Is Thomas Sowell Right About Race and Sex Discrimination Lawsuits?

In an April 3rd piece at, “The Invincible Dogma,” popular conservative columnist, Thomas Sowell, built a case that many racial and sexual discrimination lawsuits are unwarranted. According to Sowell, “at the heart of these legal charades [i.e. anti-discrimination cases] is the prevailing dogma that statistical disparities in employment… show discrimination.”

As examples, Sowell discusses “unfounded” lawsuits against Federal Express and the Sears department store chain. Plaintiffs in these cases cited statistical evidence that showed racial/gender imbalances in the workforce of these companies. For instance, Sears’ workforce was overwhelmingly male – and that struck many observers as fishy or aberrant. Sowell’s point is that, imbalances in the workforce (gender imbalances, racial imbalances, etc) are often just statistical noise. In other words, there was no intention of malice or intended discrimination. Sometimes workforces are overwhelmingly male (or white, or what have you) due to sheer chance and random distribution.

Here’s a metaphor to help you understand Sowell’s reasoning. Imagine you flip a quarter five times in a row and get heads every time. You might think that the quarter was “biased towards heads” – and indeed, the coin might be a trick coin. But according to statistics, if you did this exercise 32 times, the likelihood of getting “five heads in a row” is one out of 32 — unusual, but not outrageously so. And if 32 people flip a quarter five times, odds are that at least one person will flip heads five times in a row – even given a fully “fair” coin.

Sowell also suggests that workplaces can be influenced by staggeringly diverse attitudes, skills, and histories. Odd-looking recruitment/hiring/firing patterns may thus have nothing to do with discrimination, even if they appear like they do. He suggests that there is “little evidence or logic…behind the belief that an absence of random distribution shows discrimination.”

Sowell makes a nice-sounding intellectual argument. But even if you do believe his position, you really need to be careful not to throw out the proverbial baby with the bath water. Yes, random fluctuations can make it look like discrimination is occurring – even when there is no discrimination and no intention of discrimination.

But that same “statistical noise effect” can cloak discrimination when it occurs!

One can easily imagine a workplace with a perfectly equal distribution of male and female employees… in which the corporate structure is biased against the female employees in subtle but very real ways. The statistics might suggest that “everything is hunky-dory.” But if you examined the company’s history on a granular level, you might discover that the number of potential female executives should have been twice the number of potential male executives! In other words, the hiring and promotion processes were by no means fair and meritorious. Sowell’s model would be utterly unable to detect this kind of subtle pattern of discrimination.

The general point is that arguments like Sowell’s can seem compelling – and they are certainly worthy of consideration. But there can also be a danger of “over-intellectualizing” and oversimplifying workplace discrimination and harassment issues.

If you or someone you care about needs helps with a case, connect with the experienced employment law team at Joseph & Kirschenbaum. Call (212) 688-5640 or visit us on the web at