When Great Economists go Wrong. A tale in 2 parts

I want to say right up front, before we get going that Dani Rodrik and Greg Mankiw are excellent economists with records far better than mine will ever be. I have learned a lot from them and respect them and their work.

Yet, each has, I think, made a mistake recently. Dani in his "the NY Times doesn't get it on Trade" post on his blog, and Greg in his NY Times column on fighting recessions.

First Rodrik. Dani is not a fan of expanding or even perhaps maintaining the current world trade regime. When the Times indirectly editorialized in its favor, he responded with:

"Here's what's wrong with this argument:

1. It automatically equates any desire to reconsider trade agreements and take a breather on new agreements as "protectionist."

2. It fails to recognize the ways in which technology and globalization interact to contribute to unequalizing trends in incomes, taking refuge in the defensive statement that "There is scant evidence that trade has played a big role in holding down typical workers’ wages."

3. It follows up this statement with "There is abundant evidence that it has contributed substantially to America’s overall economic growth," ignoring what every student of trade learns, which is that large gains from trade are possible only if there are also large amounts of income redistribution."

Ever so humbly, here is what I think is wrong with Dani's argument. First and most easily, it IS overwhelmingly likely politically that reconsidering existing agreements will turn out to create more protectionism. Secondly, Rodrik gives the impression that there is a single trade theory that has not been convincingly rejected by the data that one can use to provide reliable answers to policy questions. I am pretty sure this is not true. As to "what every student of trade learns", I would have to say they learn that we have no reliable theory that has not had its brains beat out by the data. Thus the theoretical "impossibility theorem" Rodrik sets up is not produced from a well functioning model and we are not required at all to accept it (see here for more on this point).

Now to Greg. In his column pleading to keep politicians out of the recession fighting business and letting the Fed do the job he says,

"In creating the Fed, Congress wisely made it a technocratic institution free of many of the political pressures that accompany other policy decisions in Washington. Subsequent experience in the United States and abroad confirms that more independent central banks lead to better economic outcomes. That’s why, in recent years, many nations have passed reforms to insulate central banks from politics."

Now I could beef about the "free of political pressures" part, but that would be tooting my own horn, so let me focus on the part about how "Subsequent experience in the United States and abroad confirms that more independent central banks lead to better economic outcomes."

First for the US, there is no before and after so we can't judge anything in that manner. Indeed most studies on the effect of Central Bank Independence use cross country, cross-sectional regressions, and the results, as Adam Posen has so cleverly shown, are far from supporting Mankiw's assertation. There is no empirical consensus on the question of the effects of CBI.

So, my complaint in both cases is about taking things that aren't necessarily so and using them to make political or policy points that don't necessarily follow. Even though I do admit that I actually agree with both Rodrik's and Mankiw's main arguments in their essays, I am not comfortable with the way they support them.

update: Rodrik link is mended. thanks to commentors for the heads up.