Parents and Prisons

The negative effects of fathers' absence from children's lives—on kids' education, mental health, employment, and family formation, among other factors—are well-established and apparent across countries. Incarceration is one way that parents, and particularly fathers (men make up about 90% of prisoners), are separated from their children. Roughly 5 million children, about 7%, have experienced a parent going to prison at some point in their lives.

That incarceration can disrupt family life is one reason many more-traditional conservatives have found common cause with advocates of "criminal justice reform." Pointing to evidence that children with incarcerated parents do worse on a host of indicators than their peers with intact families, they argue that the costs of "breaking up" a family often outweigh the benefits to public safety, and that therefore pro-family conservatives must necessarily be in favor of decarceration.

Much of the evidence used to make this argument, however, is limited in its capacity to really establish a causal link between parental incarceration and worse outcomes. Children of incarcerated parents may already be on a worse path for reasons other than the fact of their parent's incarceration—a worse family environment precipitated by the same character traits that made their parent likely to be a criminal offender. To really separate out the effect of parental incarceration on children's outcomes, researchers need a source of randomness in a parent's propensity to be incarcerated, one necessarily uncorrelated with their children's outcomes.

A new study, forthcoming in the American Economic Review, identifies just that, and the results are startling. Contrary to the assumption that parental incarceration is destructive to children, the paper finds that parental incarceration reduces children's propensity to be incarcerated and improves their adult neighborhood quality, while having no significant effect on their academic performance.

These findings should not make us doubt the importance of intact families and fathers in the home in the vast majority of cases. But they should put a dent in the idea that incarceration is bad because of its effects on family—the story is more complicated than that.

To reach their conclusions, the paper's authors assembled a rich array of data on prisoners and their children from the three Ohio counties home to Columbus, Cleveland, and Cincinnati. About 60% of the parents in the survey are male, with an average of 1.86 children, who range in age from 1 to 19 years old at the time of their parent's court date. Using these data, they conducted a "judge assignment" study, taking advantage of the random assignment of offenders to judges with a known variability in their propensity to incarcerate as a source of randomness that permitted isolating the effects of parental incarceration from other variables. 

Their biggest finding is the effect on children's future propensity to be incarcerated: parental incarceration was linked to a 20% reduction in a child's likelihood of being charged, a 22% cut in chance of being convicted, and a 40% reduction in the chance of being incarcerated before age 25. Notably, those effects are driven by the black children in their data.

Just as surprising are the other findings. Parental incarceration actually led to a slight increase (4.1 percentiles) in the socioeconomic status of the neighborhood the child ended up living in. And it had almost-zero, non-significant effects on children's academic performance (math and reading scores, GPA, absenteeism, risk of repeating a grade) and risk of teen parenthood—parental incarceration did not improve these measures in the study, but it didn't harm them, either.

The Ohio study is actually not the only study to find that parental incarceration can have positive effects. Research from Colombia finds that parental incarceration increased children's years of schooling, while evidence from Norway offers similar null estimates.

Clearly, there is a small population of parents, prone to other profoundly antisocial behavior, whose presence in the household is harmful enough to outweigh the general benefits of intact families.

So what's behind these counterintuitive findings? And what do they say about the importance of family structure to future outcomes in general?

The original paper suggests there are three effects at play: effects on the other parent, deterrence through the example of a parent's incarceration, and incapacitation of a parent who would otherwise be a bad influence. The authors show that a parent's incarceration produces, on average, a small but significant reduction in his partner's future risk of being charged and incarcerated. And they find that the effect of parental incarceration on children is not clearly related to length of sentence, suggesting that both deterrence and incapacitation play a role in the effect.

That's a very technical way to think about the cause of these outcomes. A simpler way is to realize that the figures reported above represent average effects, combining the children for whom parental incarceration was a net positive and those for whom it was a net negative. Not all parents are the same: one would be far more likely to believe that the incarceration of, say, Charles Manson would have a positive effect on his children than, say, the incarceration of Mr. Rogers.

The subset of the population that ends up incarcerated is disproportionately likely to have character traits particularly adverse to good parenting. As my colleague Rafael Mangual has noted, antisocial personality disorder is 10 to 23 times more common in prisoners than it is in the general population, suggesting that prisons tend to select personalities at risk for harming their children. High rates of recidivism—more than 80% of prisoners released are rearrested within 10 years, the Bureau of Justice Statistics has found—similarly support the idea that prison populations are made up of those most prone to taking advantage of and harming others, including their children. The evidence on partner behavior in the paper, indicating that some parents make their partners more likely to commit crimes, also supports this notion that some parents are just dangerous influences.

The paper, then, should not cause us to revise our estimation of the importance of intact families in general, and fatherhood in particular—most dads are not the sort who will go to prison. But it should cause skepticism of the conventional wisdom that incarceration cannot be justified because of its adverse effects on families. Clearly, there is a small population of parents, prone to other profoundly antisocial behavior, whose presence in the household is harmful enough to outweigh the general benefits of intact families. In their case, incarceration may turn out to be a benefit, not a harm, to spouse and child.

Charles Fain Lehman is a fellow at the Manhattan Institute and a Contributing Editor at City Journal.