Friday, February 6, 2015

Berkeley Pay Equity Study

Inside Higher Ed today has a lengthy article summarizing a study at UC-Berkeley that looked at pay equity issues, (white) male vs. female and white male vs. minority.  The chart above from the study itself is one of several charts that show a non-standardized differential and then apply increasing levels of statistical control.  Generally, the more you control for, the lower the pay gap.  Typically, a large starting (uncontrolled) gap becomes small.  (In some fields, the gap disappears and switches from negative to positive.)  Because controlling for influences such as rank might over-control by filtering out possible discrimination in promotion, the study shows the impact of varying the degree of control.*

You can find the original Berkeley report at

The Inside Higher Ed article is at
*Note that not so long ago in the UC system, with its ranks and civil-service-style steps, if you controlled for rank and step, you would essentially have controlled-out any deviations in pay since there were very few deviations in pay from the official scale.  If there was discrimination, it would have had to come from the degree to which women and minorities advanced through the ranks and steps.  Since the budget crisis of the early 2000s, however, pay has increasingly deviated from official pay rates.

No comments: