Friday, August 17, 2012

Predicting food prices and conflict


I am generally a stickler for the peer review process. As a scientist I know the peer review process isn’t perfect, but it is a very effective way of weeding out nonsense. And on the topic of food, there is no shortage of nonsense out there.

That’s partly why I have been intrigued by some of the work coming out of the New England Complex Systems Institute (NECSI). Over the past couple of years, they have had reports on how food prices lead to climate and how speculation and ethanol can explain nearly all of recent price movements. And they have recently released a report on the effects of the current droughton food prices

These are all complex topics (which I’m guessing is what drew the Institute’s attention in the first place), and quantitative analysis can be very difficult -- the type of situation where reviews by peers can be especially useful. But they clearly have a different mode of doing science than I do, since they appear to not use the peer review process at all. Instead, they self issue reports, and do press releases on these reports that get wide media coverage.  It’s possible some versions of these reports are in peer-review somewhere, but I don’t see any mention of it on their website.

As I said, the peer review process isn’t perfect. It does not always help, and it is almost always slow. So I can understand reluctance to use it, especially when working on such topical issues. But it raises the question of how credible their work is. For example, I have gotten several questions from colleagues asking what I think of their work, questions that likely would not have occurred if the work had been peer reviewed. 

But in the case of NECSI, I think they have come up with a pretty satisfying solution – making testable predictions about the next year. For example, the figure below is from their most recent report, claiming that the drought should drive FAO’s food price index to about 240 by the end of the year. And they are already on record as saying these levels of food prices lead to large scale social unrest (they state a threshold of 210, so 240 is actually well above that). 


So by the end of the year, surely before most peer review processes would have been completed, they will have a clear test of their model. Now, it’s possible they could get things right for the wrong reasons – a broken clock is right twice a day. But they are going out on a limb, which is a way to establish credibility. Hopefully, they will be as honest and diligent in reporting their failures as their successes.

Speaking of which, FAO just released the new food price index for July 2012, which is 213. That’s 3 points above the threshold!

No comments:

Post a Comment