Trump’s statement is accurate. The official black unemployment rate reached nearly 17% in March 2010, following the Great Recession of 2008, but then plummeted during the second half of the Obama administration. In December 2017 it was was 6.8%, which is the lowest rate recorded by the Bureau of Labor Statistics since they began collecting this statistic in 1972. Still, just as it has been for the past 40 years, black unemployment remains much higher than the overall rate, which is now 4.1%. A good rule of thumb is that the black unemployment rate is a bit less than double the overall unemployment rate.

But what Trump, along with most politicians and economists, misses when talking about black unemployment is that it is actually a significantly larger problem than official numbers indicate. That’s because the US government’s employment statistics do not include people who are incarcerated.

In 2016, there were 1.5 million people (pdf) in federal and state prison in the US.  Although only about 13% of the US population in 2016 was non-Hispanic black, about 40% of those incarcerated were. The imbalance is particularly pronounced among incarcerated men: 41% of incarcerated men are black, compared with 24% of women.

If incarcerated people were counted as unemployed, which they reasonably could be, unemployment statistics would look very different. As Jeff Guo pointed out (paywall) in the Washington Post, had incarcerated people been included in the unemployment rate for men between 25 and 54 years old (considered the prime working age) in 2014, then the unemployment rate for black men would rise steeply, and only modestly among whites.

In the 1940s, when the US began collecting unemployment statistics, excluding incarcerated people didn’t make much of a difference: For most of the 20th century, the incarceration rate hovered around one out of every 1,000 Americans. But in the late 1970s, harsher sentencing policies, more stringent policing, and aggressive prosecutors sent that rate rate skyrocketing. Today, nearly 7 out of every 1,000 people are incarcerated. Black Americans had always been overrepresented in the US prison system, but mass incarceration hit that group particularly hard.

Ignoring the incarcerated does not just distort unemployment statistics either. In her 2012 book Invisible Men: Mass Incarceration and the Myth of Black Progress, sociologist Becky Pettit shows that improvements in the black high school graduation rate, and the convergence of the black-white wage gap, would be significantly diminished if the increase in incarceration since the 1970s was accounted for.

There is some good news, though. From 2008 to 2016, the number of incarcerated black men fell to nearly 490,000 from about 590,000—due largely to declines in incarceration in California and New York. Perhaps someday, official unemployment statistics will actually represent reality.

📬 Sign up for the Daily Brief

Our free, fast, and fun briefing on the global economy, delivered every weekday morning.