Wages, non-wage characteristics, and predominantly male jobs

This paper estimates the wage premium associated with working in predominantly male jobs. It also examines whether this wage premium is greater than the compensation workers demand for the less desirable non-wage characteristics of such jobs. The coefficients of the change in the proportion of men in an occupation on the change in wages for quits and layoffs provide opposing biased estimates of the wage premium; because workers who voluntarily quit move to better matches, but those that are laid off accept jobs from the representative distribution of job offers. Specifically, when the premium paid over- (under-)compensates for undesirable work characteristics, the quit estimate is a downward (upward) biased estimate of the wage premium, while the layoff estimate is biased upward (downward). Results from the U.S. National Longitudinal Survey of Youth (NLSY) and the U.S. Panel Study of Income Dynamics (PSID) indicate that: (1) the estimated bounds of the wage premium are large; and (2) the wage premium overcompensates for the non-wage characteristics of male jobs.