Tuesday, February 4, 2014

The conservative bias of economic models

Chris House has an interesting post on why economists tend to be more conservative than academics in other fields. His argument is that many economic facts have a conservative bias (e.g. when you look for evidence on politically loaded questions such as the effects of the minimum wage or high taxes).

Noah Smith replies and argues that there is no such bias in economic facts, quite the opposite, the evidence is weak for some of these questions. So where is the bias coming from? My sense is that it is coming from models to which most economists feel very attached to. And this is partly the case because these models and their economic logic is the one that allows us to show the value that our profession adds. Let me explain.

Economists do not have a great reputation when it comes to forecasting. Given the number of jokes about the inability of economists to forecast anything, about how we assume away any interesting real world phenomena, many macroeconomists stay away from practical questions about policy [As an example, when Chris Sims was awarded the 2011 Nobel Prize in economics, he was very reluctant about giving policy recommendations to get out of the economic crisis.]

But even if you avoid those difficult empirical questions you can still impress your friends by showing how the logic of simple economic concepts can go very far in understanding complex real world phenomena. Talking about opportunity cost, the difference between nominal and real variables, introducing a general equilibrium effect in a macroeconomic question or using game theory to help improve the analysis of a strategic situation are examples where economists thrive and show the added value of the discipline.

And in this spirit, when someone asks us why don't we raise wages to increase demand we use our economic logic to explain that this does not work, that increases in wages will do the opposite by reducing employment and GDP because of the effects on labor costs. And when someone suggests that the central bank should print money and distribute it to every household so that we are all richer, we explain to them that this will only create inflation. This is the origin of the conservative bias of economists, the conservative nature of the benchmark model that economists use. Of course we also have models where increases in wages could lead to higher GDP and there are situations where large increases in the money supply could not generate much inflation but those are much more complex models and they represent an exceptions to the benchmark model that economists tend to start with when they teach or explain their subject to others.

And what is interesting is that this bias does not always show in casual conversations. Many macroeconomists do not use the intuition of the benchmark model when having a casual conversation about economic policy. For example, as Paul Krugman argues today in his blog, most economists share the view that during the current crisis monetary and fiscal policy should have been more aggressive. Similarly, in casual conversations and when confronted with questions about efficiency of financial markets or labor markets or health care economists tend to describe a world full of inefficiencies and where government intervention makes sense. But it is rare that you hear the same strong message when you read their research or when they speak about it in public. Paul Krugman is right in the sense that his views are in many ways mainstream today but it is the fact that he expresses them in such an open manner that sets him aside from many other macroeconomists. I can see why some feel uncomfortable about making such strong policy recommendations given the uncertainty we have about our models or the evidence. I personally find it very refreshing to see more academics being willing to share their views to a broader audience through blogs or articles in a way that enriches the economic debate (even if, unfortunately, it might not have such a big impact on economic policy).

Antonio Fatás