The fact that economic conditions influence voters is a leading commonplace of conversation in election years. The question is: Is this fact in fact a fact? Despite the overwhelming popularity of the fact, it has received neither an explicit theoretical analysis, presumably because it is so obvious that economic adversity should create political adversaries, nor until recently a satisfactory statistical analysis.' Gerald H. Kramer has presented a multivariate analysis of congressional elections which would strongly suggest that fluctuations in the rate of unemployment have no appreciable effect upon elections, but that fluctuations in per capita real income are influential. In the following pages I propose to (1) reaffirm his finding on the electoral unimportance of ordinary fluctuations in unemployment, (2) argue that, contrary to Kramer, fluctuations in real income also do not have important electoral effects, and (3) present an argument based upon rational voter behavior for the unimportance of general economic conditions in national elections.

Harvey Leibenstein called attention in an influential article (1966) to a source of economic inefficiency which was given the awful name of X[in]efficiency. He cited studies in which misallocations of resources due to monopoly or tariffs had trifling social costs, whereas simple failure to attain the production frontier apparently led to social losses of a vastly greater magnitude. I propose to argue that this type of inefficiency can usefully be assimilated into the traditional theory of allocative inefficiency. It is a question (to be discussed below) whether one ascribes failures to reach the ultimate limits of output from given inputs in any state of technology to inadequacy of knowledge alone, or adds also inadequate motivation. Leibenstein (1966) separates the two:

The most fundamental proposition in the price literature is that in industrial markets, especially those which are oligopolistic in structure, are unresponsive to changes in general business conditions, and that this behavior is pervasive. Our study, The Behazior of Industrial Prices, is an examination of the evidence regarding this view. Our central purpose was to ascertain whether administered prices which, as reported by the Bureau of Labor Statistics (BLS), were essentially unresponsive to declines of general business conditions, were in fact unresponsive: that is, our central task was to obtain transaction rather than quoted prices. To that end, we collected buyers' on commodities that met two conditions: