Procyclical Productivity and Macroeconomic Policy: A Structural Economic Approach

Recent research on economic fluctuations has focused on the persistent empirical fact of the procyclical productivity of labor. This phenomenon is subject to a number of different explanations - a non-convex technology, labor hoarding, or technology shocks. In order to avoid a number of problems encountered in previous research, this study develops a structural econometric model capable of assessing systematically the contribution of these three hypotheses to procyclical productivity. In addressing these substantive issues, the paper presents three methodological innovations: information from the transversality conditions are incorporated in a tractable manner, new restrictions from the firm's optimization problem are exploited, and asset prices (Tobin's Q) are integrated into an economic framework in which there are non-constant returns and non-competitive markets. The analysis is particularly sensitive to the possibly distorting effects of technology shocks, which are assessed using a recently developed econometric test. No evidence is uncovered to suggest that labor hoarding or technology shocks play important roles. The estimated cost elasticity is nonetheless robust to these shocks, and is found to be less than unity. The evidence points strongly to the conclusion that procyclical productivity is due almost entirely to the shape of the technology. Marginal cost is below average cost, and falls during expansions. This finding has important consequences for the nature of macroeconomic equilibrium (equilibria) and the role of aggregate policies designed to attenuate economic fluctuations.