We hear a lot about the unemployment rate as one number.

But it’s not: It can be wildly different depending on your age, gender, education, income — and race.

We made a chart that lets you explore each of those variables.

The devastating impact of the pandemic on jobs in this country has prompted trillions of dollars of aid.

*But* not only are Black Americans and other workers of color getting hit harder now, they were long before the virus took its toll.

The Black-white gap in unemployment is unbelievably persistent: across time (it’s existed since we started measuring it, as @SandyDarity puts it), location, income, age and education. Here’s what it looks like over the past decade:

Contrary to popular belief, education does not level the playing field.

The gap exists at every education level, as this report by @ValerieRWilson and @jhacova makes clear: https://www.epi.org/publication/labor-day-2019-racial-disparities-in-employment/

The reason for this disparity is no mystery: It’s racism.

It manifests in the labor market in many ways, but for one, Black job candidates are just less likely to get hired than white ones: https://www.pnas.org/content/early/2017/09/11/1706255114 (via @JessicaJFulton)

Now, to the data itself. The unemployment rates, while based on the best available current data from the @uscensusbureau via @ipums, are still just estimates. So these numbers are correct as trends but not precise to the decimal.

That said, it’s likely that these #'s actually make the unemployment picture look *better* than it really is.

That’s bc of ambiguity in who was classified as unemployed at the start of the pandemic & bc many people have left the workforce (so don’t count in the official rate)

One major limitation of the data: small sample sizes.

There were lots of segments of the population that we couldn’t show because the numbers would have been too unreliable. Thanks to my colleague @HannahFresques for all her help with these calculations.

Relatedly, we thought it was important to show Native Americans, Asians and people of 2+ races in distinct categories.

But that means that you quickly get small samples if you divide by gender or age, for example. So a lot of those groups are missing fine-grained detail.

Finally, this piece is heavily inspired by a classic data visualization the @nytimes put out after the 2008 recession, called “A Jobless Rate for People Like You.” It’s Flash, but you can still see it here: https://archive.nytimes.com/www.nytimes.com/interactive/2009/11/06/business/economy/unemployment-lines.html?country=884

Explore the data for yourself and read more here:
https://projects.propublica.org/coronavirus-unemployment/

PS: Get notified whenever ProPublica drops a major investigation: https://propub.li/2LdoRh6