The prospect of parenthood makes people more law-abiding

Using data on more than 1m babies born in Washington state between 1996 and 2009, and records of thousands of crimes committed there between 1992 and 2015, the authors find that when women become pregnant, they are much less likely to be arrested, for a wide range of crimes. The effect is most marked for “economic” crimes, such as theft and burglary, but is also true of assaults, vandalism, and alcohol and drug offences. Arrest rates fall by 50%, almost as soon as women become pregnant and fall much further as the pregnancy goes on. Although they bounce back somewhat after childbirth, arrest rates stabilize at about half pre-pregnancy levels.

More surprisingly, the same pattern holds for fathers. Men are much likelier than women to commit crimes of all sorts in the first place, and the decline in some types of crime is less dramatic for dads than for mums. But arrest rates drop by around 15% once their partners become pregnant, and stay around this mark even after birth. In a blog post commenting on the paper, Alexander Tabarrok of George Mason University described the effect as “astoundingly large”. A study by Mr. Tabarrok published in 2007 concluded that the threat of an additional 20 years of prison made criminals 17% less likely to reoffend; the prospect of fatherhood, it seems, is more salutary than that of two decades of incarceration.

The arrest rate falls almost as soon as pregnancy begins. The comparison with prison is deliberately provocative, but it exposes an unfashionable possibility: people sometimes change after acquiring something worth protecting.

Politicians look active when they increase punishment. They get a number to raise and newspapers get a sentence containing “crackdown”. Identity, attachment and future responsibility are messier. The arrest data around prospective parenthood cannot prove a single clean cause. Arrest differs from offending. Pregnancy changes routines, relationships, health and exposure to situations where arrest might occur. People who become parents may differ from those who do not. Treating one association as a universal rehabilitation programme would be reckless.

The scale justifies a better question. What becomes newly valuable when someone expects a child? Perhaps freedom, health, income or being present. Consequences now concern a particular future date and relationship. Commercial behaviour change has a smaller version. A new supervisor begins documenting decisions because others depend on them. A first-time homeowner finally learns where the stopcock is. They had the instruction before; the new role made it relevant.

This is why endless warnings can underperform an identity people want to enact. “Avoid a penalty” asks someone to fear an institution. “Be the person your child can rely on” asks something more intimate—and can become manipulative very quickly if used without care or practical support.

A valued future cannot substitute for housing, treatment, money or justice. Those supports still have to exist. The programme designer writes the practical help beside the identity message and leaves the threat escalation out of the first draft.