Wages and the Cost of Search

Abstract

Ullman produces startling results in his article Interfirm Differences in the Cost of Search for Clerical Workers (Journal of Business, April 1968). But Ullman is not startled. Indeed he maintains that his results are precisely what would be predicted by economic theory. He is wrong. Ullman finds a negative relationship between a firm's cost of search and its wage. Ullman explains this result by wages and search as substitutes-that high-wage firms have to do less searching than do low-wage firms. This explanation is singularly incomplete. If firms were profit maximizing without constraints, we would expect just the opposite results from Ullman's. Search and wages as substitutes yield a positive relationship between the two! Only with constraints on profit maximization can a negative relationship between the cost of search and wages emerge. If this negative relationship is indeed observed, then these constraints dominate the relationship between search and the wage. This would be highly significant news about the nature of labor markets. The argument can be seen most clearly by constructing a simple model of labor market behavior. The firm distributes information about its jobs in its effort to obtain qualified applicants to whom the firm can make offers. We would expect cost of information to the firm to increase with increases in the number of offers it makes. We would also expect this cost of information to decrease with increases in the wage. The average quality of the applicants will increase so that less screening will be required as the wage is higher. A simple equation incorporates these observations.