlike value added, sales indicates the extent to which foreign affiliates are used to deliver outputs to customers, irrespective of the extent to which the output originated in the affiliates themselves or in other firms. Further, sales are more comparable than value added with regard to such variables as exports and imports, which are themselves measures of sales.
4.51. In addition to disaggregation by industry and by country (following the principles of attribution discussed earlier), other breakdowns of sales may be useful for particular purposes. One such breakdown is to distinguish among sales within the host country (local sales), sales to the country of the parent enterprise (i.e., the immediate investor), and sales to third countries.71 All three types of sales result from a commercial presence by the home country in the host country. However, only the local sales represent the delivery of output within host economies and thus relate directly to the commitments made under the GATS Mode 3 by those economies. In addition, elimination of the sales to the parent country (which would already be included as imports in that country’s balance of payments accounts) may be desired in an analysis of FATS variables in conjunction with data on the parent country’s trade with non-residents, to avoid duplication.
4.52. As a medium-term goal, countries might attempt to disaggregate sales within each industry as between sales of goods and sales of services so as to obtain a measure of sales of services. As mentioned in paragraph 4.45 above on the attribution of FATS variables, such a breakdown would represent a first step toward a product breakdown of sales. Sales of services would include both sales of services by enterprises that produce services as a primary activity and sales of services by enterprises that produce goods as a primary activity but have secondary operations in services.
4.53. The potential usefulness of this extension is considerable, and countries that are able to provide this data are encouraged to do so. It is likely that a significant share of sales of services by affiliates is accounted for by affiliates whose primary industry is manufacturing or another goods producing industry. For example, sales of computer services might, as noted earlier, be spread among manufacturing affiliates, wholesale trade affiliates and computer services affiliates. If only total sales were collected, then by default only the sales by the affiliates classified in computer services would be taken as a measure of sales of such services, resulting in a significant understatement.
4.55. In a FATS context, employment would normally be measured as the number of persons on the payrolls of foreign affiliates. Employment data are sometimes converted to a full-time equivalent (FTE) basis, in which part-time workers are counted according to the time worked (e.g., two workers on half-time schedules count the same as one full-time worker). Although FTE employment may provide a better measure of labour input, this measure is not as widely available as numbers of employees and may be difficult to implement consistently in the context of internationally varied employment practices. For those reasons, the recommendation of the present Manual is that the FATS employment variable be the number of persons employed. The number should be representative of the period covered, but in the absence of strong seasonal or other fluctuations in employment, it may be measured as of a point in time, such as the end of the year, following national practices.
4.56. Data on employment by affiliates can be used in several ways in a FATS system. They can be used to determine the share of foreign affiliates in host country employment, or to help determine the extent to which employment by foreign affiliates complements or substitutes for domestic (home country) employment by parent companies or other domestic firms. An industry breakdown of affiliates’ employment can yield further insights into the impact of foreign-owned enterprises on specific parts of the economy. Used in conjunction with data on compensation of employees—one of the “additional” variables suggested below—the employment variable may be used in examining compensation practices of affiliates relative to those of domestically owned firms.
Relation to presence of natural persons mode of supply
4.57. Although data would not, in general, appear to be available to identify separately the portion of employment by affiliates that is accounted for by employees from abroad, such information, if available, could be used for elaborating on the GATS presence of natural persons mode of supply. Annex I explores the issues that will arise in constructing a framework of statistics to use during negotiations on this mode of supply.
3. Value added
4.58. The 1993 SNA defines gross value added of an establishment, enterprise, industry or sector as the amount by which the value of the outputs produced exceeds the value of the intermediate inputs consumed. A related concept, “net value added”, is defined as gross value added less the consumption of fixed capital. Gross value added can provide information about the contribution of foreign affiliates to host country gross domestic product, both in the aggregate and in specific industries. For this reason and because it may often be easier to compute (because it does not require estimation of capital consumption) and is thus more widely available, the higher priority should be accorded to the gross measure of value added.
4.59. Although it is defined in terms of outputs and intermediate inputs, value added is also equal to the sum of primary incomes generated in production (compensation of employees, profits etc.). In some cases, depending on the particular data that are available, this equivalence may be exploited in deriving estimates of value added. This alternative could be chosen, for example, if data on intermediate consumption were lacking but information on the various incomes generated in production were available.
4.60. Because it includes only the portion of the firm’s output that originates within the firm itself, value added is a particularly useful measure from the perspectives of both GATS and globalization analysis. It is for this reason that it has been included among the “basic” FATS variables, even though, as a measure that may have to be estimated or derived from other variables, it may be among the more difficult variables to compile. For inward FATS, value added will often be available from regular industrial or enterprise surveys, but for outward FATS it may have to be derived from other variables.
4. Exports and imports of goods and services
4.61. International goods and services transactions of foreign affiliates is another basic indicator relevant to FATS. Both balance of payments data and data provided by parent enterprises and affiliates in separate questionnaires may be appropriate sources for such information. To a large extent, the possibilities for disaggregating total exports and total imports may depend on the sources used to obtain the data.
4.62. Where the data are obtained through linkages with primary data sources for balance of payments transactions, breakdowns by product and by origin or destination will often be possible. In this event, exports and imports of services may be disaggregated, not only by the primary activity of the affiliate according to ICFA but also by product on a basis compatible with the EBOPS.
4.63. Although linkages with balance of payments data may thus provide useful information, it often will be difficult or impossible to separately identify the transactions of foreign-owned firms in those data. Thus, it may be possible to develop the data on exports and imports only through the use of separate questionnaires. In that event, those same breakdowns would be useful, but it is unlikely that a large number of countries would be able to collect the necessary data with the same frequency or in the same detail as provided by balance of payments data. However, it may be possible to disaggregate exports and imports into a few broad categories in which trade with related enterprises would be distinguished from trade with unrelated parties, thus providing information for one of the last of the elements for a phased approach to implementation of the Manual.73 In addition, trade with the country of the parent enterprise could be distinguished from trade with other countries. If possible, those breakdowns should be obtained separately for goods and services. For inward FATS, for example, that would mean disaggregating the affiliate’s exports of goods and exports of services into (a) exports to the parent enterprise, (b) other exports to the country of the parent, and (c) exports to third countries. Imports would be similarly disaggregated.
5. Number of enterprises
4.64. The number of enterprises (or establishments, where that is the statistical unit) meeting the criteria for coverage by FATS statistics is a basic indicator of the prevalence of majority ownership by foreigners in the host economy. That number may be compared with the total number of firms (or establishments) in the economy. It may also be assessed in relation to the other FATS variables because it allows the computation of ratios - such as value added or number of employees per enterprise - that may be compared with the same ratios for domestically owned firms, thus giving an indication of the behaviour of foreign affiliates.
4.65. It should be recognized that the number of firms alone may not give an accurate picture of the overall importance of foreign-owned firms, because of differences in size between these firms and those that are domestically owned. If the foreign-owned firms tend to be larger, for example, then their share in the total number of firms would be smaller than their share in the various measures of operations and would thus tend to understate the role and importance of these firms in host country economies.
4.66. Typically, information on numbers of enterprises will be a natural by-product of collection of data on other FATS variables, rather than a separate object of the data-collection effort. As such, the number is likely to be affected, often significantly, by the level of company consolidation and by thresholds for reporting on surveys. To assist users in interpreting counts of enterprises (or establishments), countries are encouraged to indicate in explanatory notes how the numbers were derived.
6. Other variables
4.67. Although not included as priority items, there are other FATS variables of significance, perhaps for certain countries of equal or greater importance than some of those previously discussed. As is the case for the priority items, comparisons with the total economy and with specific sectors can be effected and used for assessing the impact of foreign-controlled enterprises on home and host economies.
4.68. Among such variables are the following, which are already collected by some countries (definitions are drawn from the 1993 SNA, which may be consulted for additional details):
Assets: entities over which ownership rights are enforced and from which economic benefits may be derived by their owners by holding or using them, including both financial assets and non-financial assets, whether produced or non-produced;
Compensation of employees: the total remuneration, in cash or in kind, payable by an enterprise to an employee in return for work done by the employee during the accounting period;
Net worth: the difference between the value of all assets—produced, non-produced, and financial—and all liabilities;
Net operating surplus: measured as value added (gross), less compensation of employees, consumption of fixed capital and taxes on production, plus subsidies receivable;
Gross fixed capital formation: measured by the total value of a producer’s acquisitions, less disposals, of fixed assets during the accounting period, plus certain additions to the value of non-produced assets realized by productive activity (fixed assets are defined as produced assets that are themselves used repeatedly or continuously in processes of production for more than one year);
Taxes on income: these consist of corporate income taxes, corporate profit taxes, corporate surtaxes etc. as well as taxes that accrue to owners of unincorporated enterprises as a result of the income of those enterprises. Taxes on income include only taxes in the host country of the affiliate and not any taxes paid by the parent in the home country as a result of income earned or distributed by the affiliate. Taxes on income are usually assessed on the total income of corporations from all sources and not simply on profits generated by production;
Research and development expenditures: expenditures for activities undertaken for the purpose of discovering or developing new products (goods and services), including improved versions or qualities of existing products, or discovering or developing new or more efficient processes of production.
G. Compilation issues
4.69. There are two basic approaches, which are not necessarily mutually exclusive, to developing FATS statistics. The first is to conduct surveys that directly request information on the operations of resident affiliates of foreign firms and foreign affiliates of domestic firms. The second, which can be used only for inward investment, identifies the subset of existing data on resident enterprises that is accounted for by foreign-owned firms.
4.70. Whichever of these approaches is taken, there are likely to be links to existing data on foreign direct investment. Where there are surveys for FATS statistics, registers used in collecting FDI data would typically be used to identify majority-owned affiliates for which FATS variables should be collected. Alternatively, key FATS variables might be added to existing FDI surveys. However, because FDI surveys may be conducted more frequently than FATS statistics are compiled (e.g., quarterly rather than annually) and require a quick turnaround, as well as because FATS statistics are needed for only the majority-owned portion of the FDI universe, separate surveys probably would offer a better solution in most cases. Where existing domestic statistics are used as the source of information on FATS, links to FDI data will often provide the means of determining which resident enterprises are majority owned and should be included, as well as the means of identifying the country of owner. In that approach, FATS statistics would be obtained as an aggregation of statistical variables across the foreign-owned statistical population.
4.71. Each approach has its own advantages and drawbacks and it is possible to outline some of the intrinsic differences. However, the criteria for deciding whether an enterprise is foreign owned would be the same in both cases.
4.72. The FDI framework, whether implemented through adding questions to existing surveys or through institution of new surveys covering the majority-owned subset of the FDI population, allows for the compilation of outward as well as inward statistics and provides more options for tailoring the data to specific FATS needs. However, the activity classification used in FDI statistics is generally rather aggregated, and it seems difficult to go beyond basic statistical variables, such as turnover and employment, without designing completely new surveys, which might raise concerns about resource availability and respondent burden. Also, if that approach is taken, special care will be necessary to ensure compatibility with the domestic statistics with which the FATS statistics may be compared.
4.73. The picture of FATS statistics as a subset of enterprise statistics is quite different. It is not possible to compile outward statistics within that framework. However, the activity classification used may be quite detailed and product detail may be available for sales or turnover. In addition, a comprehensive set of statistical variables can generally be provided.
4.74. In many cases, a mix of the two approaches may work best, with the “FDI context” being used for the compilation of outward FATS statistics and for the identification of foreign-owned companies, and the “enterprise statistics” context being used for the compilation of inward FATS statistics, with a more detailed activity breakdown and a more comprehensive set of variables. An extended business register might be a suitable means of maintaining such information. That approach has already been adopted in some countries that use them to maintain data on foreign ownership. The present Manual bases its recommendations on both approaches, recognizing the advantages and disadvantages of each and the need for countries to have flexibility in adapting the recommendations to their individual statistical infrastructures and maximizing the use of existing data.
4.75. One of the main challenges for the new statistical domain stems from the fact that the collection and definition of FATS statistics may draw on expertise and responsibilities that are scattered among multiple institutions, such as central banks, national statistical offices and various ministries. Thus, as statistics are developed, there will be a need for close cooperation among the different institutions involved.
H. Summary of recommendations
4.76. The main recommendations of the present chapter on the compiling of FATS statistics may be summarized as follows:
FATS statistics should cover those affiliates in which the direct investor (or an associated group of investors acting in concert) holds a majority of the ordinary shares or voting power. However, countries are encouraged to provide supplemental statistics, covering cases in which foreign control may be deemed to be present, even though no single foreign direct investor holds a majority stake;
FATS variables should be compiled for all foreign affiliates, not only those affiliates in services. However, the activity classification to be used for reporting to international organizations provides more detail for services than for goods;
For statistics on foreign-owned affiliates in the compiling economy (inward FATS), the first priority for geographical attribution should be the country of ultimate beneficial owner. However, to facilitate linkages with FDI data, countries are also encouraged to provide some data in which attribution is based on the country of the first foreign parent. Statistics for foreign affiliates of the compiling economy (outward FATS) should be attributed based on the country of location of the affiliate whose operations are being described;
Because it is the basis required for some variables as well as the basis on which data probably are, at present, most widely available, an activity basis is recommended as the initial priority for FATS statistics. However, data on a product basis is recognized as a longer-term goal, and countries are encouraged to work toward providing product detail for those variables that lend themselves to this basis of attribution (namely, sales (turnover) and/or output, exports and imports);
For reporting to international organizations, FATS variables should be disaggregated according to the ISIC Categories for Foreign Affiliates (see table 3). Any product detail that is developed should be disaggregated on a basis compatible with EBOPS (see table 2);
The Manual recommends that the FATS variables to be collected include at least the following basic measures of foreign affiliate activity:
Sales (turnover) and/or output;
Exports and imports of goods and services;
Number of enterprises.
Additional measures are suggested for countries wishing to expand FATS data collection beyond this basic set;
The Manual envisages that a variety of sources and methods may be used to collect and compile FATS statistics. Separate surveys may be conducted or links may be made with domestic enterprise statistics that are already collected. In either case, there are likely to be links with existing data on foreign direct investment.