More abortion, less crime?

Inconvenient doesn’t mean untrue

John Donohoe and Stephen Levitt argue that Roe v Wade (1973), now the subject of a pending judgement by the US Supreme Court, led to an increase in abortion which in turn substantially reduced crime in the US. This effect comes, they say, because children born from unwanted pregnancies are more likely to commit crime later in life, so reducing unwanted pregnancies through abortion reduces crime.

The authors first made their case in a 2001 paper; it has received methodological criticism over the years which seems to have had the effect of strengthening it. The update was produced in 2019; the authors talk about it at length on this episode of Freakonomics podcast (Levitt is the co-author of the original Freakonomics book).

Donohoe and Levitt’s results seem to vary in their replicability across other countries. For example, Kahane et al argue that it does not hold in the UK. Different states have different histories of abortion, and of course different demographics and other social variables.

The authors are only too aware that neither side of the Roe v Wade debate finds their link between higher abortion and lower crime an attractive one. Supporters of the status quo in the US (at the time of writing) do not like the association with the bloodless utilitarian idea of crime reduction via abortion; supporters of returning abortion to pre-1973 rules do not like the association with higher crime. C’est la vie.

Donohoe and Levitt make the important point that the abortion-crime reduction link they propose does not impact upon the anti-Roe v Wade argument, insofar as if the foetus is given the same status as a living person then thousands of lives saved through reduced rates of serious crime are dwarfed by the those lost to increased abortion. And while the claimed abortion-crime reduction effect does on its face amount to a strong utilitarian argument going the other way, don’t expect to see it being employed anytime soon…