I’m a big fan of Chris Blattman’s blog (as the number of ‘hat tips’ – [h/t] – on this one demonstrates), but he lost it a bit in his recent post 
on food riots. Here’s what he says:
‘Globalization and growth should reduce price spikes in future. More countries are producing crops. Climate shocks in Argentina are not that tied to climate shocks in Russia or China, and so price volatility from supply shocks should be going down. Falling transport costs also mean that more substitutes are available, further reducing price volatility. So things should be getting better over time, not worse, especially if trade allows countries to diversify their diet. Envision a future of diminishing instability.’
This reminds me of the apocryphal French diplomat arguing in a Brussels punch-up ‘I can see it works in practice, but does it work in theory?’ Here’s the practice – you can clearly see food prices pretty smooth up til 2007, then going haywire.
Time to adjust the theory, rather than deny the reality, Chris?
Where I agree with him is that the triggers for food riots, as for most things, are often local rather than global – government policy, bad 
policing etc. And not just in developing countries – the incompetence of the Seattle police was in my mind largely responsible for the collapse of the WTO ministerial in 1999 – they had plenty of tear gas but no crash barriers (see pic). But it’s the interaction between local and global phenomena that we need to understand better – price rises in 2008 did coincide with a wave of food riots, no doubt all with their local contributory factors too. And to argue that this was mere coincidence isn’t very credible, in my view – A at least partly caused B. So yes, worry about local politics, but worry about the graph too.