Philip Bump’s Argument for DEI in the Federal Workforce Is Pretty Weak

Philip Bump has a story at the Washington Post today arguing that President Trump’s elimination of DEI in the federal government is a perpetuation of white privilege. But the data he offers to make his case doesn’t really prove what he’s trying to prove, at least not about race.

When the Government Accountability Office (GAO) published its assessment of the composition of the federal workforce a few years ago, it revealed an interesting data point: As of 2021, people with disabilities were about three times better represented in government jobs than in civilian jobs.

This isn’t an accident, obviously. The government is deliberate about ensuring that physical constraints aren’t barriers for employees — deliberate as a matter of law, yes, but also as a matter of intent. The government has systems and processes that encourage people with disabilities to seek and obtain employment.

That GAO report also detailed other data about the federal workforce, indicating that the demographics of government employees hadn’t changed much over the preceding 10 years. In 2011, the start of the window considered in the report, about 72 percent of federal employees were White (including White Hispanic people). In 2021, the most recent year with data, 68 percent were. The percentage of Black (and Black Hispanic) workers stayed steady at about 20 percent of the workforce; the increase in diversity was mostly among other racial groups, particularly Asian Americans.

Here’s a chart from the GAO report he’s relying on:

As you can see, the percentage of white federal employees has been trending down while the percentage of black, Asian, American Indian and mixed race people has been trending up. Here’s how Bump sums up the data.

The result was a workforce that looked a bit different than the overall labor pool — a bit less White, a bit more Black — thanks in part to deliberate efforts to “advance equal opportunity” among all Americans. The Office of Personnel Management’s Office of Diversity, Equity, Inclusion, and Accessibility (DEIA) was the leading such office in the federal government, the report noted, providing other agencies with “concrete strategies and best practices” to hire and retain a “diverse, results-oriented, high-performing workforce.”

OPM’s DEIA office is now gone, eliminated by one of the first executive orders President Donald Trump signed upon his return to the White House this week.

He wants to frame this as a terrible outcome but based on the numbers he’s provided what we have is not a federal workforce that looks like American but one that is “a bit less White, a bit more Black” than America. In other words, whites are slightly underrepresented (and trending down) and Black and Asian Americans are overrepresented and trending up.

So this isn’t a situation where Black Americans are underrepresented in the federal workforce compared to their percentage of the US population, it’s one where there are overrepresented. This is the system that Bump thinks needs to be preserved for the benefit of “historically disadvantaged racial groups.” Bump doesn’t linger on this fact but instead immediately turns to a different question: How are black and white federal workers represented in different levels inside the federal government? 

Consider another aspect of that GAO report. It noted that, though about 2 in 10 entry-level employees were Black women, only 1 in 10 senior-level employees were. About 3 in 10 entry-level positions were held by White men — while White men held a little under half of senior-level positions. Black men made up a consistent portion of each level of employment; White women were decreasingly represented at each higher tier.

Here he may be on to something but it’s hard to tell because the only factor he’s interested in is race and gender. So, for instance, the possibility that white people are overrepresented at the mid and top levels because they have higher levels of education isn’t considered even though it might explain some of the difference. About 40% of Black Americans complete a 4-year degree (within 6 years) compared to 62% of white Americans. This isn’t just a meaningless status thing. Federal workers are mostly white collar workers and are actually more highly educated than the general workforce.

As a whole, federal workers are more educated than the overall civilian workforce. Nearly a third of federal workers (31.5%) have a bachelor’s degree, compared with 27.7% of all employed Americans. And almost 22% of federal workers have an advanced degree, versus 17.6% of all workers.

In other words, that 22% gap in who completes 4-year degrees probably makes a big difference when it comes to who advances to leadership positions. But Bump doesn’t even mention education except in passing as a sign of “lingering discrimination.” He wants to present this difference as the result of systemic racism.

Which really raises another question. If systemic racism is still dramatically in force inside the federal workforce, doesn’t that suggest that decades of DEI efforts aren’t really working? If some form of racism is the only explanation here, then apparently the current solutions have not worked well at all.

Again, I think there are probably other factors that explain some of these differences but Bump is sidestepping those because they might tend to undermine his conclusion:

That upending affirmative action programs would entrench the advantages of White men is not a countervailing argument. It is very much the point.

Ending DEI and affirmative action is an attempt to shore up white supremacy. That’s Bump’s argument. But is a federal workforce that is “a bit less White, a bit more Black” than the civilian workforce really proof that DEI and affirmative action are needed to achieve demographic representation that matches the makeup of the country? I guess we’ll see over the next four years how the numbers change.