these points is discussed below. Moreover, insofar as is feasible, the panel advocates adaptation of the double-entry structure of the core accounts for use in any supplemental accounts that may be developed. One implication of this last position is that every effort should be made to value outputs independently of inputs. The panel recognizes, however, that implementing such a structure for measuring nonmarket outputs will pose significant challenges, and that methodological compromises may be required during intermediate steps of the accounts’ development.

INTEGRATING CORE AND SATELLITE ACCOUNTS—OR NOT

Ideally, an experimental set of expanded accounts would integrate market and nonmarket components. From this perspective, the System of National Accounts (SNA) may offer some lessons.2 Though it does not endorse nonmarket accounts, a major feature of SNA is that it is a multi-element set of accounts linking production income, consumption, accumulation, and wealth.3 Both the SNA and the National Income and Product Accounts (NIPA) allow for the possibility of satellite accounts; for example, in the early 1990s, BEA developed the Integrated Environmental and Economic Satellite Accounts which could be integrated with the NIPA accounts.4 This integrated feature has become a stated goal of the U.S. accounts and would, if feasible, also be a desirable element in augmented nonmarket accounts (Nordhaus, 2002, p. 4).

A less ambitious, simpler alternative—and perhaps necessary first step—might be to construct free-standing accounts for individual nonmarket components. Under this alternative, the core accounts would remain much as they are, and the satellite accounts would provide alternative perspectives on selected areas of activity—such as household production, human capital, and health. The shortcomings of this “interim solution” are obvious: Given the overlap of nonmarket areas, with each other and with market areas already captured in the national accounts, a system that fails to integrate core plus satellite accounts would produce numbers that do not add up to a meaningful total and would be inconsistent with current practices at BEA—the virtues of which have already been enumerated.

One advantage of the nonintegrated approach, relative to a fully integrated one, is that it might allow experimental accounts to be produced much more quickly. However, for various nonmarket areas, experimental work might follow heterogeneous approaches—particularly with regard to valuation methods. The nonintegrated approach would also allow work in each area to proceed at its own pace as data and concepts evolve, without seriously disturbing the existing accounts. This seems to be the thinking behind the new satellite household production account in the United Kingdom, for instance.5

2  

SNA guidelines were prepared by the Inter-Secretariat Working Group on National Accounts, a joint effort by the International Monetary Fund, the European Union, the Organization for Economic Co-operation and Development, the United Nations, and the World Bank. The effort was mandated by the Statistical Commission of the United Nations to oversee international coordination in the development of national accounts

3  

Another research project under way at BEA would create wealth accounts by integrating the Flow of Fund Accounts with the National Income and Product Accounts.

4  

Congress directed BEA to suspend this work in 1994.

5  

For a description of this account, see http://www.statistics.gov.uk/statbase/Product.asp?vlnk=9338 [4 April 2002].



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement