Essays on wage differentials

Date of Award


Degree Type


Degree Name

Doctor of Philosophy (PhD)




J. David Richardson


Labor market, International trade, Wage differentials

Subject Categories

Economics | Labor Economics | Social and Behavioral Sciences


Why do wages of observationally similar workers vary with their industry of employment? In the literature, this question has been explored along several lines, including a variety of "efficiency wage models" and models with compensating differentials. These theories, which are not mutually exclusive and are often overlapping, account for about half of the overall residual variation of wages in the U.S. economy. My dissertation attempts to shed some light on the unexplained half.

I suggest an alternative approach to explaining industry wage differentials--one based on industry specific human capital. Recognizing the empirical difficulty of identifying specific factors, I develop two cases: (i) where a researcher has perfect knowledge of whether an observation represents a specific or a shared factor, and (ii) where such distinction is not possible and only the industry of this individual's employment is observed. I show that in case (ii), the estimated industry wage differentials are contaminated by the inclusion of the shared factor whose price is equal across industries.

I find empirical evidence of factor specificity in unequal returns to education across industries and occupations, in the expected relationship between labor of differing mobility and industry wage differentials, and in inter-temporal pattern of wage differentials at several levels of aggregation. The assumption of perfect labor mobility is not supported. I argue that industry marginal effects as a measure of industry wage differentials are preferred to traditional differentials in the specific-factors model. I support this argument by comparing the performance of industry marginal effects and traditional differentials as dependent variables in several regression specifications, which themselves contain a number of novel explanatory variables.

The effect of trade on wage differentials is found statistically significant, but its magnitude is small compared to other explanatory factors, such as technological change and industry location. The relationship between industry wage differentials and industry location in a high-wage state is positive, but does not determine the direction of causality: geographical wages dispersion might stem from differences in industrial composition, rather than cause industry wage inequality. In the last part of my dissertation, I explore this issue and find that state wage differentials are indeed influenced by state industry mix.


Surface provides description only. Full text is available to ProQuest subscribers. Ask your Librarian for assistance.