Tuesday 9 August 2011

The economics of rioting

Let me start by pointing out that I am not an expert on civil unrest. But given events in London (and other UK cities) I was interested to see what the urban economics literature had to say about the causes of riots. Initial impressions suggested 'not a lot', but a little more reflection suggests this conclusion was a little hasty. Let me try to spell out what I learnt and how it helps frame the debate about the causes of the London riots.

First, rioting is a complex issue. But the complexity is much more about the when, than the where and the who. In other words, its hard to predict when rioting will break out. But once it does it's easier to say things about what kind of people and neighbourhoods will be involved. [To clarify - in case this isn't blindingly obvious - I am talking about tendencies rather than suggesting that the police should be able to predict precisely where riots will occur and who will be involved with them a la minority report]

Second, the narrow economic approach to rioting would follow Becker and ask whether rioters are simply responding to the private benefits and costs of rioting. So rioting should be more common where the financial gains to rioting are high and the time costs to rioting are low (in terms of the opportunity cost of time and the probability of imprisonment). Consideration of the who and where suggests that these factors are at play in the London riots.

Third, it is often claimed that rioting is related to community level grievances. Again, this is something that we have seen with the London riots. Here, however, is where the economic analysis of the causes of riots get difficult. Because individual and neighbourhood deprivation tend to be very highly correlated it is very hard to distinguish between individual and neighbourhood explanations of socio-economic phenomena. With rioting, this general problem becomes even more difficult because there are several possible mechanisms at work. It could be that 'organisers' solve the coordination problem in specific neighbourhoods, or that peer pressure changes underlying norms (e.g. in attitudes to crime or responses to opportunities for crime) or that events in specific communities move them from no-riot to riot 'equilibrium'. With the London riots, it is quite possible that all of these mechanisms are at work. But it is equally possible that none of these mechanisms are at work and that individual factors are driving everything.

In summary, the economics literature tells us that separating out whether individual or community factors drive rioting is incredibly difficult. This suggests we should be very wary about believing anyone who claims to know otherwise.

[PS: For those of you who would like to read more, try Dipasquale and Glaeser's paper on the LA riots which formed the basis for the arguments that I laid out here].