This paper describes and evaluates BEA data on U.S. trade and
investment in services, discusses recent efforts by BEA to improve its
data, and makes suggestions for further improvement.
Services are generally defined as economic outputs that are
intangible and invisible. In analyzing international services
transactions in this paper, this definition is interpreted broadly to
include the output of such diverse industries as petroleum services;
construction; transportation, communications, and public utilities;
wholesale and retail trade; banking; finance (except banking),
insurance, and real estate; agricultural services, hotels and loging
places; and a variety of business, technical, professional, and personal
The first section of the paper reviews the growing policy and
statistical interest in services transactions over the last decade and
BEA’s response to it. The most significant of the steps BEA
already has taken to improve its services data is a new annual survey of
U.S. direct investment abroad, which was approved in July 1984. Other
improvements, including some that would require new legal authority to
implement, are underway. The most significant is a benchmark survey of
services transactions between U.S. residents and unaffiliated foreign
residents. Such a survey would help fill the pressing need for
information on “other private services” (see below).
Following a summary of the paper in the second section, the third
section discusses definitional and methodological issues. The
discussion of definitional issues–for example, that goods and services may be sold jointly and not priced separately–indicates the difficulty
of isolating services activity. The discussion of methodological issues
provides the framework, used in the fourth section, for evaluating
existing BEA data. The issues relate to (1) the basis used for
classifying data, (2) the detail in which data are, or can be,
disaggregated by country, industry, or type of transaction, (3) the
coverage of transactions by the existing reporting system, and (4) the
basis used for recording transactions.
The fourth section transactions, evaluates existing BEA data on
international services transactions. BEA data are from two sources. The
first source, the international transactions (balance of payments)
accounts, covers transactions, including services transactions, between
U.S. and foreign residents, both affiliated–that is, between U.S.
parent companies and their foreign affiliates and between foreign parent
companies and their U.S. affiliates–and unaffiliated. The balance of
payments data cover travel, passenger fares, other transportation,
royalties and fees from and to affiliated foreigners, royalties and fees
from and to unaffiliated foreigners, other private services (such as
foreign contract operations of U.S. companies, international
communications, reinsurance, and film rentals), direct investment
income, and other private receipts and payments (such as interest and
dividends on portfolio investment). The second source, direct
investment financial and operating data, covers various aspects,
including services-related aspects, of the financing and operations of
U.S. parent companies, their foreign affiliates, and U.S. affiliates of
foreign companies. Instances where no coverage exists–such as
transactions between unaffiliated U.S. and foreign persons for
advertising, accounting, legal, and medical services–are noted.
The fifth section discusses methods of combining data from the two
sources to obtain a measure of total sales between U.S. and foreign
residents, whether made through trade or investment channels. It
presents a tabulating scheme, using engineering services as an example,
that shows how the necessary information can be derived by utilizing,
and building upon, BEA’s system of international surveys.
The last section presents illustrative data on services from the
balance of payments accounts and the 1977 benchmark survey of U.S.
direct investment abroad. The latter data are using to address two
analytic questions: (1) whether data classified by industry of
enterprise are adequate proxies for data classified by industry of
activity, and (2) to what extent, and in which industries, data on sales
can substitute adquately for data on value added.
References to related studies and papers are provided.