Macroeconomic Implications of Adjustment Costs: A Further Note

Abstract

Standard macroeconomics involves a basic inconsistency, since the production function for consumption goods is defined in two ways. In general, this function is G (N, K, I), where the inputs are labour, N, capital, K, and investment, I, respectively. The properties of this function that are always assumed are: GN, GK > 0 and GNN, GKK, GI < 0. The inconsistency stems from analysts making two different assumptions concerning GI,. When macroeconomists include the production function in their models they tend to write G (N, K, I) = F(N, K) -I, so that GI, = 0. However, when the production function is referred to as part of the micro-underpinning for the investment function, nonlinear adjustment costs are normally assumed, with GH < 0. In an earlier paper (Bailey and Scarth, 1980), we chose two particular forms of the production function to illustrate the potential significance of correcting this inconsistency. The comment on our paper by Baron, Black and Loewenstein (1983) (hereafter BBL) provides the opportunity for us to clarify some of the issues involved, and to stress that our demonstration of the significance of this logical inconsistency survives the extensions suggested by BBL.