Testing for Wage Discrimination in US Manufacturing in 1832

Wage discrimination occurs when the relative wage of a particular group of workers does not equal their relative marginal productivity. While wage discrimination is not the only form of discimination, an understanding of labor market discrimination surely must include an understanding of when and where wage discrimination was present. Unfortunately, the most commonly used method of measuring wage discrimination is so flawed as to be largely uninformative. Studies of the wage gap usually use household survey data to estimate wage equations as a function of individual characteristics such as education, experience, and other observable characteristics. These wage equations allow the researcher to divide the wage gap into explained and unexplained components. The unexplained portion is upper bound on the portion of the wage gap due to wage discrimination.1 Unfortunately this measure is fairly uninformative because the unexplained portion of the wage gap may be the result of wage discrimination, or it may be the result of failure of the explanatory variables to capture important differences in productivity. Often the unexplained portion of the wage gap is fairly large, suggesting a wide range of possible levels of wage discimination, including no wage discrimination