John Cochrane has a new post on his site that is quite dismissive of the ideas of Richard Thaler. Thaler is the famous behavioralist who has been getting quite a bit of airtime due to an excellent new book he has out. Thaler wrote a NY Times piece a few weeks back which Cochrane called a “straw man” because he says that all economists understand the limits of the simplifications in their own models. Cochrane’s view, ironically, looks like a bit of Dunning Kruger at work!
In one story Thaler explains how his students used to complain about their grades when based on a 100 point scale because the average score was just 72. But by changing the grading scale to 137 the average score rose to 96 and the complaints seemed to stop. Although this is the same score of 72% the number 96 felt better to the students. A totally irrational response that is consistent with the way humans respond to many things in life (including the way the economy works). Cochrane dismisses stories like this one saying that the conclusions are “small potatoes” in the grand scheme of things. He goes on to basically dismiss Thaler’s entire work as nothing all that relevant. Pretty bold claims to say the least.
Anyhow, Cochrane says that economists simplify lots of things for the purpose of experimentation which explains why their models don’t always perfectly reflect reality. And he’s right. But only up to a point. See, the thing is, many economic models aren’t just simplifications. They are alternate realities. For instance, before the financial crisis, it was pretty rare for economists to include the financial sector in the monetary economy because this sector was viewed as primarily being intermediaries in the real economy. Of course, we all know this wasn’t just a minor oversight now. It was a catastrophic simplification.
Cochrane rants about Thaler’s lack of showing a “monumental and devastating prediction” – well, there it is. That little tiny financial crisis we went through where economists consistently said the financial sector wasn’t a big concern and then, when it did prove to be a big concern, responded by totally botching the potential outcomes from the recession’s largest policy program, Quantitative Easing, because many of them misunderstand how banking works. I don’t expect economists to be able to predict things like recessions or financial crises, but a model that properly embedded the financial sector would have had a much better chance of identifying the substantial risks in addition to better predicting how QE might impact the economy.
And therein lies the problem. When good scientists perform experiments in a vacuum (a simplified version of reality) they are able to take those experiments and test them more accurately. But the results, no matter how well they work in the vacuum, must still work in reality in order for them to be useful in real life. And that’s the problem with a lot of economics. It’s performed in a vacuum that never gets translated to the real world. We could call this the “vacuum problem”. And because of this vacuum problem a good deal of economic theory and modelling just doesn’t reflect the way things actually are. This leads to a world that isn’t just simplified, but literally an alternative reality. I understand the need to simplify things. I do it all the time. But the conclusions reached from those simplifications still need to reflect reality. Otherwise, they’re only useful in the realm of the theoretical.