Effective Use of Econometric Models in Macroeconomic Policy Formulation

At the beginning of each year, the Economic Report of the President of the United States makes projections of GNP in nominal and real terms for the coming year, the unemployment rate and the inflation rate, and states the major fiscal and monetary policies required to achieve these target rates. For example, the Report of January 1976 estimates real GNP to be 6 to 6.5 per cent higher in 1976 than in 1975 (p. 19), the unemployment rate to fall by almost a full percentage point and the inflation rate measured by the rise in the GNP deflator to be about 6 per cent (p. 24). The associated fiscal policies include a proposed Federal outlay in fiscal 1977 of $394 billion, a cut in taxes beginning in July 1976 of about $28 billion relative to what they would be under the 1974 law (p. 22). The rate of growth in the money supply M₁, as announced by the Federal Reserve, ranges between 5 1/2 and 7 1/2 per cent, but the Report asserts that maintaining a rate of money growth at the upper limit of this range would hinder the progress toward lower inflation rates (pp. 21–2). Assuming that econometric models are being used for policy analysis, this paper presents a systematic approach to apply some recently developed techniques of stochastic control to improve the formulation of macroeconomic policies and the accompanying economic projections.