The Case against Conservative Macroeconomics: An Inaugural Lecture

Keynesian economics has now reached middle age. With these advancing years the claims of her traditional rival, the quantity theory, have shown their endurance, while a young offspring of the old classical theory, called the New Classical Economics, threatens to take the centre of attention from both of the older theories. This lecture is a defence of Keynesian economics. It proceeds by the admittedly unchivalrous method of pointing out the faults of the two rivals. The issue at hand is what it has always been between Keynesian and conservative macroeconomics, whether the government is able or unable to control real income by macroeconomic interventions. To organize our thoughts it is convenient to choose one exemplar of conservative macroeconomics. For that purpose I have chosen Sargent's model of the economy (Sargent, 1973).1 I do wish to emphasize that the Sargent model is chosen because it is a typical representative of the New Classical Economics. Therefore the remarks here concerning the Sargent model in particular apply with equal force to the New Classical Economics in general. The Sargent model has three equations. The first is an IS curve. Keynesian economists and conservative economists agree, it appears, on the IS curve, at least to a first degree of approximation. The second equation is an LM curve. New Classical economists allow this LM curve to have a non-zero interest elasticity; the old-style conservative economists, who espouse the quantity theory, question the empirical significance of such an interest elasticity. With a significant income elasticity and a zero interest elasticity of the demand for money, fiscal policy can have no effect on money income regardless of the degree of slack in the economy. If not only the interest elasticity, but also the income elasticity, of the demand for money is low, however, fiscal policy can be effective in changing real income. The fault of the quantity theorists, I will argue in this lecture, is to dismiss the very reasonable possibility that fiscal policy works because the short-run income elasticity as well as the short-run interest elasticity is low. The first part of this lecture will discuss the theoretical and empirical merits of a short-run money demand function in which the income elasticity of demand is low. The third equation of the Sargent model is the aggregate supply schedule. According to this equation the supply of goods (and implicitly of labour too) depends upon the gap between actual prices and expected prices. Actual prices (and wages) are determined in such a way that goods and labour markets clear. The greater is the excess of actual over expected prices, the greater will be aggregate supply, since unexpected price changes are mistakenly interpreted as relative price changes. Unemployment in excess of the natural rate can persist only so long as the time it takes market participants to formulate their expectations correctly. The second part of this lecture takes issue with this view on empirical grounds; for it is shown that most unemployment belongs to spells of