DYNAMIC ADJUSTMENTS UNDER BOND FINANCE: AN EXTENSION OF THE BLINDER AND SOLOW MODEL

Abstract

In their well-known analysis of fiscal stabilization policies [1], Blinder and Solow clarify the meaning of the government budget constraint. However, the transitional, dynamic process by which government expenditures are transmitted is not explicitly considered. The purpose of this paper is to explore implications of the adjustment process of income in the analysis of a bond-financed increase in government expenditures.
In the IS-LM framework, income is instantaneously determined at the intersection of the IS and the LM curves, while stock variables such as bond balances, money balances, and physical capital are assumed to be constant. As the result of public expenditures, income increases instantaneously. If the impact government spending multiplier, 1/(1-marginal propensity to spend) is sufficiently large, the instantaneous increase in tax revenues will create the government budget surplus at the initial moment. The government budget constraint implies that the surplus is turned to the redemption of bonds. Since we are concerned with the effect of bond-financed budget deficits on income, the case in which government spending leads to a budget surplus at the initial moment is neither interesting nor relevant. Blinder and Solow assume away this troublesome case by omitting income from the investment function in order to lower the marginal propensity to spend. we shall present a more generalized framework and consider the case where induced investment is allowed.
We shall explicitly introduce an adjustment process in the commodity market, so that fiscal expenditures increase tax revenues only with delay. In this way we can avoid the complications noted above, and at the same time explicitly consider the time development of the system. Contrary to the conclusion by Blinder and Solow that the additional changes in bonds have a positive net effect on income if the system is stable, we shall show in our generalized framework that the subsequent effects may be contractionary even if the system is stable. Thus, Blinder and Solow's conclusion holds on narrower grounds than they seem to claim; that is, only when the parameter of policy is government purchases of goods and services and the marginal propensity to spend is sufficiently weak.
As regards the steady-state multiplier considered, it will be shown that the positive wealth effect on consumption is a necessary condition for the effectiveness of fiscal policy. The question as to whether or not the wealth effect is positive is not at all out of place here. However, the long-standing Keynesian-monetarist controversy concerning whether or not the impact effect of the increase in bonds on effective demand, E_B, is positive has no bearing on the analysis of the steadystate multiplier, because the sign of E_B is not relevant to the long-run effectiveness of fiscal policy. In Model II (the parameter of policy is government expenditure including interest payments), even if E_B is negative, it is still possible for the system to be stable and for fiscal policy to be effective.
As regards the dynamic adjustments under bond finance considered, we shall show in our consideration of the adjustment process that fiscal policy is normally effective during the initial phase where the government budget remains at a deficit, as long as the impact effect of the increase in bonds is positive. Fiscal policy in depression mostly corresponds to the situation where the government budget is in deficits. Therefore, in most of the relevant cases to be encountered, fiscal policy can be deemed to be effective in spite of some less favorable implications to be found in the paper.