Trump's Disparate Impact Executive Order

The mainstream media seems stumped by how to respond.

It’s interesting how little mainstream media attention has been devoted to the Trump Administration’s April 23rd Executive Order against the use of the disparate-impact liability concept in civil rights enforcement. For example, the New York Times has run one sentence on the news in the week since:

Another restricts the use of the so-called disparate impact rule, which civil rights groups have long said is an important tool for showing discrimination against minorities.

Here are the top Google hits for Disparate Impact Executive Order:

a Heather Mac Donald essay in City Journal and four technical posts for lawyers.

Disparate income theory goes back to the 1971 Griggs decision by the Supreme Court that, in effect, put the burden of proof on employers to prove the “business necessity” of their hiring policies that have unequal outcomes by race (later defined by the EEOC as one race being hired at least 20% less than another race). After the Supreme Court rolled it back somewhat in 1989, Congress rewrote it into law in the 1991 Civil Rights Act signed by George H.W. Bush.

Perhaps in 1971 the social science that equal opportunity wouldn’t lead to equal results between the races wasn’t yet well understood. The federal Coleman Report of 1966 had been a time bomb, but was mostly ignored at the moment. Daniel Patrick Moynihan led seminars at Harvard over the next few years digging into the Coleman Report. By the early 1970s, there were a new generation of social scientists who had started to digest the realities, but you could hardly expect Supreme Court justices to be up to date.

By 1991, however, we were a quarter of a century into the new era of social science, so Congress and the elder Bush are less forgivable.

Today, though, we are 59 years out from the Coleman Report.

Keep in mind that disparate impact law is unreasonable but not absurd. Firms can sometimes prove the validity of their hiring practices. For example, my impression is that Procter & Gamble spent a large amount of money to validate their famous hiring exam.

On the other hand, the marketing research firm I worked at on the P&G account had long used as a client service hiring exam the Quantitative Methods in Marketing Research final created by one of the founders, a Big Ten professor. When I took it in 1982, I thought it was a beast of a test and I spent 3 hours working on it.

But then after they graded it the next day, the CEO called me up to spend a day interviewing.

Judging by the quality of the people hired, it appeared to be highly valid. But after decade or so the firm was big enough to get on the EEOC’s radar. When challenged to prove it was valid, management punted rather than spend the money and threw out the test. The quality of new hires seemed to decline.

But it’s hard for the New York Times to explain any of these realities to its eleven million subscribers — Despite six decades of social science research, it sounds racist to paying NYT customers say that blacks can’t achieve equal outcomes statistically without racial favoritism or at least without crippling good hiring practices — so it’s mostly ignoring Trump’s disparate impact executive order even though it’s a big deal.