Working Paper: NBER ID: w7310
Authors: Mark Bils; James A. Kahn
Abstract: Manufacturers' finished goods inventories move less than shipments over the business cycle. We argue that this requires marginal cost to be more procyclical than is conventionally measured. We construct, for six manufacturing industries, alternative measures of marginal cost that attribute high-frequency productivity shocks to procyclical work effort, and find that they are much more successful in accounting for inventory behavior. The difference is attributable to cyclicality in the shadow price of labor, not to diminishing returns in fact, parametric evidence suggests that the short-run slope of marginal cost is close to zero for five of the six industries. Moreover, while our measures of marginal cost are procyclical relative to output price, they are too persistent for intertemporal substitution to be important. We conclude that countercyclical markups are chiefly responsible for the sluggish response of inventory stocks over the cycle.
Keywords: Inventory Behavior; Business Cycles; Marginal Cost
JEL Codes: E3
Edges that are evidenced by causal inference methods are in orange, and the rest are in light blue.
| Cause | Effect |
|---|---|
| procyclical behavior of inventory investment (E22) | misinterpretation as evidence of lower production costs during expansions (E23) |
| shadow value of inventories (G31) | procyclical behavior of inventory investment (E22) |
| anticipated sales (F17) | inventories (G31) |
| sales-stock ratio (C69) | procyclical behavior of inventory investment (E22) |
| sluggish adjustment of inventory stocks (D25) | high marginal costs relative to discounted future marginal costs or output prices during expansions (D25) |
| countercyclical markups (D43) | sluggish response of inventory stocks to changes in demand (E41) |