Another missive from the economics profession to the Queen - I bet she's regretting ever having raised the issue.
In my view this version - put together by Geoffrey Hodgson, editor of the Journal of Institutional Economics - gets it right. The letter complains that 'economists has become largely transformed into a branch of applied mathematics, with little contact with the real world' and that economists missed the credit crunch because of 'the preference for mathematical technique over real-world substance (which) diverted many economists from looking at the whole picture'.
The letter pinpoints the questionable assumptions upon which economic models are mostly based - the rationality of most human behaviour, and the tendency of markets to operate efficiently - as the reason for mathematical modelling mostly missing the point. These are not, of course, new criticisms, but the fact remains that there are powerful ones, and that the mainstream economics taught in prestigious institutions tends pretty much to ignore them.
Unfortunately, a chunk of political science has been heading the same way in recent years, even though if anything the rationality assumption is even less likely to explain things in the political world. Of course, we don't have much to brag about either. Just like the economists missed an $8 trillion housing bubble, political scientists have failed to anticipate a host of big events, such as the collapse of Soviet communism or the ability of Islamist militants to launch a successful attack on the United States. But the lesson of economics is that most of the time mathematization is unlikely to improve our understanding of the world, especially if it is based on the crude assumptions of rational choice theory, whose empirical failings were beautifully documented 15 years ago by Green and Shapiro's Pathologies of Rational Choice Theory.
Pathologies was ignored by rat choice political scientists, who have continued down the economists' path of ever more elaborate formal models and ever more advanced statistical techniques. But the poverty of the basic formal reasoning (over-simplification and weak conceptualization) and the lack of good data on political phenomena (small samples, biased or unreliable data sources) leaves us pretty much where we were in 1994. Until I get a coherent rat choice story about why most people vote in Western democracies - ie a solution to the "voter's paradox" - I'm going to stick to inductive theorizing, the source of just about everything useful that poli sci has to offer.