Boys who grew up in families in the bottom tenth of the income distribution are about 20 times more likely to be in prison in their 30s than boys born into families in the top tenth:
That conclusion, reached by two researchers who leveraged millions of individual tax records, may not be surprising but should be shocking.
The Brookings Institution’s Adam Looney and the Federal Reserve Board of Governors’ Nicholas Turner, the two economists behind the paper, both used to work at the Treasury Department. That gave them access to unusually rich tax return data sets and made it possible to link people’s prison histories with their earnings and the earnings of their parents.
A few years ago, such a study wouldn’t have been possible. But beginning in 2012, the federal government started requiring state governments and the Federal Bureau of Prisons to report data on prisoners in their custody to the IRS. The main goal was to make it easier to audit prisoners. “The idea was, ‘Let’s audit prisoners to make sure they’re not claiming fraudulent tax refunds,’” Looney explains.
In trying to enable more audits, though, Congress also created the first national administrative database of all state and federal prison inmates. Really, the first: Before, there was just a jumbled mix of state data. The national database allowed Looney and Turner to look at a massive group of 2.9 million prisoners, which they winnowed down for their study to people with sentences of a year or more who were imprisoned as of December 31, 2012 — a group of about 1.4 million prisoners.
What the data tells us about employment before and after prison
Perhaps one of the most surprising findings of the study is that there’s little difference between the earnings and employment rates of prisoners before and after their incarceration:
After people are released from prison, they tend to have similar earnings, and about as many are employed as before they went in.
It’s important not to interpret these charts as causal; they don’t mean that prison is harmless as far as employment and earnings go. Looney notes that a significant number of these people may have been on their second or third prison stay, or at least second or third criminal conviction, and so the negative effect of conviction or imprisonment might have been baked in already for them.
This is also a group that was incarcerated in 2012, and so might have had unusually low earnings right before prison (in 2009 and 2010, say) and unusually high earnings after (in 2014 or 2015), due to the timing of the national economic recovery.
More to the point, other research better capable of isolating the causal effect of prison has found, unsurprisingly, that it reduces earnings and employment. Mike Mueller-Smith, an economist at the University of Michigan, has found that defendants randomly assigned to harsher judges (and thus given tougher prison sentences for basically arbitrary reasons) see a dramatic drop in employment and earnings. Out of felony defendants serving at least a year and with stable jobs before prison, prison reduces employment by 24 percentage points. Put another way: A quarter of these people (who, again, used to be gainfully employed) are pushed out of work by prison.
That said, Looney thinks the lack of difference between pre- and post-incarceration earnings and employment he finds in the IRS data also points at another bleak truth: Most people the US locks up grow up in poor disadvantaged areas where even before prison, your life prospects aren’t great.
Out of working-age men in prison during the period the study covers, only 49 percent worked three years prior to being incarcerated, and the median man who did work had earnings of only $6,250. One year after prison, 55 percent of prime-age men worked, and the median man working earned only $10,090.
On the one hand, the increase in employment rate and earnings after prison is surprising. On the other hand, the important point is that both before and after numbers are absolutely terrible.
“If you have grown up in a terrible neighborhood in a poor family, and you’re unable to get a good education, and you face discrimination based on race or class, there are a lot of challenges for you in the job market even if you haven’t been to jail,” Looney says.
Incarceration rates are highest in high-poverty minority neighborhoods
Sure enough, when they broke down odds of incarceration by zip code (the kind of analysis enabled by this detailed tax data), the zip codes where children face the highest rates were disproportionately poor and disproportionately black as of 2000, when most of the young adults in their sample were growing up:
If there was any doubt that the carceral state was locking up large shares of black communities, this list should remove it. There’s a 93 percent black zip code in Nashville; black neighborhoods in Fort Myers, Jacksonville, and Orlando, Florida; black parts of the Hampton Roads area in southern Virginia.
By contrast, the least-incarcerated zip codes, with rates of 0.0 percent, include posh areas like La Jolla, California, or Princeton, New Jersey, or über-white neighborhoods in Minneapolis and Portland, Oregon.
The data Looney and Turner use doesn’t include racial demographics. But it does let them see incarceration by parental income, which in part reflects racial income gaps. They also incorporate other data sets allowing them to estimate the size of the ex-offender and jailed populations so they can add them to the prisoner population they describe and see the full reach of the criminal justice system across the income scale. The resulting picture is stark and disturbing:
More than 30 percent of men ages 30 to 34 born to the poorest families were either in prison, in jail, or former prisoners. These numbers, in turn, go a long way in describing the growing population of American men who are outside the labor force. “Roughly one third of all not-working 30-year-old men are either in prison, in jail, or are unemployed former prisoners,” Looney and Turner write.
No data set can answer every important question, and Looney and Turner are clear about the limitations of this one. It can’t say, for instance, whether the massive difference in imprisonment rates between rich and poor Americans is the result of a difference in rates of actual criminal offense or a result of pervasive discrimination and aggressive policing of poor communities.
But whatever explains this gap, it’s a massive societal problem, and one concentrated in poor and disproportionately black and Latino areas. “In many of those places, just an enormous share of the male population will end up in prison,” Looney notes. That doesn’t have to be the case.