Economic Stabilization in the Presence of Limited Discretion

Abstract

Several authors, including Chow [4; 5] Cooper and Fischer [7] and Garbade [9; 10], have investigated the dynamic behavior of stochastic macroeconomic models when policy in each period is set as a function of observed economic activity. The results of those investigations indicate that fluctuations in aggregate activity can be reduced if policy-makers apply discretionary policies rather than fixedly following some pre-selected pattern of instrument choices. None of the studies, however, fully considered the relative importance of limiting the discretionary variation of specific instruments. Garbade treats only the two polar cases of discretionary versus non-discretionary selection of all instruments. Cooper and Fischer, using the St. Louis model [2] with its two instruments of money supply and high employment federal government expenditures, do treat the case where one instrument varies in response to the evolution of economic activity while the other follows a fixed sequence. Those authors, however, limit their choice of fixed policy sequences to either a steady rate of change in the instrument or to the observed historic choice of policies over the interval of simulation. They also limit their discretionary policy to the class of proportional and/or derivative stabilizers [15], where the instuments respond to changes in