Predatory pricing: When bright line cost-based tests are possible

In an earlier article I argued that the familiar Areeda-Turner (AT) test for predation may not always be unambiguously interpretable.  Complications  arise when the firm's desideratum is the maximization of intertemporal, rather than single-period, profit.  Of particular interest is the fact that, in the presence of learning-by-doing and demonstration effects, intertemporal profit maximization may require that price be set below average variable cost, the accepted proxy for marginal cost.  Add to this the complications that  inhere in the multi-product case, and the ambiguity of  cost-based tests is enhanced.

The purpose of this paper is to show that, under the right conditions, cost-based tests retain their basic relevance. In particular, the ambiguity of test results is significantly reduced in the case of single product firms that engage in cost-plus pricing. That said, it remains true that the question of proper cost taxonomy must be faced.  Among other things, the temptation to invoke the "semi-fixed" cost convention is particularly problematic.  Finally, in the case of a vertically integrated firm, determination of the appropriate transfer price(s) is a matter of signal importance.