Political and economic determinants of county government pay

Abstract

With rare exceptions the literature on wage determination in the public sector has focused upon the impact of collective bargaining. (Shapiro, 1978; Freund, 1974; Ashenfelter, 1971; Ehrenberg, 1972). On the other hand, there is a more limited discussion of government wage determination which implicitly assumes the existence of a social welfare function which generates labor demand curves from equating social marginal rates of substitution to relative factor price ratios (Ehrenberg, 1972; Ashenfelter and Ehrenberg, 1975; Smith, 1976.) However, all of this liberature neglects the public choice theory and empirical evidence which, in extremis, denies the existence of a well-behaved social objective function. Unfortunately, the major public choice work on this subject by Niskanen (1971: 45) assumes that a bureau is a 'competitive purchaser of factors.' The conclusions of this assumption appear to contradict the facts of relative government pay as adduced by Smith (1976; 1977), Lentz (1976a) and Perloff (1974). Similarly, the works of political scientists on public sector employment in the non-unionized context, Sorauf (1956), Tucker (1969) and Grump (1971), focus upon the political participation outcomes of public employment without any rigorous treatment of the economic incentives used to induce these outcomes. In this article, I simply wish to argue that we need a model of public sector wage determination which (1) predicts the wage outcomes in the absence of unionization, (2) takes explicit account of the fact public employment labour demands are the outcome of a democratic political process, and (3) accounts for the major non-wage terms and conditions of public employment. Reder (1975) has advanced this sort of model as have I in Lentz (1976a). In this paper I will merely summarize my earlier work and test the model put forward using data for county government employees. In the analysis proposed here and in Lentz (1976a), the fundamental assumption is that public employers seek to maximize their long-run expected, i.e. loyal, voter support rather than the output of traditionally-defined