Does prison reduce crime? Yes, obviously

Sometimes the common-sense solutions really are best

I’m giving my first important TedTalk to a crowd of international executives and global ‘thinkfluencers’. I’m promoting a new airport psychology book, which is filled with curious and quirky counter-intuitive stories about human behaviour — all the good stuff you can repeat over dinner and drinks to sound intelligent and plugged in.

Beginning with a historical anecdote to prove my point, I recall how in 2006 the Italian government released one-third of all its prisoners en masse – and a curious thing happened to its crime rate…

I take a pause, press for a new slide. The audience is curious. I wonder what happened, they ask themselves? Counter-intuitively, maybe crime went down. Perhaps there was no change because offending rates are all down to age cohort or employment and maybe they fixed it with this one neat trick.

Well, I say, what happened is that crime massively went up. Of course it did. Obviously letting loads of criminals out of jail is going to lead to loads more crime, something anyone could tell you. Well, almost anyone.

In my first ever Substack post I wrote about how clever counter-intuitive thinking characterised the early Covid debate about masks; although mask-wearing was widespread in east Asia, actually we were told, they are counter-intuitively not helpful in preventing infection. This was bandied about by experts — even by the United States surgeon general — despite being obviously untrue. A complete idiot could have told you that.

This kind of counter-intuition is also popular in thinking about crime. Actually, prison is not only punitive and barbaric, it can actually increase crime. It’s a university of crime! This argument was used fairly recently by Yale’s Jason Stanley in his best-selling How Fascism Works, the author stating that ‘studies indicate that incarceration itself contributes substantively to an increase in crime rates,’ because of subsequent unemployment. He quotes crime expert David Roodman and his line ‘more time in prison, more crime after prison’.

Stanley believes that anti-crime rhetoric is just fascism: ‘What has not accompanied this shift is an awareness that the underlying motivations for the hard-on-crime rhetoric and policies were fascist, set up to establish an us-versus-them dichotomy and reinforce preexisting hierarchical stereotypes.’

In fact, ‘fascist law-and-order rhetoric is explicitly meant to divide citizens into two classes: those of the chosen nation, who are lawful by nature, and those who are not, who are inherently lawless.’ He used the case of the Central Park Five in 1989 to illustrate the phenomenon of law-and-order moral panic, although personally if I was to argue that fear of crime was just people being hysterical, I would probably not use New York in the late Eighties as my starting point.

The university-of-crime argument has a nice counter-intuitive ring, because on a basic level it seems obvious that if someone is in jail, then unless they have been granted the God-given power of bilocation, they can’t be committing crime outside. The case against prison is not only counter-intuitive but, because putting people in jail is an unpleasant and unseemly task, it’s also a relief.

Yet even the progressive alternative of more ‘education’ seems to accept the primal, atavistic case for jail. I’ve even read the argument that, rather than relying on prison, we could reduce crime by raising the school age to 18, because there is evidence to show that crime rates drop during school hours. That is, when teenage boys are forced inside government buildings and not allowed out, they don’t commit crime outside – no way! It’s almost like secondary school is treated as a prison, but one which everyone has to suffer because society sees adolescent males as a nuisance.

Similarly, people’s real beliefs can often be inferred indirectly. Whenever a progressive politician wants to show how seriously they take racism or homophobia, one of the first proposals they come up will be stiffer sentences for hate crimes. But if longer prison reduces hate crimes, why don’t we apply it to non ‘hate’ crimes such as assault, GBH and murder?

As with mask wearing, the actual evidence seems to back up people’s intuitive beliefs. A very interesting new post by Substack author Mugwump looks at the effect of longer prison sentences on crime, and concludes that, while prison only has a limited impact as an individual deterrent, as a form of incapacitation it is effective. This is because a huge amount of crime is committed by a relatively tiny number of individuals who, once released, will repeatedly reoffend. Mugwump writes:

Thomas Mathiesen Prison on Trial famously concludes that incarceration does not “rehabilitate, deter or incapacitate”. I’m with him on the first, ambivalent about the second, and against him on the third. Here’s why. 

In 2012, Siddhartha Bandyopadhyay, a professor from the University of Birmingham made a series of requests under the Freedom of Information Act to 43 police forces, covering the period between 1992 and 2008. Utilising public information to come up with an average sentence, in combination with this sleuthed information, he was able to find the impact of the length of incarceration on local crime rates. He found:

· increasing the average sentence length for burglaries from 15.4 to 16.4 months would reduce burglaries in the subsequent year by 4,800; and 

· increasing sentences from 9.7 to 10.7 months for would result in a reduction of 4,700 offences a year. 

Given the total number of annual burglaries was 962,700 and annual fraud offences was 242,400, a mere 1 month increase sentences is clearly not a panacea. But Bandyopadhyay found something else that was interesting: if, instead of automatically releasing prisoners at the half-way point, we made them serve two-thirds of their sentences, burglaries would drop by 21,000, robberies by 2,600 and fraud by 11,000. That is much more significant, and gives an indication that longer sentences work in reducing crime.

Mugwump looks at the evidence around Italy’s insane ‘collective pardon’, where ‘Buonanno and Raphael’s most conservative regressions estimate that the mass release increased reported crime by 57 per 100,000 residents per month. Following this increase in crime, Italian prisons inevitably started to fill up again.’ Most crime is committed by a very small number of persistent criminals, whose impact on wider society is enormous. A Dutch study found that in 10 regions where judges were able to impose ‘enhanced’ sentences for prolific offenders, this reform led to a 25% drop in thefts. 

He points to various analyses backing this up, including a 2017 study by David Roodman. Mugwump writes: ‘Amusingly, Vox - referencing Roodman’s work - ran an article entitled “A massive review of the evidence shows letting people out of prison doesn’t increase crime”. A curious headline given Roodman’s actual findings (which include the conclusion that all the incapacitation-studies he looks support incapacitation effects).’

The prison as ‘university of crime’ effect also works in reverse, too. Not only does early release lead to an individual reoffending, but it also means his friends are more likely to reoffend, too. A study from last year found ‘that one additional peer incarcerated at the time of release decreases the probability of arrest (incarceration) by 2.4 (1.9) percentage points within the first-year post-release. For more narrow definitions of peers such as family members and former criminal partners, their findings suggest a more than 10 percentage point decrease in the probability of recidivism. The mechanism is allowing the offender to return and integrate into society without bad influences or criminal incentives.’

Prison works, and it doesn’t even need to be punitive. I’m not that interested in punishing criminals; I’m not opposed to luxury Norwegian-style prisons that compare well with the average twenty-something Londoners’ dwelling. I just believe that a relatively small number of men make life miserable for everyone else if allowed to roam free.

Placing our faith in the system’s ability to reform people is also naïve to the point of irresponsible. Recidivism is the norm in every society. A worldwide study of reoffending  found that ‘reported re-arrest rates were between 26% and 60%, reconviction rates ranged from 20% to 63%’, with those lower figures mainly found in very wealthy, rural Scandinavian countries.

So if a rapist or wife-beater or mugger or burglar is given a feeble sentence, there is a 20-63% chance that he will then do the same thing to someone else within a couple of years. Those are risks no one would impose on their loved ones – why make innocent strangers suffer?

In Britain there are countless cases of violent offenders being given minimal sentences, as a result of which blameless people lose their lives or suffer horrible injuries or trauma. I find the extent to which we just endure this very strange; there are countries where people don’t fatalistically accept that their bicycles will be stolen or houses broken into and the police will do nothing, where women walk around at night wearing headphones, where people don’t get randomly attacked with a crossbow in broad daylight.

These things aren’t inevitable, they are direct consequences of decisions made by individuals. Mugwump’s conclusion is that, while lawyers and judges get the blame, it is entirely the result of government policy.

That surely seems, true. Britain’s needless crimes are not just the result of clever counter-intuitive ideas gaining hold, but also a Conservative government that fails to build prisons because the Treasury is allergic to investing in anything. As with so many things – transport, for instance – this is a failure to see beyond the immediate bottom line. Not only does crime cause pain and misery, but active criminals are a huge financial strain on society, and spending money on new prisons would lead to savings elsewhere. Instead, after 12 years of Tory prime ministers in Number 10, all the ‘law and order party’ has given us are hundreds of closed police stations. As with so much else, their record on crime is one of almost total failure.