"Across Europe and the US, crime has fallen, regardless of which states lock up a high or low proportion of offenders."
Polly Toynbee, The Guardian, 23 October 2012
The question of whether or not 'prison works' is one that has been asked for almost as long as criminals have found themselves behind bars. Full Fact has itself looked at the issue from a number of different angles in the past.
However following Prime Minister David Cameron's speech on criminal justice earlier this week, during which he promised a "tough but intelligent" approach to tackling crime, the effectiveness of a custodial sentence as a tool for lowering crime rates has again been called into question by some.
Writing in the Guardian yesterday, Polly Toynbee claimed that crime rates in Europe and the United States had fallen regardless of how often spells behind bars were being issued by the courts. So is there really no correlation between crime and custodial punishment?
As we've seen in the past, gathering comparable data on crime and prisons across a number of different countries and criminal justice systems isn't easily done.
Eurostat - the statistical office of the European Union - does have figures available on the number of people in prison across Europe (and also the USA), as well as the number of crimes recorded in each country.
From this we can get a rough-and-ready idea of which countries have seen increases and decreases in each metric, and whether or not there is any correlation between crime rates and prison numbers.
[Data for small countries has been excluded, however the line of best fit does include Switzerland and Turkey who have anomalously high changes in recorded crime, and so should be interpreted with this in mind]
The first thing to note is that, by this measure, crime hasn't uniformly fallen across Europe and the US. In fact, of the 31 countries considered (smaller countries were excluded because of their volatility), 14 actually saw increases in the number of crimes recorded, including some of the EU's larger countries, such as Italy and Spain.
We do however need to be cautious in how we interpret this data. As we've shown before, analyses of recorded crime over time are hamstrung by the changes that are made in the ways in which offences are recorded.
In the UK alone, the methodology has been changed on several occasions, notably in 2002 and 1997. This means that the Home Office has warned that trend analysis shouldn't be undertaken across these divides. Once we factor in the differences between countries as well, this makes any comparison all the more complicated.
When it comes to the numbers behind bars in Europe and the US, we can see that the majority of countries have locked up more criminals - 23 increased their prison population, compared to eight that saw a decrease between 2000 and 2009.
Of the eight that did see a reduction in the number of people incarcerated, only two saw increases in crime rates (Portugal and Latvia) whereas the other six saw drops.
Of course in isolation this doesn't tell us a great deal: rising crime could be the stimulus to rising prison populations, or 'tougher' sentencing could lead to reductions in crime. We don't know which is the causal driver.
Ms Toynbee has however asserted something slightly different: that crime has fallen regardless of whether or not more criminals have been locked up.
Internationally comparable figures on the proportion of offenders given custodial sentences is not easy to come by, so we got in touch with the International Centre for Prison Studies, who pointed us towards the European Sourcebook of Crime and Criminal Justice Statistics.
The fourth edition, published in 2010, compiles data from 2006 that shows what proportion of punishments handed out to convicted offenders in that year resulted in a spell in custody (table 18.104.22.168).
Again, we can compare this to recorded crime to see what relationship there might be between the two:
From this we can see that of the 13 countries which saw falling crime levels between 2000 and 2009 (treating the three UK jurisdictions separately), 11 locked up a smaller proportion of offenders than the European average, while two - Lithuania and Romania - were above it.
We need to be a little cautious in interpreting this. Differences in the law between countries might mean that courts in some jursidictions may deal with a higher share of 'serious' criminals than those in other countries.
On the face of it however, these figures would lend support to Ms Toynbee's claim that a tougher stance on prison doesn't go hand-in-hand with falling crime, and may actually suggest that the reverse is true.
Interestingly however, if we look at how different countries' use of custodial sentencing has changed over the 2000-2009 period, the data suggests that most countries that have seen a fall in crime have also taken a 'tougher' stance on sending criminals to prison (using the number of prisoners per 1,000 crimes recorded as a proxy for the proportion of offenders that are given custodial sentences).
Of the 17 countries to have seen falls in crime levels, only two - Estonia and Romania - have locked up fewer criminals between 2000-2009. Most others, including the home nations,France, Germany and the United States, have taken 'tougher' stances on custodial sentences over the period. This might suggest that there is some correlation between the proportion of offenders locked up and the number of crimes committed.
So what can we conclude from this data? Looked at in different ways and at different points in time suggests seemingly different trends.
Academics that have looked at the topic have also noted that there is little by way of a pattern here. The respected crimonologist Tapio Lappi-Seppala studied the purported link between the prevalence of prison sentences and the level of crime for the International Centre for Prison Studies and concluded:
"The development of prisoner rates in 1980-2005 showed no consistent patterns with total recorded crime. In different times different countries showed different patterns. These results are in agreement with the conclusion from previous literature which states that differences in the use of imprisonment cannot be explained by the level and trends in criminality."
It is notoriously difficult to get internationally comparable crime and prison data, and more difficult still to trace trends over time, given the inherent differences in how crime is recorded.
However using the data that is available to us, there doesn't seem to be any simple link between imprisonment and crime levels, as Polly Toynbee suggests. For example, while most countries that have seen falls in crime in Europe have below average use of custodial sentences, the majority have also taken 'tougher' stances on prison over the same period.
However it isn't strictly true to say that crime has fallen uniformly "across Europe and the US." if we're dealing with recorded crime. While this is true for the UK, the USA, Germany and France, among others, a significant number of countries - including larger ones such as Italy and Spain - have seen rises in recorded crime over the past decade.