Analysts have thus had a long time to assess the effects of benefit sanctions, and one interesting insight comes from a 2004 paper by UCL's Stephen Machin with Olivier Marie. In the article, Machin and Marie use quasiexperimental methodology to compare crime rates in in areas before and after the introduction of JSA, which at the time was seen as introducing a more punitive benefits regime.
Machin and Marie's results are quite fascinating. They found that in areas where there were more sanctions there was also a corresponding rise in crime. Thus, they argue that "benefit cuts and sanctions embodied in the JSA appear to have induced individuals previously on the margins to engage in crime".
The Coalition (and the whole political class in general) is widely convinced that there needs to be ever greater sanctioning in the benefits system. The argument for this is generally split in two: (a) the threat of sanctions is a matter of fairness (you get a benefit, we expect you to look for work); and (b) economic incentives will encourage people to find work.
While (a) is generally a moral issue, the evidence on (b) - as Machin and Marie show - is far from clear. We all want to see more people in work, but when we try achieving this by pushing people to the extremes of economic hardship, we might just find some very unintended - and some very undesired - consequences.