This paper assesses evidence on imperfections in employer information about new workers in the labor market, and their role in generating wage differentials by race and sex. The methods used borrow heavily from research by Foster and Rosenzweig (1993, hereafter FR) to study statistical versus taste discrimination in developing countries, applying their methods to data from the U.S. that might be informative with respect to some of the same questions they consider. In addition, the paper presents some innovations relative to their methods. The analysis proceeds in three steps. First, using FR’s methods directly, evidence is presented on whether wage differences between apparently equal ly-productive white and minority workers are better characterized as reflecting taste discrimination or statistical discrimination.
As noted above, even in the simple statistical discrimination model, minority workers who are as productive as white workers will receive lower wages than comparable white workers if average productivity of minority workers is lower-where “comparability” is determined based on actual productivity, rather than expected productivity which determines starting wages in the statistical discrimination model. Thus, evidence that looks like taste discrimination may stem solely from statistical discrimination. Second, extending FR’s methods, the empirical analysis attempts to distinguish between imperfect information on the part of employers and measurement error in the productivity proxies available to econometricians, which have identical empirical implications for the test of taste discrimination versus statistical discrimination, but quite different implications for modeling wage setting and for policy. Finally, evidence is presented on whether employers have better information about some groups of workers than others. As suggested by the Rothschild-Stiglitz model, worse information about a particular group could lead not only to the usual statistical discrimination result, but also to group discrimination, so that average wage differentials exceed average productivity differentials. advance payday loans

The policy implications of this analysis are potentially quite important. If taste discrimination accounts for the unexplained lower wages of women and minorities, then anti-discrimination legislation may be the only appropriate response. On the other hand, if statistical discrimination is important, especially in conjunction with some other factor that leads to group discrimination, then better means of assessing workers’ productivity—including apprenticeships, skill certification, job testing, etc.—may contribute to the reduction of discrimination at the individual or group level.