Do longer prison sentences lead to a situation where we have more hardened criminals due to a longer stay in chokey or do they lead to a situation where criminals are ‘scared straight’? The reason I’m asking myself this question is that I was reading Super Crunchers by Ian Ayre the other night and the subject cropped up. He discusses the use of regressions in finding answers to all kinds of questions – including the extent to which social security payments disincentivise people from finding employment (the answer? not nearly as much as people thought, apparently), and the one I’m looking at here. Ayre wrote that the findings of the study into recidivism and sentencing were that, all in all, the length of a sentence had pretty much bugger-all effect on the likelihood of a convict reoffending once released from prison. I had a quick google today and found a study by Gendreau, Goggin and Cullen (which is not the one referred to by Ayre). They found that:
None of the analysis conducted produced any evidence that prison sentences reduce recidivism. Indeed, combining the data from the more vs. less and incarceration vs. community groupings resulted in 4% (f) and 2% (z±) increases in recidivism.
In addition, the results provided no support for three other hypotheses. The prediction that recidivism rates correlate with sentence length in a U-shaped fashion was not supported. The view that only lower risk offenders would be deterred by prison sentences was also not confirmed. The lower risk group who spent more time in prison had higher recidivism rates.
The Gendreau, Goggin and Cullen paper is online here. It doesn’t tally with the findings of the study discussed by Ayre (which found no difference).
The one I’m interested in looking at had an intriguing method for working out the effect of harsh sentencing. They assumed that longer sentences would be given to the ‘worst’ offenders and that therefore simply comparing recidivism rates of longer-term against shorter-term prisoners would be hamstrung by this. Instead, they compared recidivism rates of persons convicted by hanging judges with those convicted by bleeding hearts. They found that longer sentences were not associated with an increased likelihood of reoffending due to their [relatively] long incarceration (so no ‘hardening of criminals’ effect was seen), but that, equally, longer sentences were not associated with a reduction in rates of recidivism. According to this study, then: length of sentence has no particular effect in terms of reducing rates reoffending nor in terms of ‘hardening’ criminals.
The study referred to by Ayre was conducted by Joel Waldfogel. A similar study appears here, written by Danton Asher Berube and Donald P. Green. The authors “find little evidence that incarceration reduces the probability of recidivism, as would be expected based on specific deterrence and incapacitation hypotheses”.
I have been interested while reading Super Cruncher to see just what people will test. Another point of interest in the same chapter of this book that deals with sentencing and recidivism is a note that Senator Moynihan introduced a change in the law that meant states could experiment with new ideas only if they had a control group. Brilliant – a politician introducing a regulation that only allows trials when they have a control group. I can’t imagine anyone from Durham County Council coming up with an idea like that.
Incidentally, the message I took from these studies on longer sentencing was that it seems to be useless in terms of prevention of crimes on release. There are other reasons to lock people up for longer – affording society a longer period of peace (if a menace to society is locked up for 10 years rather than five, I guess the public are protected from them for a longer period of time) or for reasons of punishment. The Canadian study I linked to earlier includes this in the discussion: “the primary justification for use of prisons is incapacitation and retribution, both of which come with a “price”, if prisons are used injudiciously”. If anyone has any contradictory evidence or (even better) a systematic review of studies into recidivism and length of sentence then please feel free to leave a comment linking to it.
Edit: from the book, re the rationale for choosing to look at hanging judges versus bleeding hearts – “Ten-year inmates might have a higher recidivism rate – not because prison hardened them, but because they were worse guys to begin with. Waldfogel’s randomisation insight provided a way around this problem. Why not look at the recidivism rates of criminals sentenced by different judges? Since the judges see the same types of criminals, differences in the judges’ recidivism rates must be attributable to disparities in the judges’ sentencing”. Berube and Green also get a mention and a study by Jeff Kling is referred to. Kling found that post-release earnings of people sentenced by hanging judges were “not statistically different from those sentenced by the judicial bleeding hearts” (a convict’s earnings after prison are, apparently, a strong indicator of recidivism). This study is available as a pdf here. There’s some abstracts of Waldfogel’s papers here: Index.
Comments, corrections and criticism are welcome as always. In fact, if anyone prefers links to open in the same window then tell me now. I’m never sure which is better – same window or new window?