ON MEASURING FISCAL POLICY

Abstract

SINCE THE END of World War II, Federal fiscal policy has been regarded as one of the Government's primary tools for promoting high-level employment and economic growth. On a theoretical level, the economic impact of over-all fiscal policy, as well as of particular fiscal measures, has been explored under a wide variety of assumptions. The critical problem now seems to be less a matter of further refinements of theory than the more mundane testing of how the economy, and subsectors of the economy, do in fact behave in response to various fiscal policies. Despite extensive discussion during the postwar period, our practical knowledge of fiscal impact, except in the very broadest terms, is still quite primitive. Why should this be so? It is the thesis of this article that a good part of the answer can be found in the lack of appropriate empirical data suitable for examining fiscal impact, even though published budget data are more than plentiful. Most economic analyses using budget data rely either on the cash budget or on the Federal budget in the national income accounts. There has been a lively debate for some years about the relative merits of these two budget accounts, and no consensus has yet emerged-unless one can call a consensus the statement that both budget series are needed to give a complete picture of the Government's fiscal role.' Notwithstanding this consensus, empirical analyses generally do not use data from both budget series, and for good reason; there is almost a complete absence of linkages or bridges to move from one series to the other. Moreover, there are formidable problems in trying to construct bridges between statistics on Government operations and statistics on the operations of the private sector. This inevitably limits how much empirical research can discover with respect to how, when, and to what extent Government programs affect the economy or subsectors of the economy. If fiscal policy is to be developed into a more precise art, if not a science, then it is crucial to be able to pinpoint more accurately the timing and the extent of impact of fiscal measures. In order to suggest what kinds of additional data are needed for this purpose, it is first necessary to evaluate the major virtues and deficiencies of the national income and the cash budgets. Based on the conclusions of this survey, suggestions are offered for some additions to the currently available data.