A Solution for the Gender Pay Gap

A Google engineer just touched the untouchable in office politics by openly arguing that biological “differences may explain why we don’t see equal representation of women in tech and leadership.” However, the claim and many furious responses are mostly lack of evidence. It is understandable that people resort to emotional statements when an ideologically-rooted issue is discussed. But can “data” in gender gaps, say salary differences, offer insight to this debate? Unfortunately, the answer is no.

If we can have two people who are identical in their genes and life experiences, except for their gender, then the pay difference from the same employer can offer insight to whether there is discrimination or biases at work. All other data interpretations suffer the omitted variable bias. That is, there is always some other factors that we cannot document in explaining the difference. For example, we know skill can partially explain the pay difference, but to define and quantify the skill difference itself is an impossible task.

Even if we are able to control all internal factors, there are still external ones to deal with. Use my industry as an example. One of the possible reasons why female professors are paid less than their male colleagues is the difference in supply of female Ph.D.s. In disciplines such as sociology, plenty of female candidates with impressive c.v. and excellent caliber fight for very scare openings, while disciplines like accounting have a hard time getting any candidates. The colleges and universities will naturally offer higher salaries to accountants than to sociologists. Since women dominate sociology, but are more equally represented in accounting, the end result in higher education is a gender pay gap. But is it wrong for the administrators to save some money by having a pay gap? Hardly. But is there discrimination hidden behind the market force? We don’t know and will never know.

Nevertheless, data can still play a constructive role. Data, especially price data, are signals to suppliers and consumers. We should publish and publicize the gender pay gap. If the pay difference is caused by biases, it means opportunities for the executives who care about the bottomline to hire more female employees. If they are foolish enough to place discrimination above productivity, why get in the way of a natural balancing act? More importantly, the detailed pay differences across industries will guide aspiring young women to choose careers with better pay. That is the right way to get to the natural equilibrium without biases.