Monday, November 26, 2012

Ignorance as an excuse

Via Greg Mankiw I read a response to the argument by Peter Diamond and Emmanuel Saez that the top marginal tax rate in the US should be raised to about 73%. I do not want to enter into the substance of the debate but instead discuss the logic used by the criticism of Diamond and Saez work.

The authors of the response present a contrast between the willingness of Peter Diamond to offer a concrete policy recommendation with the answers that two other Nobel Prize winners (Tom Sargent and Chris Sims) gave after receiving their prize. When asked in 2011 what should the government do to help growth, Sims answered:

"I think part of the point of this prize in the area that we work in is that answers to questions like that require careful thinking, a lot of data analysis, and that the answers are not likely to be simple. So that asking Tom [Sargent] and me for answers off the top of our heads to these questions — you shouldn’t expect much from us."

And when asked for a specific policy conclusion he added:

"If I had a simple answer, I would have been spreading it around the world."

The authors praised Sims' answers as the "model of how academic economists should behave when facing questions about specific policy." I find this to be an odd and depressing conclusion as it portrays a very pessimistic view on what academic economists should (can) do. I understand that some of what we do as academics is not useful enough for policy makers, and in these circumstances is better to be honest and stay out of the debate. But as Sims points out in his statement, one can find answers to those questions after careful thinking and a lot of data analysis. That's what Diamond and Saez have done. One can disagree with their analysis but one cannot simply disregard it because they are academic economists.

The authors of the response then go into the details of the analysis and they present their arguments on why the calculations of Diamond and Saez are wrong. But when presenting their arguments they protect themselves from criticism by saying that they do not have an answer to this policy question: "if we had a simple answer we would be spreading it around the world". Although "they can be pretty sure that the answer if significantly lower than 73%". Isn't this a simple answer? Or maybe the fact that they are not willing to provide a number but a range ("significantly lower than 73%") makes it a complex answer?

There is no simple answer to the question of what should be the top marginal tax rate. But policy makers need to chose a number, not a range. Diamond and Saez presents their arguments and data analysis in a way that is at least as competent as any other analysis on the same subject. They can be criticized on their assumptions or calculations but not on their willingness to advance the knowledge on an issue of great policy relevance. If any, they should be praised as academics who want to go beyond writing great papers to make those papers useful for policy makers or society at large.

Antonio Fatás