Parenthood Helps Mom and Dad Quit Crime

Having a child is a profoundly important milestone in any adult’s life. You may change where you live, where you work, and with whom you spend your time. Perhaps unsurprisingly, a new paper published by the National Bureau of Economic Research finds that parents also become far less likely to commit crimes.

The study, authored by economists Maxim Masenkoff and Evan Rose, uses a unique dataset that combines records on the arrests, births, marriages, and divorces associated with over a million parents in Washington State. Using these data, the authors address a deceptively simple question: what happens to mothers’ and fathers’ propensity to commit crimes when they have a child? More specifically, Masenkoff and Rose are interested in what the qualitative literature calls the “turning point” model of parenthood, which has many first-time parents reporting that the birth of their first child was when they got their lives together.

The effects of parenthood on arrest risk, it turns out, are large. For both mothers and fathers, risk of arrest declines across the four categories of offense—drug, alcohol, property, and “destruction.” To test whether these observed declines are more than just age, Masenkoff and Rose compare the decline among parents to the risk of arrest at the same time among future parents, who act effectively as a control group. The results are quite stark: having a child reduces mothers’ risk of arrest by about 50% and reduces fathers’ risk about 20% (although risk for domestic violence goes up significantly among men following birth). 

Childbirth, in other words, shifts women’s age-crime profile to that of women who have their first child two to three years later, and shifts fathers’ profile to men who have their first child one to two years later. The motherhood reduction, Masenkoff and Rose note, is on par with the deterrent effects of doubling police funding or the prison population—a massive change.

These effects, the paper then argues, are uniquely attributable to the dispositional change that parenthood causes. They note that these effects show up for both men and women, and persist even after the child is born, indicating that the effect is not driven by the physical impact of pregnancy. And they are more apparent in first childbirth than second, showing that the effect is a function of the state of change from non-parent to parent, rather than the resource constraints imposed by childrearing (which scale with the number of children).

More controversially, they find that the effect is not due to marriage cooccurring with childbirth. They find that the effects are driven overwhelmingly by unmarried parents—married parents, starting from a much lower base rate, see no durable reduction. More surprisingly, they find that declines in arrest precede marriage generally, not the other way around, suggesting that (within their sample at least), people get married after they have otherwise “settled down,” rather than settling down because they are married. As the paper puts it, “marriage itself marks that end of a long-period of desistance rather than a turning point for criminal behavior.”

It's possible that marriage has a causal effect on risk for crime outside of the paper’s sample. We know, for example, that testosterone falls after marriage, reducing aggression and therefore propensity to offend. And we know that a peer’s marital status—e.g. getting a divorce—can affect your marital status, suggesting that individual marriage can affect community norms. (No effect on propensity to offend, incidentally, does not mean marriage does not have a lot of other benefits.) But in another sense, the paper’s findings simply reinforce the urgency of marriage in the context of childrearing.

It’s notable, after all, that the effects in the paper are overwhelmingly driven by the unmarried, while married parents see little effect of childbirth on their propensity to offend. This is likely because married parents have already “gotten their lives together” in advance of the baby, even if the baby was unplanned—their turning point has already happened. For unmarried parents, by contrast, the “turning point” comes with the news and birth of their first child. Unmarried parents are playing a sort of catch-up, risky-behavior wise, as their risk of offending falls towards the previously lower level of the married.

While childbirth can encourage men and women to desist from criminal offending, in other words, that doesn’t mean it’s a particularly good solution to the problem of antisocial conduct. Taking the steps culminating in marriage before having a child puts individuals at lower risk of offending to begin with—creating, thereby, a better environment for children to thrive in.

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