Peltzman on regulation and politics: Reply

Abstract

Few useful economic theories can, on their logic alone, rule out conflicting outcomes. Both upward and downward sloping demand curves are compatible with the logic of constrained utility maximization; by the same logic a rise in the price of corn can either decrease or increase the price of wheat, etc. The essential thrust of Goldberg's critique is that he expects more of my theory. This is flattering, but not very helpful. Goldberg's prime example centers about a utility commission which, for illustrative purposes, I had trading wealth between producers and consumers of electricity. In that scheme, a tendency (not an unambiguous result) emerges for increases in potential profits to be translated into lower electricity prices. In Goldberg's proposed variant, the commission trades wealth between electric and gas producers and the same increase in potential electricity profits produces fiigher electricity prices. Goldberg is disturbed that there is no a priori basis for choosing between the conflicting possibilities. But, just as one would not discard consumer choice theory for its failure to predict that wheat and corn are substitutes or complements, a skillful analyst would not be paralyzed by Goldberg's lament. He or she might notice, for example, that some real world commissions regulate electricity rates others regulate both electric and gas. These institutional facts could then suggest a working hypothesis e.g., that the tendency toward higher prices adumbrated in Goldberg's Figure 2 is likely to be more important where the commission regulates both gas and electric rates. Then a fairly clear, testable proposition emerges: If costs fall or demand increases, the resulting price change is algebraically smaller where the commission regulates only electricity rates. There is nothing very novel about any of this; it is the ordinary road we travel from the inherent ambiguities of general models to their specific, falsifiable predictions. Goldberg is also bothered that the theory cannot define the relevant constraints. Cannot subsidies be granted or taxes imposed to alter the regulator's opportunity set? If so, isn't the constraint politically endogenous? The answers are, of course, 'yes' and 'yes.' But, to the more relevant question, 'How important is all this?' The only sensible answer has to be, 'It depends on