Here’s How Government Data Minimizes Racial Inequality

As you may have heard, in order to justify the recent military occupation of the nation’s capital, Donald Trump alleges that the D.C. Metropolitan police department has been lying about crime numbers in the District.

Supposedly, they’ve been cooking the books to make it seem as though crime has been dropping dramatically, even though “everyone knows” that D.C. is, as Stephen Miller puts it, “more dangerous than Baghdad.”

And by “everyone,” they mean that guy, “Big Balls” from DOGE, who got his ass kicked by a 15-year-old would-be carjacker a couple of weeks ago, prompting Trump to call in the troops.

Anyway, although the D.C. police may have provided inaccurate numbers that downplay the extent of crime in the District — I mean, lying is sorta what cops do, and if it makes them look better at their job than they are, all the more incentive, I guess — still, the president hasn’t provided evidence to that effect.

Just like he can’t provide numbers to counter the embarrassing ones from the Labor Department, indicating how anemic job creation has been in recent months, which led him to fire the head of the Bureau of Labor Statistics.

Oh, and speaking of the BLS.

I agree that the Labor Department’s statisticians have a problem with accuracy, but not when it comes to the monthly job creation count. Despite regular revisions, sometimes up and sometimes down, they generally do a pretty respectable job on that front, and whatever difficulties they have getting it right stem more from non-cooperation by employers and industries surveyed for the monthly projections than from anything internal to the BLS itself.

The same cannot be said, however, for the way the Bureau handles unemployment data, specifically unemployment by race and ethnicity.

Here, they have long presented data on that subject in a way that diminishes the extent of racial inequality in the job market, and especially the ongoing massive advantages enjoyed by white Americans when it comes to finding work.

So here’s the deal.

Taking a look at the most recent job numbers for July 2025, you notice that even in the official data there are racial gaps in unemployment that are meaningful and troubling: 7.9 percent for Black folks and 5.1 percent for Latinos, compared to a much smaller 4 percent for whites, meaning that according to the BLS, Black unemployment is just a bit shy of double the rate for whites, while Hispanic unemployment is about one-fourth higher than the rate for white Americans.

But in reality, things are worse than this.

The fact is, the Labor Department essentially double-counts Hispanic folks. First, they pull them out in a category of their own, so we can get a sense of the unemployment status of persons in the Latino community, which makes sense.

But then they also fold Hispanics into the three main racial groups for whom data is provided: Whites, Blacks, and Asian Americans. They do this because, as they note, Hispanics can be members of any racial category and are not a separate race, per se.

As it turns out, according to a methodological note in the department’s annual report, Labor Force Characteristics by Race and Ethnicity, nearly 9 in 10 (89 percent) of Hispanics in Labor Department data are classified as white, racially. Meaning they will also be included in the totals for whites in the workforce and among whites who are unemployed. Here’s that receipt from the most recent annual report:

Press enter or click to view image in full size

Why that’s a problem should be obvious.

If Latinos are more likely to be unemployed than non-Hispanic white folks (and they are), then folding 89 percent of those Latinos into the white numbers, rather than disaggregating the white totals and presenting a number for non-Hispanic whites specifically, will artificially inflate the unemployment numbers for white people. This will then artificially diminish the evidence of racial disparity in the job market, making the problem of inequality seem much smaller than it is.

Let’s take a look at how the BLS distorts the truth in its latest monthly employment report.

First, let’s look at the numbers for persons in the civilian labor force, for whites (including the Latinos who are folded in with the white group), and then for Blacks and Asians. The labor force number is the important first step because this number will be the basis for calculating the unemployment rate, once we get to the number of unemployed persons below. Here’s the data for whites (again, including 89 percent of Hispanics in the labor force):

Press enter or click to view image in full size

As you can see, there are 129.5 million “whites” in the labor force. But now we need to figure out how many of these aren’t really white in the sense most people consider the term, because if we’re trying to get an idea about the labor market position of whites compared to people of color, it makes no sense to have large numbers of those people of color actually sitting in the white group’s data.

So what does that data look like? It’s here:

Press enter or click to view image in full size

As you can see, there were roughly 34.6 million Latinos in the civilian labor force in July. And if 89 percent of those are classified as white, that means 30.8 million of those Hispanics are also to be found back in the white column in the previous table, and would need to be extracted from it to get a sense of the number of “real whites” in the labor force.

When we do that calculation, it leaves us with 98.7 million whites (non-Hispanic) in the labor market.

Then here are the numbers for Blacks and Asians: 22.3 million and 12.6 million, respectively.

Press enter or click to view image in full size

Press enter or click to view image in full size

Now, let’s get the unemployment numbers. For simplicity’s sake, I am using the non-seasonally adjusted numbers; however, you could also use the seasonally adjusted ones, and the results wouldn’t change the trajectory of the argument I’m making below. The numbers themselves would be slightly different for all groups, but the racial disparities would remain functionally the same.

According to the BLS, there were 5.15 million “whites” unemployed in July, but again, this includes lots of Latinos who are classified racially as white. There were 1.78 million Latinos unemployed that month, and if 89 percent of those are classified as white, then roughly 1.6 million of these would need to be backed out of the white unemployment numbers to give us an accurate picture of how many non-Hispanic whites are out of work.

That means that rather than 5.15 million unemployed whites, the real number is more like 3.55 million. So there are 1.6 million fewer unemployed white people than the BLS would have us believe!

That same month, there were 1.77 million unemployed African Americans and 524,000 unemployed Asian Americans.

Next, we can determine the unemployment rates for each group. In the officially released data, “white” unemployment was 4 percent, but that was when 1.6 million out-of-work Latinos were still in the white column. By backing them out, the white unemployment rate for July drops almost a half a point to 3.6 percent (3.55 million/98.7 million)

This means that rather than Black unemployment being 1.9 times greater than white unemployment the actual ratio is 2.2 to 1 (7.9 to 3.6) Hispanic unemployment, rather than being 1.25 times higher than white unemployment is really 1.4 times higher (5.1 to 3.6), and even Asian American unemployment, which was essentially the same as for whites in the official data, is about 17 percent higher than for whites (4.1 to 3.6).

While these may seem like relatively minor alterations, in truth, they represent evidence that the racial gaps in labor market outcomes remain wildly unequal, and more so than is imagined. If the Labor Department is exaggerating white unemployment by more than a million and a half people each month, they are contributing to the white racial grievance fueling much of MAGA at present, and diminishing the reality of white advantage relative to folks of color.

None of this is to say that millions of white people aren’t hurting. They are, to be sure, struggling with the costs of healthcare, higher education, rent, groceries, and everything else. But in a nation where white people are being told we are the new victims, the forgotten, and the left behind, the data correction serves as a stark reminder of how nonsensical that argument is.

Black folks, Latinos, and Asian Americans, too, continue to have a harder time finding work than whites. And by the way, according to that annual report I mentioned earlier — a new one of which should be released within a few months — even when those folks of color have college degrees they are between 25 and 67 percent more percent likely to be out of work than their white counterparts in most years (again, once Latinos are backed out of the white totals).

So when you hear Trump and his supporters complaining about “fake data” when it comes to crime or job creation numbers, remember, they have no receipts. But when it comes to downplaying racial disparity in America, the data is as phony as the president’s business acumen.

We need an honest conversation about racial disparity and the way official data minimizes the problem. What we don’t need are white grievance grifters and racists acting like we are the ones being marginalized. Stop whining. Facts matter.