The Wall Street Journal published a story this week about DEI supporters who think corporate DEI needs a makeover. It turns out that a lot of the stuff being promoted in the name of DEI either doesn't work or borders on illegal.

Some of the most visible steps that businesses took after George Floyd’s murder in 2020 turned out to be ineffective, according to Harvard sociologist Frank Dobbin, whose research on racial and gender disparities in management is often cited as evidence for robust diversity programs. Those pledges companies made to fill certain percentages of leadership roles with women and people of color? Besides being legally risky, the goals seldom materialized, he says.

And the unconscious bias training that was supposed to produce antiracist allies? Turns out people hate being told they have hidden prejudices.

Dobbin is an interesting character. In 2018 he wrote a critique of corporate DEI training titled "Why Doesn’t Diversity Training Work?" Here's a sample:

...two-thirds of human resources specialists report that diversity training does not have positive effects, and several field studies have found no effect of diversity training on women’s or minorities’ careers or on managerial diversity. These findings are not surprising. There is ample evidence that training alone does not change attitudes or behavior, or not by much and not for long. In their review of 985 studies of antibias interventions, Paluck and Green found little evidence that training reduces bias. In their review of 31 organizational studies using pretest/posttest assessments or a control group, Kulik and Roberson identified 27 that documented improved knowledge of, or attitudes toward, diversity, but most found small, short-term improvements on one or two of the items measured. In their review of 39 similar studies, Bezrukova, Joshi and Jehn identified only f ive that examined long-term effects on bias, two showing positive effects, two negative, and one no effect.

But an article Dobbin co-wrote for the WSJ last year seemed to lean heavily on implicit bias testing (though the authors said it was being misused by DEI trainers to shame people). But a response to that article in the City Journal points out the deeper problem: Implicit Bias is junk science.

The problems with DEI trainings are not in their tone, however, but in their substance. The implicit-bias theory (also called unconscious-bias theory) on which these trainings are based has no scientific basis, as years of examinations have consistently demonstrated. Lee Jussim puts it politely in his “12 Reasons to Be Skeptical of Common Claims About Implicit Bias,” but the Open Science Foundation’s archive of Articles Critical of the IAT and Implicit Bias renders a harsher verdict. In 2011, Etienne LeBel and Sampo Paunonen reviewed evidence that measures of implicit bias possess low reliability. In other words, when you test for implicit bias multiple times, you rarely get the same result. Their conclusion was that some part of “implicit bias” is really “random measurement error.” In 2017, Heather Mac Donald’s intensive examination of the theory and its empirical basis (or lack thereof) concluded that the “implicit-bias crusade is agenda-driven social science.” And Bertram Gawronski’s 2019 review of the scholarly literature on implicit-bias research also concludes that there’s no proof that people aren’t self-aware enough to know what’s causing their supposedly “implicit” or “unconscious” biases; and that you can’t prove that there’s any relationship between how people do on the test and how they behave in the real world...

Professional critiques of implicit bias have shown, politely but repeatedly, that there is nothing there. Activists and scientists who think that science should serve political objectives want to believe in the existence of massive systemic bias to justify their goals of imposing “equity” by law and by litigation. Implicit bias is a pseudoscientific theory made to order for this purpose. It’s a house of cards, and governments and the private sector should terminate every program based on it.

All that to say, Dobbin is right that DEI training doesn't work but there's no rescuing a field that is based on junk science that doesn't measure anything. The other claims supporting corporate DEI also have serious problems.

The argument that more diverse leadership teams are directly linked to greater profits has come under fire. Still, companies know they risk missing talent if they don’t actively recruit a range of people, says Fayruz Kirtzman, a senior client partner at Korn Ferry who advises companies on using DEI to help the bottom line.

She says a lot of businesses stumbled in recent years by publicizing demographic targets. Intended to keep companies accountable to the numbers, the goals were easily misconstrued as reverse discrimination.

“By the time the message trickles down, it’s heard as, ‘They won’t hire a white guy,’” she says.

It's probably heard that way because that's actually what is happening in many cases. Companies, like colleges, will travel to the ends of the earth to find a competitive black candidate and in the process ignore or overlook the white (or Asian) candidate standing on their doorstep. As mentioned, the justification for a lot of this is the dubious claim that diversity leads to greater profits. That claim has been trumpeted for several years by McKinsey & Company. But academics who attempted to replicate their results based on public information couldn't do it.

Dame Vivian Hunt, McKinsey’s managing partner in the UK and Ireland at the time and a coauthor on all four of McKinsey’s studies, crystalizes McKinsey’s view that greater racial/ethnic diversity in a firm’s executive team drives better firm financial performance: “What our data shows is that companies that have more diverse leadership teams are more successful. And so the leading companies in our datasets are pursuing diversity because it’s a business imperative and driving real business results”

...the goal of our paper is to revisit McKinsey’s results by means of a quasi-replication...

In contrast to McKinsey’s results, the key finding of our study is that we observe no statistically significant difference between the likelihood of financial outperformance as measured by the industry-adjusted EBIT margin of S&P 500® firms during the years 2015–2019 in the top vs. bottom quartiles of S&P 500® firms ranked on McKinsey’s executive racial/ethnic diversity metric measured in mid-2020. Instead, we find that 54.0 percent of S&P500® firms in the top executive race/ethnicity-ranked quartile have a positive industry-adjusted EBIT margin vs. 51.2 percent in the bottom quartile, with the z-statistic on the difference of 2.8 percent being a not statistically significant 0.5 (p-value = 0.65).

The two accounting professors behind this paper found no statistically significant difference and therefore no evidence that diversity yielded greater success/profits.

To sum up, DEI training doesn't work, is based on junk science and doesn't produce measurable results for the companies that adopt it and yet, 95% of companies claim to have a commitment to DEI. Something is very wrong with this picture.