R. F. Kahn's recent review article on the literature of “full-cost pricing” brings out the fact that the controversy which began with the article by Hall and Hitch in 1939 remains unresolved. Because the actual pricing of manufactured products takes place within a very restricted range of plant capacity, and under dynamic conditions which limit the entrepreneur's horizon, some writers have argued that the pricing process is necessarily arbitrary and based on rule of thumb. Most theorists, on the other hand, maintain that, however restrictive the data from which business-men form their decisions, the pricing process may be interpreted by marginalist concepts of profit maximization.

The gap between these points of view has been substantially narrowed in recent years. Extended subdivision of the theory of imperfect competition into individual market situations has brought that theory close to the methodology of “industry studies,” at least of those studies which have included some attempt at generalization. There remain, however, several obstacles to closing the gap between the deductive and the inductive analysis of the price mechanism. In particular, the available data are usually so unrefined that appeals to empirical cost and revenue schedules are misleading, or inconclusive, or contradictory. Joint costs, indeterminate demand schedules, and the natural reluctance of producers to reveal their private information all hinder the theorist searching for empirical support. In the face of these formidable difficulties some economists (notably the exponents of full-cost pricing) have turned to the technique of the questionnaire. This procedure is open to a serious methodological criticism: information derived from questionnaires is more likely to reveal the entrepreneur's idealized conception of his own behaviour than to show what his behaviour actually is. The phrases “usually full cost,” “cost is the normal guide,” or “the initial guide” recur throughout the replies to the Hall-Hitch questionnaire. These qualified answers are unrevealing for the same reason that the Marshallian long run (when all the various forces have had time to work themselves out) is a poor guide to real-world price determination. Normative statements such as those obtained in the studies mentioned above cannot be regarded as empirical data.