A Flow of Funds Analysis of Crowding out: Some International Evidence

Abstract

In a pioneering article in this Journal, Zahn [22] investigated the separate and combined influences of increase in government spending and government borrowing on the time path of income. The investigation was undertaken within the framework of a small quarterly model of the U.S. economy consisting of both commodity and financial (flow of funds) sectors. By explicitly using government borrowing, Zahn tested the crowding out hypothesis as one involving both government spending and borrowing effects on the flow of funds between commodity and financial markets. The combination of both the expansionary influences of government spending and contractionary influences of government borrowing on the time path of income determine the extent of real crowding out. Zahn distinguishes between pure and partial multipliers. The influence of an increase in government spending, financed strictly by an increase in government borrowing, is measured by the pure multipliers. They are based on the assumption that increases in endogenous net taxes as a result of an expansionary fiscal policy are held in a Treasury account and offset in the government budget constraint by equal decreases in the monetary base. On the other hand, the influence of an increase in government spending, financed in part by endogenous net taxes and the remainder by government borrowing, is measured by the partial multipliers. The major results in Zahn [22] may be summarized as follows. The pure multipliers show that the greatest influence of an increase in government spending, financed strictly by borrowing, is registered during the first two quarters but afterwards crowding out occurs. The partial multipliers show similar results, but crowding out is substantially less because borrowing requirements are reduced. Lastly, the cumulative multipliers indicate that a sustained increase in government spending, financed either by borrowing or by both endogenous net taxes and borrowing, increases income during the sample period, but at a decreasing rate, and after many years the influence falls to zero. In other words, crowding out occurs as a result of the negative sign of the government borrowing multipliers, but it is not large enough to outweigh the expansionary influence of government spending except in the very long run, when the steady state government spending multiplier approaches zero. Zahn's conclusions are, in principle, compatible with those found in most previous and