Thursday, January 11, 2018

Measuring Bias, or Reflecting Reality?

The above graphic, from today's NY Times, is intended to show bias in economics textbooks. I am not so sure that is the right interpretation.

In the world, only 4 percent of CEOs (of Fortune 500 companies) are women, so does the figure of 6 percent shown above demonstrate underrepresentation of women in textbooks or an accurate reflection of reality? Similarly, policymakers mentioned in texts are most often Presidents or Fed chairs. Historically, only one woman has been a member of this group. Economists mentioned in texts are most often important historical figures (Smith, Ricardo, Keynes) or prominent modern economists, such as Nobel laureates. Once again, 8 percent is higher than for the population being sampled.

To be sure, the role of women in society is changing, and in some circles there is some bias. But measuring the amount of bias is hard. The graphic above is not a useful gauge.

Or maybe in my next edition, I should add a discussion of Paulina Volcker's disinflation.


Update: Some twitter commentators seemed to misinterpret my cheeky last line. At the risk of being pedantic, let me explain: Textbooks reflect reality, which includes a history in which men played a larger role than women in some spheres of life. If a history professor were to write a text on the history of presidential politics, and you were to find that there were more mentions of men than women in the book, would that be evidence that the historian is biased? I don't think so. The writer has to reflect what occurred and is not free to change the gender of historical protagonists.