Introduction

Sargent's Keynesian textbook model of economic growth, which we studied in chapter 2, was basically a supply-side model. Accordingly, firms' production is determined by the solution of their short-run profit maximization problem, with no quantity constraints involved. In connection with the standard neoclassical production function, this amounts to the marginal productivity principle for labour. We have on several occasions expressed our uneasiness with the implications of this hypothesis. This unease is so profound that, as has already been announced, we now drop it completely. In doing this, three components of the model have to be conceived anew: (a) the production technology, (b) firms' output, and (c) firms' price setting. The rest is, essentially, taken over from the Sargent model, perhaps adjusted slightly to fit the modified context, or extended slightly.

The usual alternative to the neoclassical production function is a fixed-coefficients technology. In point (a) we accept this supposition as far as long-run equilibrium growth is concerned, but allow for variations in capacity utilization as well as for procyclical fluctuations in (detrended) labour productivity. Regarding (b), we employ the concept that the output of firms satisfies current demand. In familiar parlance, the model may thus be termed a Keynesian demand-side model. More formally, the short-run time horizon of goods and financial markets is specified in IS-LM fashion, where, in contrast to the Sargent model, the price level is predetermined. Regarding the third component (c), prices remain invariant within the (infinitesimally) short period and are changed between these periods.