Here. (HT - Tyler Cowen). This part is particularly good:
"Now that I'm a social scientist myself, or at least as close to being one as we manage to get in these early days of human civilisation, what do I think of Asimov's belief that we can, indeed, conquer that final frontier – that we can develop a social science that gives its acolytes a unique ability to understand and perhaps shape human destiny?
Well, on good days I do feel as if we're making progress in that direction. And as an economist I've been having a fair number of such good days lately.
I know that sounds like a strange claim to make when the actual management of the economy has been a total disaster. But hey, Hari Seldon didn't do his work by convincing the emperor to change his policies – he had to conceal his project under a false front and wait a thousand years for results. Now, there isn't, to my knowledge, a secret cabal of economists with a thousand-year plan to save our current civilisation (but then I wouldn't tell you if there was, would I?). But I've been struck these past several years by just how much power good economics has to make correct predictions that are very much at odds with popular prejudices and "common sense".
To take a not at all arbitrary example, a standard macroeconomic approach, the IS-LM model (don't ask) told us that under depression-type conditions like those we're experiencing, some of the usual rules would cease to apply: trillion-dollar budget deficits wouldn't drive up interest rates, huge increases in the money supply wouldn't cause runaway inflation. Economists who took that model seriously back in, say, early 2009 were ridiculed and lambasted for making such counterintuitive assertions. But their predictions came true. So yes, it's possible to have social science with the power to predict events and, maybe, to lead to a better future.
That said, it's a long way from getting the medium-term path of interest rates and inflation more or less right to predicting the overall course of civilisation centuries in advance. Asimov's psychohistory evidently integrates economics with political science and sociology, which are much harder subjects than economics – economics is, after all, largely about greed, while other social sciences have to deal with more complex emotions. There are wonderful, insightful political scientists and sociologists working today, but their fields have yet to develop even the (very limited) degree of intellectual integration that makes doing economics sometimes feel like we're living in at least the very early dawn of Hari Seldon's psychohistory.
But maybe those fields will come along too."
My single quibble is with his "largely about greed" line, but he's writing for a popular audience and he says "largely", so that's not such a big deal.