CROSS-SUBSIDIZING: A COMMENT

Abstract

XV. G. SHEPHERD'S article, 'Cross-subsidizing and Allocation in Public Firms',' evokes both admiration and exasperation. On the one hand, Shepherd is to be congratulated for insisting that the debate about the choice of allocative rules in nationalized industries should by now have been reduced to a series of empirical questions with numerical answers; and for the contribution he has made to our knowledge of the actual performance of the National Coal Board. On the other hand, in his zeal to refute earlier criticisms of the 'correctness' of the N.C.B.'s policy, Shepherd appears to slip into some conceptual errors, and to reach some conclusions not supported by his evidence. 1. Shepherd's major theoretical answer to those who have alleged that cross-subsidizing has occurred is to deny that the critics are using an appropriate criterion of efficient allocation: 'In several types of situations, cross-section patterns of profits and losses within public firms may be wholly consistent with efficient allocation.'2 Where such a pattern appears, applying private-market rules (i.e. attempting to maximize profits) would thus not necessarily reduce misallocation. This appears to be an impregnable theoretical argument. But if 'the market criterion' is an inappropriate one to use in cases where there are disequilibria or divergences of private from social costs, why does Shepherd go ahead to use it in his own empirical work ?3 Shepherd might simply be making the valid debating point that, although he does not accept the conceptual appropriateness of the profitability criterion, many of the critics of the nationalized industries have used the criterion, and so it is helpful to his case if he can show empirically that the critics are wrong even on their own conceptually inappropriate criterion. If that is the explanation of the structure of his argument, however, Shepherd must now go on to tell us what usable criteria for allocative efficiency he is himself adopting. At the moment he has left us no way of testing empirically such propositions as: 'The N.C.B.'s investment policy is efficient.' For Shepherd is apparently now inclined to explain any failure