The high economic costs of incarceration

The United States leads the world when it comes to incarceration rates. According to one estimate, there are 2.4 million individuals in U.S. jails or prisons—a massive number that stands in sharp contrast to the decline in violent crimes over the past four decades. The increased incarceration also caused a steep rise in the imprisonment of so called marginal defendants, or those who are on the edge of going to jail or staying out of it. The costs of this incarceration rate to the U.S. economy are strikingly large, according to a new paper harnessing a large and impressive data set.

The paper, by Michael Mueller-Smith, a PhD candidate at Columbia University, looks specifically at the effects of incarceration in Harris County, Texas, which includes the city of Houston. While the paper only looks at one county, the data set Mueller-Smith uses covers every single defendant in the county. The data cover all 1.1 million defendants from 1980 to 2009. The data also connect via court records to other data that reveal information about future criminal activity, wages, reliance on government assistance programs such as the Supplemental Nutrition Assistance Program (known as food stamps), and marriage records.

Mueller-Smith’s impressive array of data enables him to demonstrate causality due to the sentencing procedures used by Harris County. The county court system randomly assigns defendants to courtrooms, judges, and prosecutors. Because judges and prosecutors vary quite a bit in their desire or willingness to put defendants in jail or prison, this variation (which is determined randomly) lets Mueller-Smith identify the causal effects of incarceration.

In his calculations, Mueller-Smith focuses on the effects of incarceration on the marginal defendant, which means his study factors out the kind of defendants that society expects to be incarcerated, such as rapists or murders, so that he can zero in on those who commit crimes that could be punished with probation.

Given Texas’s commitment to be “tough on crime,” it’s likely that the marginal defendant in Harris County is committing crimes deemed less dangerous in other legal jurisdictions around the country. This means Mueller-Smith’s estimates might be biased toward finding more costs to incarceration. Furthermore, the research literature on the labor market effects of incarceration are mixed.

Even so, his findings are striking. He finds that the incarceration of marginal defendants increases their future criminal activity, reduces their employment prospects, increases their use of public benefits, and reduces their opportunities to get or remain married.

Specifically, Mueller-Smith finds that each additional year of being incarcerated increases the probability of facing a charge for a new crime by 5.6 percentage points per quarter after release. And he finds that released individuals are more likely to escalate the type of crime they commit after release, such as moving up to theft and burglary or drug crimes.

Incarceration also causes severe labor market problems for these marginal defendants. For each year a person is in jail or prison, their employment drops by 3.6 percentage points. For those convicted of a felony, the drop is even more severe: five years after release their employment is reduced by 24 percent. These marginal defendants also come to rely more on public assistance programs such as food stamps, another sign of a fragile attachment to the labor force.

The effects bleed into the defendant’s family life, too. Mueller-Smith finds that incarceration reduces marriage rates for younger defendants and boosts divorce rates for older defendants.

Putting all these effects together, Mueller-Smith calculates the social cost of incarceration. According to his estimate, one year of prison or jail for a marginal defendant has a social cost of ranging from $56,200 to $66,800.

Incarceration is supposed to punish convicted criminals but also reform prisoners to help them prepare for a life after jail or prison. Mueller-Smith’s results show that incarceration severely harms those marginal defendants who are best suited to be rehabilitated by reducing their labor market opportunities, damaging their families, and increasing their criminal activity.