Occupations, Human Capital and Skills

Economists have long recognized that occupations can be used as proxies for skills in wage regressions. Yet the potential
existence of non-market factors such as discrimination and occupational choice (sorting) on the basis of job attributes that
are separate from, but potentially correlated with, wages makes occupations an imperfect control for skills. In this paper,
we consider whether inter-occupational wage differentials that are unexplained by measured human capital are indeed due to
differences in unmeasured skill. Using the National Compensation Survey, a large, nationally-representative dataset on jobs
and ten different components of job requirements, we compare the effects on residual wage variation of including occupation
indicators and these skill requirements measures. We find that although these skill requirements vary across 3-digit occupations,
occupation indicators decrease wage residuals by far more than can be explained by skill alone. This indicates that “controlling
for occupation” does not equate to controlling for only these skill measures, but also for other factors. Additionally, we
find that there is considerable within-occupation variation in skill requirements, and that the amount of variation is not
constant across skill levels. As a result, including occupation indicators in a wage model introduces heteroskedasticity that
must be accounted for. We suggest that caution be applied when using and interpreting occupation indicators as controls in
wage regressions.