PROBABLY the single most important statistic measuring the impact of government fiscal policy on the economy is the magnitude of the government surplus or deficit. Justifiably or not, this single figure has taken on such importance that the Congressional Budget and Impoundment Control Act of 1974 established a process whereby the Congress is forced to consider overall receipts and outlays and commit itself under a binding resolution to these totals. Because of the importance of the budget figure, economists have continually sought to modify the raw data to better assess the impact of the government on aggregate demand. A major breakthrough was the concept of the FullEmployment Surplus, which was popularized by the Council of Economic Advisers in their 1962 report, although its roots go back to World War II. I This concept recognized that due to the endogeneity of expenditures and particularly tax receipts, the actual budget deficit or surplus, uncorrected for the level of output, gives a biased indication of the stance of fiscal policy. Although the concept of the full employment surplus was an improvement, dissatisfaction with the single summary statistic remained. Several economists have experimented with different weights on the components of taxes and spending, recognizing the differential spending propensities out of the different sources of income.2 This approach has not been successful due to the difficulty of reaching a consensus on a weighting scheme. Another problem with the full employment surplus was the failure to adjust for changing price levels or rates of inflation. Fixed nominal income tax brackets, specific taxes, and taxes on nominal interest and capital gains all suggest that the impact of the budget should allow for the behavior of price level variables.3 Although standardization to an exogenous level of potential output seems plausible (given growth or frictional unemployment), there is no normal price level associated with any given level of output. Furthermore, with the breakdown of the Phillip's Curve relationship in recent years, there appears to be no rate of inflation associated with a full employment level of output. This paper attempts to demonstrate that price level changes are important for measuring fiscal impact apart from any attempt to determine what price level behavior would exist at full employment. Since the deficit is equivalent to the amount of government bonds sold to the public,4 analysis of the impact of the supply of bonds is incomplete without allowing for changes in the real value of the debt caused by inflation or deflation. Real value accrual accounting, which is making considerable headway in private sector accounts, can also be employed in the public sector.5 A redefinition of the deficit along these lines can be integrated easily with the full employment surplus or any variant definition to yield a better measure of fiscal impact. Section II of this paper presents a brief theoretical discussion of real value accrual accounting. Section III Received for publication April 4, 1977. Revision accepted for publication April 28, 1978. * University of Pennsylvania. I would like to thank Milton Friedman, Ben McCallum, and the members of the University of Chicago's and University of Virginia's Money and Banking Workshops for their helpful comments on earlier versions of this work. Steve Thompson provided valuable research assistance and computer programming. 1 Initial references to the concept were Ruml and Sonne (1944), Committee for Economic Development (1947), and Friedman (1948). For an excellent discussion of the concept, see Okun and Teeters (1970). 2 In particular see Gramlich (1966), Musgrave (1964), Okun and Teeters (1970), and Hymans and Wernette (1970). Warren Smith in Okun and Teeters (1970) gives yet another impact statistic. 3 See Committee for Economic Development (1947), Gramlich (1968) and Okun and Teeters (1970) for attempts at price level standardization. 4 For simplicity it is assumed throughout this paper that the money supply is held constant. One can alternatively regard the profits (seigniorage) from steady-state monetary expansion as tax revenue, so that monetary financing is considered only as an inflationary tax on real cash balances. 5 For an excellent summary of purchasing-power accrual accounting as applied to the corporate sector, see Shoven and Bulow (1975, 1976).