Is Non-Monetarist Arithmetic Just as Unpleasant?

Abstract

What should the central bank do if the fiscal authority persists in running a deficit? If the monetarist policy prescription of bond financing deficits is adopted, McCallum [8] and others have shown that the real growth rate must exceed the after-tax real interest rate if macroeconomic instability is to be avoided. Analysts such as Miller and Sargent [10] and Burbidge [1] doubt whether this condition is met and thus argue in favor of greater flexibility in the rate of monetary growth. But will that lead to instability as well? Is the requirement for a unique solution and macroeconomic convergence just as stringent under the non-monetarist option of money-financed deficits? If so, monetarists can deflect criticism of their financing option from those who argue money financing is preferred on feasibility grounds. This paper addresses this question and concludes that money financing involves much more pleasant arithmetic than does bond financing. Analysis is given for both closed and small open economies. Considering what is available in the existing literature, the open economy extension may be the most important contribution of the paper. In addition to addressing the feasibility of alternative financing options, we use the model to comment on the welfare cost of inflation in an open economy environment. The feasibility of money-financed deficits has recently drawn renewed attention among macroeconomists. Erbas [2], using a model which imposes the Ricardian equivalence hypothesis and assumes no growth, concludes that money financing is preferred as an automatic stabilizer as it offers greater short-run stability of the level of income. Scarth [14] comes to the same conclusion using a model which does not impose the Ricardian equivalence hypothesis but which allows economic growth. Liviatan [6] and Mitchell [11] investigate the feasibility of bond and money financing using long-run models of a closed economy in which government bonds constitute a part of net wealth. In this paper we consider both open and closed economy long-run models, and we adopt the Ricardian equivalence hypothesis. We emphasize Ricardian equivalence since most monetarists are comfortable with this assumption. Thus unlike the previous long-run studies, we examine the feasibility of monetarist versus non-monetarist financing options using a model acceptable to most monetarists. This is important when the results do not support monetarism. The remainder of the paper is organized as follows. The closed economy analysis is given in section II, the open economy case is discussed in section III and concluding remarks are contained in section IV.