Economists: too much ideology, too little craft

Paul Krugman argued yesterday that the belief in the need for new economic thinking after the financial crisis was incorrect, but led to some crazy but influential ideas that suited big money and the political right. While I agree with a lot of what Paul says, I would want to add something. These thoughts are strongly influenced by the fact that I’m in the middle of reading Dani Rodrik’s new book, called ‘Straight Talk on Trade’ (of which hopefully more in a later post).

In the preface to that book he tells the story of how 20 odd years ago he asked an economist to endorse a previous book of his called ‘Has Globalisation Gone Too Far?’. The economist said he couldn’t, not because he disagreed with anything in the book, but because he thought the book would “provide ammunition to the barbarians”. Dani Rodrik argues that this attitude is still commonplace. That attitude is, of course, both very political and very unscientific.

I suspect that something similar might have been going on before the financial crisis among economists working in finance. Paul Krugman is certainly correct that mainstream economics contained models that could explain much of why the GFC happened, so little new thinking was required in that sense. But one reason why so few mainstream economists used those models before the event owed at least something to an ideological aversion to regulation, and perhaps also not wanting to bite the hand that feeds you.

One of the features of mainstream economics today is the huge diversity of models that are around. Academic prestige tends to come to those who add to that number. But how do you decide which model to use when investigating a particular problem? The answer is by looking at evidence about applicability. That is not a trivial task because of the probabilistic and diverse nature of economic evidence, and Dani Rodrik describes that process as more of a craft than a science.

So, in the case of the GFC, good craft was in seeing that new methods of spreading risk were vulnerable to system wide events. Good craft was to see, if you had access to the data, that rapid increases in bank leverage should always be a concern. And more generally that arguments that ‘this time was different’ do not generally end well.

In my own discipline, I can think at least one area that should not have got off the ground if the craft of model selection had been applied well. RBC models were never going to describe business cycles because we know increases in unemployment in a downturn are involuntary. If you do not apply the craft well, then what can replace it is ideology, politics or simple groupthink. This is not just an issue for some individual economists, but can sometimes be a concern for the majority.