RDP 8809: The Intertemporal Government Budget Constraint and Tests for Bubbles 1. Introduction

Much attention has recently been focussed upon the appropriate stance of fiscal policy. Many macroeconomic analysts and policy makers, both in Australia and overseas, have argued that protracted fiscal deficits do not aid economic performance and might adversely affect future real economic activity by crowding out private investment expenditures and generating inflation. In addition, concern has recently been expressed about the sustainability of trends in the growth of public debt and this has led some governments to embark upon programmes of fiscal restraint in order to reduce their levels of outstanding debt. As Blanchard, Dornbusch and Buiter (1986) point out, an interesting question in this regard concerns the optimal path of debt stabilisation because it is far from clear that the optimal policy involves the fastest possible adjustment.

The purpose of this paper is to examine the appropriateness of fiscal stance in relation to the government's intertemporal budget constraint. The latter relates the sustainable growth of government debt to the non-interest fiscal deficit, the growth of nominal output and the service costs of outstanding debt. Specifically, the government's intertemporal budget constraint provides the sustainable limits to the growth of government debt given the performance of the macroeconomy. Although this constraint must hold in the long run, there is ample scope for governments to exceed it over short periods of time by engaging in so-called bubble financing. Tests for the existence of bubble financing have been reported by Hamilton and Flavin (1986), Hakkio and Rush (1987) and Trehan and Walsh (1988) for the US and by MacDonald and Speight (1987) for the UK. On close examination, however, a number of problems exist with the method in some of these studies and this paper presents the corrected tests. We have developed the framework to allow for the effects of income growth on debt sustainability and applied the tests to Australian data over the period 1953/54 to 1986/87. Amongst the main findings are the lack of evidence of bubble financing in Australia together with an indication of the historical importance of monetisation for the financing of non-interest fiscal deficits.

Section 2 presents the framework of the government's intertemporal budget constraint and explains how it constrains the sustainable growth of debt. Section 3 generalises the tests of bubble financing utilising the cointegration methodology of Engle and Granger (1987) and the approach of Hamilton and Flavin (1986) to allow for income growth. The tests are applied to the data in Section 4. The final section summarises the paper and draws conclusions.