In December 2019, the Bureau of Economic Analysis (BEA) published updated statistics on the characteristics of international services traders in its Survey of Current Business article "A Profile of U.S. Exporters and Importers of Services" and for the first time released statistics on trade in services by industry and by firm type in tables 6.1 and 6.2 in its international services interactive tables. These statistics shed light on the characteristics of the U.S. firms that trade services. For services traders that are U.S. multinational enterprises (MNEs), BEA provides a fuller set of characteristics by linking to its extensive information on MNEs collected on its activities of multinational enterprise (AMNE) surveys of U.S. parents and U.S. affiliates.
This FAQ contains a description of the selected services considered in scope for BEA's current statistics on the characteristics of international services traders; definitions for the types of firms that trade services; technical explanations of how nonsurvey-based trade was attributed to specific companies, how the industries of companies were assigned, and how the BEA trade in services and AMNE surveys were linked; and a comparison of the scope of the selected services and firms included in BEA's current statistics with those BEA released in the past.
What selected services are included in the current statistics?
BEA's analysis of firms that trade services is limited to a subset of international services trade statistics that is based on data collected by BEA, specifically financial services; insurance services; charges for the use of intellectual property; telecommunications, computer, and information services; and other business services.
These selected services cover most international services traded between businesses and accounted for 57 percent of total U.S. services exports and 51 percent of total U.S. services imports in 2017. The remainder of trade in services consists of travel, which includes transactions involving individual persons; government goods and services (including military); and transport and maintenance and repair services (including those traded among businesses). The services that are out of scope for the characteristics of international services traders statistics are mainly based on source data other than BEA-administered surveys; exceptions are certain transport and maintenance services. The surplus on trade in these services totaled $79.0 billion in 2017, mostly reflecting a surplus in travel services.
How does BEA define the types of firms that trade services for its statistics on the characteristics of services traders?
MNEs featured in BEA's statistics on the characteristics of services traders are U.S. business enterprises that are involved in a direct investment relationship with a foreign resident. Direct investment is distinguished from other types of international investment by the implied degree of control or influence.
- U.S. parents are persons, resident in the United States, that own or control at least 10 percent, directly or indirectly, of a foreign business enterprise, which is called a foreign affiliate.
- U.S. affiliates are U.S. business enterprises that are owned or controlled at least 10 percent, directly or indirectly, by a foreign resident, which is called a foreign parent. The term "U.S. affiliates" is used in the characteristics of international services traders statistics to refer to majority-owned U.S. affiliates, which are U.S. business enterprises that are owned, directly or indirectly, more than 50 percent by a foreign resident.
- Non-MNEs and minority-owned U.S. affiliates, which are referred to as non-MNEs in the characteristics of international services traders statistics, are persons, resident in the United States, that are neither U.S. parents nor majority-owned U.S. affiliates. This includes U.S. business enterprises that are minority-owned (by at least 10 percent and at most 50 percent) by a foreign resident. Although minority-owned U.S. affiliates are in a direct investment relationship with a foreign resident, they are not included in the U.S. affiliates statistics on characteristics of services traders, because minority owners typically exert a smaller degree of influence and control and, as such, face less risk and derive less of the benefit of ownership.
There is an overlap in the statistics for U.S. parents and U.S. affiliates because a portion of U.S. parents are foreign owned.1 The statistics for U.S. parents and U.S. affiliates each include the overlap of firms that are both U.S. parents and majority-owned U.S. affiliates. For this reason, the sum of statistics for U.S. parents and U.S. affiliates does not equal the total for MNEs, and the sum of statistics for U.S. parents, U.S. affiliates, and non-MNEs does not equal the total for all services traders in international services interactive table 6.2.
How was nonsurvey-based trade attributed to specific companies?
The selected services included in BEA's current statistics on the characteristics of services traders are composed of services transactions that are primarily collected on surveys administered by BEA and therefore can be attributed to specific companies. However, within the service types included in the total of services, there are source data for certain transactions that are not survey based. These include data for financial brokerage services related to debt securities that come from external data sources as well as data for computer services imports from Canada and trade unions transactions between the United States and Canada, which are provided through a data-exchange agreement with Statistics Canada, without any accompanying company information. In addition, aggregate-level adjustments are made to direct insurance and reinsurance transactions in the estimation of insurance services to better reflect long-term economic conditions and economic accounting concepts. These adjustments are proportionally assigned to individual companies based on their reporting. It was not feasible to leave these transactions out of the selected services included in the characteristics of international services traders statistics, because doing so would risk the disclosure of information of individual companies contributing to the remaining data. Therefore, for each of these cases, the following steps were taken to allocate the transactions to individual companies that did report on BEA surveys or to categorize the transaction as non-MNE and assign an industry.
For financial brokerage services related to debt securities, transactions were assigned proportionally to companies that reported on the Quarterly Survey of Financial Services Transactions Between U.S. Financial Services Providers and Foreign Persons, according to the value of their transactions in brokerage services related to equity transactions and other brokerage services (for forwards, futures, options, and swaps and other financial instruments). For computer services imports from Canada, the values were distributed proportionally to the companies that reported such imports from Canada on BEA surveys.2 Trade union transactions, which measure dues paid by unions in Canada to their U.S.-based parent organizations and financial benefits paid by U.S. trade unions to their Canadian affiliates, were classified as non-MNEs, so they were not linked to BEA's MNE surveys and were assigned to "other industries." For direct insurance and reinsurance, the adjustments to smooth premiums and losses and estimated premium supplements were assigned proportionally to all companies according to the premiums they reported on the Quarterly Survey of Insurance Transactions by U.S. Insurance Companies with Foreign Persons.
How were firms classified by industry?
U.S. services traders were classified under the industry that best describes the major activity of the consolidated U.S. enterprise. The fully consolidated U.S. enterprise includes (i) the U.S. corporation whose voting securities are not owned more than 50 percent by another U.S. corporation and, proceeding down each ownership chain from that U.S. corporation, (ii) any U.S. corporation whose voting securities are more than 50 percent owned by the U.S. corporation above it.3 Entities classified in a particular industry could have activities in a variety of other industries. For example, a firm classified in manufacturing may have secondary activities in information. For MNEs, we use the industry reported on BEA's activities of multinational enterprises surveys and for non-MNEs we use the industry reported on BEA's trade in services surveys, as of the fourth quarter of a given year.
In a small number of cases, there was no industry assigned to companies that reported on the trade in services surveys in the early years covered by these statistics. For these companies, an appropriate industry was assigned by using information that had been obtained from the company on an earlier or later survey, by employing assumptions about companies reporting on certain surveys, or by other means. In a small remainder of cases, the industry of the reporter could not be identified, and the reporter was classified in "other industries." This mostly affected the early years covered by the characteristics of international services traders statistics, meaning that the industry attribution is less precise in those years than in later years. The companies for which an appropriate industry was not assigned accounted for less than 1 percent of total U.S. exports or U.S. imports in any given year, and less than 0.1 percent in 2017. This issue only affects non-MNEs since the industry of MNEs is derived from the MNE surveys, for which industry assignments were available for all companies.
How were the BEA surveys of international trade in services linked to the surveys of activities of MNEs?
BEA conducts several surveys to collect information on international trade in services and on the activities of MNEs. The data collection system for each survey is distinct. Therefore, data on trade in services may only be combined with the richer detail available on the MNE surveys via a bridge, which links reporters from one survey to reporters on another survey. In 2012, BEA undertook the first-ever project to link data from its three major services surveys—BE–120/125 Survey of Transactions in Selected Services and Intellectual Property with Foreign Persons, BE–140/45 Survey of Insurance Transactions by U.S. Insurance Companies with Foreign Persons, and BE–180/185 Survey of Financial Services Transactions Between U.S. Financial Services Providers and Foreign Persons—to data from the surveys of the activities of MNEs, including both the operations of U.S. parent companies and foreign-owned U.S. affiliates, for 2006–2009.4 In 2019, BEA updated the link to extend to 2017 and made several improvements to the quality and coverage of the link for earlier years dating back to 2008.
The initial 2012 linking project used employer identification numbers (EINs), which are collected on most BEA surveys, as well as standardized names and manual review of large reporters to construct the set of links that compose the bridge. The 2019 link update also utilizes EINs and names, but it introduces more advanced text mining techniques for both the establishment and verification of name-based matches. In addition, the update utilized information on email addresses, mailing addresses, and location to assist in and automate the process of establishing and verifying matches. Careful manual review was still a part of the linking process, but the use of automated methods reduced the time required to validate a set of matches and to reject false matches. The effects of these improvements resulted in an even larger share of trade in services being attributed to MNEs than in the initial 2012 project (see "How does the scope of BEA's current statistics on the characteristics of services traders compare to that of past statistics?", below, for more details).
How does the scope of BEA's current statistics on the characteristics of services traders compare to that of past statistics?
BEA's current statistics on the characteristics of international services traders, first released in December 2019, provide an updated view of services traders for 2008 through 2017. The last presentation of this kind was published in 2012 and covered data for 2008.5 The value of trade in services for 2008 presented in the current statistics is different than that presented in the 2012 article because of revisions to BEA's trade in services data and minor changes to the service types considered to be in scope for the analysis. Specifically, exports of the selected services in scope for the current statistics for 2008 totaled $294.4 billion and imports totaled $197.9 billion, compared with in-scope exports of $302.3 billion and imports of $195.6 billion published in the 2012 article.
In addition to revisions to the underlying data and changes to the services considered in scope for the current statistics, various improvements have been introduced to BEA's methods for linking reporters between surveys. Improvements to BEA's survey linking methods led to a larger share of trade in services being attributed to MNEs than in the previous profile. Based on the updated linked data, MNEs accounted for 87 percent of exports and 87 percent of imports of selected services in scope for the current statistics in 2008, as compared with 82 percent of services exports and 83 percent of services imports as reported in the 2012 article.
1 In August 2019, BEA published, for the first time, statistics featuring U.S.-headquartered parents that are ultimately owned or controlled by a U.S. resident. These statistics exclude information on U.S. parents that are foreign owned.
2 BEA collects this information, as it does for imports of computer services from all other countries, on one of its quarterly services surveys, but its published values of computer services imports from Canada are based on the quarterly value provided by Statistics Canada because BEA views Statistics Canada data as having more complete coverage.
3 The fully consolidated domestic U.S. enterprise excludes foreign branches and other foreign affiliates.
4 The selected services covered by the BE–120 and BE–125 surveys differ from the selected services included in the characteristics of international services traders statistics. For more information on BEA's surveys of international services transactions, see International Surveys: U.S. International Services Transactions. For more information on the surveys of the activities of MNEs, see International Surveys: Foreign Direct Investment in the United States and International Surveys: U.S. Direct Investment Abroad.
5 See Kevin Barefoot and Jennifer Koncz-Bruner, "A Profile of U.S. Exporters and Importers of Services," Survey of Current Business 92 (June 2012).