Do long prison sentences work? Economics has the answer
Sentences deter different groups in different ways – and behavioural economics shows how, says Paul Ormerod
Does prison deter crime? The recent riots have rather left the government facing both ways on the matter.
On the one hand, the Starmer administration has not just gone along with the previous policy under which most inmates served just 50 per cent of their allotted sentences. They have reduced this to 40 per cent.
If prison works, the effect will be an increase in crime. This can hardly be the intention of ministers. So the implication is that they believe that prison has at best only a marginal impact in deterring crime.
In stark contrast, we see the lurid threats made against the rioters. Being charged with terrorism, given 10 years – on this view of the world, prison is a very effective deterrent.
There is in fact a large literature in economics on criminal deterrence which has some subtle policy implications. A major survey was published on this topic in the Journal of Economic Literature, one of the world’s top academic journals, in 2017 by Aaron Chalfin and Justin McCrary. It only covers important papers, but even so there are nearly 300 references used in the survey.
Nobel Laureate Gary Becker of the University of Chicago started it all off in 1968, in a paper with the splendid title of “Crime and Punishment”. For Becker, crime is a gamble. The potential criminal weighs up the various expected costs and benefits of getting involved in crime before making a decision.
One of the key features of Becker’s theoretical model is the attractiveness or otherwise of opportunities in the legitimate labour market. The more jobs there are locally and the better they are, the lower will be crime.
Becker’s paper sparked off a large number of empirical studies. And the 2017 survey article concludes that these give firm support to his theory. Lower unemployment and higher wages reduce crime.
This particular conclusion might not seem very surprising. But perhaps what is unexpected is that it comes from a body of work which is firmly based on rational, free-market thinking rather than on liberal do-goodism.
The question as to whether prison sentences deter other potential criminals is more complex.
The issue here is that the potential “cost” of committing a crime is a mixture of the chances of being caught and the potential fine or jail sentence.
But the latter brings in more complications. Imagine that you are offered the choice of being given £100 now or £110 in a year’s time. In this thought experiment, in order to remove the uncertainty about whether you would in fact be paid next year, imagine also that by some means or other you are certain to be paid it.
Some will take the £100 now. If next year’s amount is reduced to £105, more will do so. But if the choice is between £100 now and £100 next year, only the truly eccentric will opt for the latter.
Economists call this “discounting” the future. And behavioural economics shows clearly that less educated and skilled people in general place much more weight on the here and now than they do on the future compared to, say, graduates from Cambridge.
So the implication for the recent riots is that the most powerful deterrent is to increase the chances of being caught. The potential length of the sentence is less important. Catch them quickly and hold them on remand now.
Intriguingly, though, in order to deter, say, more middle-class groups of criminals, the implications from economics are that heavy sentences are the answer. They discount the future much less than the rioters, and so the “cost” of prison to them is much higher.
The strategy followed by the police regarding the riots has been exactly in line with the findings of economics.