Richard Stone and Measurement Criteria for National Accounts

John Richard Nicholas Stone was a prolific and innovative writer. His contribution to economics covered a wide range of areas including empirical analysis of consumer behavior, economic growth, economic demography, and education. His work also covered topics such as index numbers, time series and cross-section surveys, environmental and sociostatistics, models of financial markets, and optimization problems related to economic growth. Stone's contribution to these topics wasshaped by his concern with measurement, with integrating theory and facts, and abstract and practical reasoning and expressing results in a measurable form so that they could be used to solve concrete problems. In no other area did Stone's work achieve more recognition than in the field of national accounts (see Meade [1944–46] 1990, 62; Deane 1997, 389; Harcourt 1995, 153).

The origins of national accounting can be traced back to William Petty's early estimates of capital, income, and expenditure for seventeenth-century England and those of Pierre de Boisguillebert in France. Studenski (1958) provides a comprehensive discussion of the evolution of the national accounts that shows how the history of national accounts could be classified into different periods: (1) early estimates (1571–1888); (2) the period 1900–1917; and (3) the period 1918–39, which Studenski calls the period of “the extraordinary flourishing of estimates” (1958, 149). It is within this last period that we can find the origins of the modern national accounting and of Stone's work.

The history of modern national accounting started with A. L. Bowley's, Colin Clark's, and Simon Kuznets's estimations of the principal macroeconomic accounting values for the United Kingdom and the United States, respectively, for some years in the 1920s and 1930s. Clark and Kuznets have been described as coauthors of ‘the statistical revolution' that followed the revolution in macroeconomic theory of the 1930s (see Patinkin 1976, 1104; Arndt 1988, 2). Production of national statistics boomed during the 1930s in countries such as Hungary, Germany, Sweden, Canada, the Netherlands, the United States, and the United Kingdom (see Carson 1975 and Kendrick 1970). During the 1940s, when new institutions were created with the aim of calculating national statistics, many contributors, such as Milton Gilbert, Morris Copeland, George Luxton, E. F. Denison, O. Aukrust, and J. Tinbergen, to name just a few, helped the development of national accounting. Some of this work was aimed at the collection of national level data rather than national income data. Thus it was not uncommon to find a variety of diverse isolated statistics—such as those relating to production levels of food or minerals—being used to characterize the evolution of national income. Indeed, the beginnings of this ‘statistical revolution' seem to be characterized by statistics being produced without much consistency.

In the history of the development of the national accounts, relatively little attention has been paid to the issue of what criteria should be used to assess the quality of measurement. Stone's contribution to the measurement of national accounts is particularly important in this respect and can be examined by looking at the different criteria to assess the measurement of these accounts that he developed in his work. The criteria suggested by Stone were influenced by a set of practical considerations, and this essay is about Stone's attempts to develop these practical criteria for national income accounts.

It seems that most of the criteria Stone developed over a period of about forty years were all originally conceived in his early work. In fact, the main criteria of logical consistency, flexibility, invariance, and standardized forms (as well as his error adjustment methods) were enumerated by Stone as early as in the beginning of the 1940s. Stone appeared to have thought of these criteria within a dynamic perspective according to which some criteria would have to be achieved first in order to provide a basis for subsequent extension of the accounts; in the...