Deterrence is contagious!
Civitas, 16 February 2011
Via Chris Dillow, we learn of another fascinating study by Francesco Drago and Roberto Galbiati based on Italy’s experience of its 2006 Collective Clemency Bill. It suggests that creating credible threats to return ex-prisoners released on license to prison if they re-offend does not just reduce their re-offending rate. It also reduces offending amongst their peers.
In 2006, with its prison system suffering from tremendous overcrowding, the Italian government enacted one of its occasional mass clemency bills, releasing around 40 per cent of offenders back into society long before their sentences were complete. Unsurprisingly, this led to a dramatic increase in crime. However, some features of the terms of clemency made it a useful opportunity to measure the specific effects of a prospective prison sentence on offenders. If re-convicted of any offence in the following five years, the released offenders were to be re-incarcerated for the residual of their sentences (as well as receiving a penalty for their subsequent offence). In other words, they were to complete the sentence they were originally given if caught reoffending.
A previous study suggested that this suspended sentence had a significant impact on the individuals released: the longer their residual sentence, the less likely they were to re-offend. What is interesting about this new study is that the crime reduction impact does not appear to have been limited to those individuals faced with those residual sentences. By looking at the nationality (for foreign-born prisoners) and place of residence for Italian prisoners, the researchers were able to establish peer groups for these offenders. These peer groups also had lower offending rates. In fact, a 1% reduction in offending amongst ex-prisoners subject to the clemency rules was associated with a 2% reduction in offending within their peer group.
This is an interesting result as it suggests some mechanisms through which prison, and the threat of prison, reduce crime. The relative absence of offending amongst peers seems to reduce other people’s disposition to offend themselves. Perhaps this is through a change in the communal perceptions regarding the relative costs and benefits of committing (if an ex-prisoner fears returning to jail, then so might their friends and acquaintances), or through a lack of offending behaviour to imitate. These dispositions to counter-act offending behaviour seem, in this case, to have spread through communities and localities rather than, for example, through knowledge of official Government policy or the national media.
This study also might represent a challenge to one argument against the efficacy of deterrence that was aired in the Oxford Journal of Legal Studies. Robinson and Darley claim:
‘the ability and the motivation to make the calculations required for deterrence can be influenced by a variety of contextual effects… Perhaps the most important of these stems from the facts that crimes are often committed by groups. When offenders commit crimes in street gangs, for instance, several effects can temporarily reduce the possible impact of a threatened future prison terms on current law-breaking activities: an ‘arousal effect’ leads to sprees and reduced sensitivity to risk, and an increase in the immediate rewards can arise from an increase in esteem in which the group holds the member who boldly breaks the law.’
While there is no doubt plenty of truth to this description of group crime, the logic of the scenario can actually run in the opposite direction. If some members (perhaps the potential ringleaders) of a gang are individually fearful of a prison term (because of the condition of their release from prison) then that might impact on the offending behaviour of the group, including those that do not have a particular reason to fear prison. If group perceptions impact on individual behaviour, then individual perceptions also seems to have an impact on the group’s behaviour.
This might suggest that far from being inefficacious, deterrence in some contexts has a multiple impact on group crime. Merely targeting some members of a peer group can be sufficient to reduce the criminal disposition of the collective. The power of this contagion effect of deterrence is clearly going to depend on individual circumstances (not least the specific individuals targeted and their status in the group) but what this new Italian study suggests is that the impact of deterrence does not end with the individual. That is just where it begins.