Macroeconomic adjustment to external shocks : essays on the behaviour of individuals and markets

Abstract

This thesis consists of three distinct sections. The first two sections present theoretical models of the response of private agents to an unanticipated terms of trade shock and to changes in government net saving, respectively. The third section examines the joint behaviour of interest rates and the exchange rate. The link between the sections is that they were all motivated by the response of the Australian macro economy to the 1985/86 adverse terms of trade shock, and by the government's reaction to that shock. The first section of the thesis (Chapter Two) analyses the impact on the current account of a permanent terms of trade deterioration in a model in which an immortal intertemporally optimizing household is assumed to exhibit an explicit sensitivity to falling consumption levels. When the household is sufficiently sensitive, it is optimal for its consumption to be continuous over time and then the Harberger-Laursen-Metzler effect holds, i.e., following the trade shock, the current account deteriorates. It is argued that this model is more realistic than earlier models which predict a current account improvement in response to a permanent adverse trade shock. The second section of the thesis consists of two sub-sections which are both concerned with the Ricardian equivalence hypothesis. The first subsection (Chapter Three) reports the results of two surveys. Over six hundred economics students were asked to estimate the level of outstanding Australian Federal government debt, and eleven academic economists were asked to predict the proportion of students with 'a rough idea of the amount of this debt' (in a sense defined precisely). Student knowledge is very meagre and the average academic overestimates it fivefold. The relevance of these results for the Ricardian equivalence hypothesis is discussed. The second sub-section (Chapter Four) develops a model of an immortal, intertemporally optimizing consumer who is ignorant of the link between bonds and future taxes. The consumer is exposed either to the changing level of Federal government debt in Australia or in the U.S. over the twenty five years, 1963 - 1987. "Best" estimates of the cost of ignorance are about $A2 per annum for an Australian consumer or about $7 per annum for a U.S. consumer - i.e., less than 0.1% of annual income in both instances. When uncertainty about future income and the existence of progressive taxes are included in the model, these estimates are substantially reduced - even from this 0.1% level. It may therefore be optimal for consumers to ignore the link between bonds and future taxes. The final section of the thesis (Chapter Five) examines the large shortterm real interest differential between Australia and the US since late 1984. The chapter provides a detailed examination of an arbitrage condition for a representative US investor. There is some evidence for a risk premium until late 1985. The chapter also examines the possibilities that either the foreign exchange market is inefficient or that the market has continually and rationally expected real depreciation of the $A, but that such depreciation has not yet occured. Show more