Monday, October 31, 2011

Riots in Retrospect

Christian Davenport and Will Moore have a new podcast, where they discuss their views on media coverage of occasions of dissent/repression/human rights violations.  They do a very nice job with their first effort of talking about how they think about conflict and media coverage of it.  It is not jargony at all, but does a nice of applying some good social science ideas to the topics at hand.  They also provide a nice counter-balance since they tend to see dissent as a good thing (or at least not as inherently bad). 

The first one is on the London riots.  I just want to speak to one bit of the conclusion where they consider a prediction someone made about riots next occurring the US after the Arab Spring and after London.  CD suggests Detroit as a potential place to spark, given how badly it has done as a city.  I would suggest that a different way to approach it is to figure which cities have the worst records of police violence and police reputations.  The key spark in London was a guy (a black guy) who got shot by the cops, and the cops were less than responsive when the community sought more information.  The LA riots were not just in the aftermath of the Rodney King beating and the officers being acquitted--we learned later that police abuse--Rampart division in particular--was pretty extreme. 

So, if we had good data (which we probably do not have), we could predict (if I am right) where riots would be most likely.  Still, all it takes really is for one person to get shot by the police to raise questions about the legitimacy of the police and about who has authority.  When that happens, we tend to see anarchy (from the IR sense--absence of government) writ small--alliances, arms races, and conflict spirals.  It was the LA riots that initially inspired by adaptation of the Security Dilemma to Ethnic Conflict (Barry Posen published it first, but I had already been playing with the idea).

Anyhow, I look forward to future CD/WM podcasts, and so should you.


No comments: