Articles What and where is the evidence that reducing prison population works?

What and where is the evidence that reducing prison population works?


Howard White of the Campbell Collaboration believes arguments put forward by the Howard League’s Commission on English Prisons Today are undermined by the fact they missed the opportunity to cite the rigorous evidence that reducing the prison population will not increase crime.   

Recently the Howard League’s Commission on English Prisons Today issued its report following a two year inquiry. The report calls for significant reductions in the prison population, replacing short prison sentences with community-based sentences.

I read the report from a ‘What works?’ perspective. What evidence do the Commissioners produce to support their argument that reducing the prison population will not increase crime, indeed may well reduce it?

Unfortunately they produce weak evidence. Their arguments do not pass muster as showing a causal relationship between imprisonment and re-offending. The report relies on correlations, expert opinion and turning a blind eye to uncomfortable facts. All of which is a shame, as there is rigorous evidence to support their position. But apparently the two year review did not find time to include it.

Much emphasis is based on the fact the several countries, most notably Finland, have substantially reduced their prison populations at times of falling crime. But this is a correlation. Maybe other factors explain the correlation. Perhaps liberal governments also have good social programmes and it is these, not falling prison population, which explain the reduction in crime. But we also often find correlations in data which mean nothing – the Oxford economist David Hendry famously showed how annual rainfall affected inflation! The best we can say from this evidence is that decreasing the prison population has not automatically led to an increase in crime.

At the same time the authors ignore correlations which don’t suit their position. We are told that the prison population has been rising in the UK since the 1980s despite the fact that crime has been falling since the 1990s. Hang on a minute, ‘despite’. Why not ‘and so’? It shouldn’t be ‘and so’.  Correlation is not causation. But the authors should not be using correlations on one page but not on the next when it doesn’t suit their purpose.

The Commissioners ignored other evidence which doesn’t suit them. We are told to follow Scotland’s bold example in cutting prison populations. That decision was made in 2008, since when the prison population has continued to rise. So are there lessons to be learned from the barriers to successful implementation of the policy? No! Rather than ‘waiting for the lessons’ we should ‘seize the initiative’.

They also collected ‘a great deal of evidence’ from visits to several countries including Finland when they heard from experts that imprisoning too many people was wrong. So people who think like they think told them they are right to think what they think. And this is evidence? It is true that many reports rely on expert opinion. But experts used to apply leeches, carry out bloodletting and tell us that the earth is flat. We should rely on experts to give us the best evidence on the matter on hand, not state their opinions. And each of every one of us should learn to critically appraise that evidence.

And, as I said above, there is good evidence. There have been several randomized controlled trials and natural experiments which allow us to compare reoffending rates between similar offenders given custodial and non-custodial sentences. A review of this evidence published by the Campbell Collaboration shows that there is no difference, both are equally (in)effective. So the Commission are right that reducing the prison population won’t increase crime. But they are not right for the right reason.

There is also evidence that some prison regimes are actually detrimental. Another review published by the Campbell Collaboration summarizes evidence from nine studies of scared straight programmes intended to straighten out youth by exposure to prison life. But they do the opposite.  Youth taking part in the programme are more likely to turn to a life of crime. Another review of boot camps for young offenders found them to be ineffective.

So, the Commission is right in its conclusions. But wrong in its argument. It is to be hoped that in future commissions setting out to examine a policy area will pay attention to the quality of evidence they will consider at the outset. They could start with an evidence and gap map. And then commission either new studies or reviews of existing evidence depending on what is out there already. Only once we move beyond correlation and anecdote will we really know what works so be better able to shape better policies for a better future.


Views expressed are the author’s own and do not necessarily represent those of the Alliance for Useful Evidence. Join us (it’s free and open to all) and find out more about the how we champion the use of evidence in social policy and practice.