INVENTORIES, PRODUCTION SMOOTHING, AND THE ACCELERATOR: SOME EMPIRICAL EVIDENCE.

The article presents information on inventories, production smoothing, and the accelerator. In this note a simple model of production smoothing that does not include "inventory smoothing" is presented and fitted to monthly data for six industries and four cross-section samples for one industry. Paul Darling and Michael Lovell argue that production smoothing does not represent an acceptable substitute to the flexible accelerator. Both hypotheses are consistent with maximizing behavior, and "the choice between the two alternative approaches must be resolved empirically." In choosing such a plan, the firm must balance the costs involved in changing the rate of output against those of inventory investment. If there are no costs of changing the level of output, the firm will produce in each period an amount equal to expected sales in that period. The results presented in this paper indicate that for the samples studied the behavior of inventories is inconsistent with that implied by the accelerator model and can be successfully explained by the production-smoothing hypothesis.