Tuesday 29 November 2011

Christina Romer shifts my priors

 ... a speech on estimating the economic effects of austerity and/or expansion. Her conclusion: there is solid evidence that fiscal stimulus works, and that austerity is painful. Not something I' naturally am inclined to believe, but when the evidence is gathered so carefully, it shifts my priors.

Dodgy science

Alex Tabarrok spanks the BMJ over an article on the brain drain.

The BMJ and Lancet are publishing a lot of social science these days. The slant is usually Left-ish; more importantly, the methodology is often weak. When did social science overtake medical science in analytical rigour?

Monday 21 November 2011

Blame the economists!




It is widely thought that the financial crisis shows up the flaws of neoclassical economists, who were so immersed in nerdy mathematical theory that they forgot about the real world.

However, many economists did predict aspects of the crisis. Not all were marginal voices, either. Some were very distinguished figures from the mainstream.

Here is Marty Feldstein in 1997 (ungated link), arguing that the Euro is an economic liability.

Here are Obstfeld and Rogoff in 2005 (ungated link), worrying about opaque networks of cross-holdings and counterparty risk (in the context of the US current account deficit).

Economists probably are not blameless, and certainly there were naïve optimists – I read a fabulous article recently, from about 2005, which said [re the strength of the dollar, but it seems to typify that era] “the markets can be wrong, but they can't be wrong for a decade”. But as a group, macroeconomists probably had a clearer and earlier sense of the dangers than anyone else. The problem is that their knowledge did not percolate to the wider public.

This seems like a growing problem. Social science has been getting better and stronger in the past two decades – we have rigorous techniques for untangling multiple causes, ever better datasets and now the possibility to integrate field and laboratory experiments into our theory. But very little of this knowledge has spread to the world in general. Popular social science (Freakonomics or The Tipping Point) has been in fashion recently, but we are still far from the point where people turn to social scientists for ideas about what is going on. For example, think of the referendum on the Alternative Vote in the UK this May. There is a lot of research on how electoral systems shape political and economic outcomes – not enough to decide definitively either way, but certainly relevant evidence. This research never got mentioned in the debate. (There was some very simple stuff, like simulations of how the 2010 election would have gone, but nothing of modern research in political economy.)

I am not sure how this could be fixed. We definitely need more Levitts and Harfords. Perhaps we need more social scientists blogging.