Occupational Segregation and the Gender Wage Gap Revisited:

This paper reexamines the relationship between occupational sex segregation and the gender gap in earnings. While there is no dispute that predominately female occupations pay less than male occupations, there is considerable disagreement as to whether this wage gap actually reflects gender discrimination in the labor market or productivity differences among workers. Several recent papers have used longitudinal data of workers who change occupations in an attempt to arrive at a more precise estimate of the effect of gender segregation on wages. While these studies find statistically significant effects, the magnitude of the effect is small, suggesting that occupational segregation is of relatively minor importance in determining the male/female wage gap. This paper argues that measurement error in the coding of respondents’ occupations significantly reduces the size of the observed effect. There is evidence that 85% of occupation changes in survey data are spurious. Simulation results suggest this may attenuate the observed effect of occupational segregation on wages by as much as 80%. Finally, I use multiple reports of occupations from successive months of data from the 1986 to 1992 Current Population Survey to suggest a lower bound of the true effect of occupational segregation on the wage gap.