Must-Read: This does not seem to me to be terribly helpful. Dani Rodrik says, apropos of NAFTA:

  1. “The aggregate [distributional] effects were rather small…”
  2. “Impacts on directly affected communities were quite severe…”
  3. “Overall… a ‘welfare’ gain of 0.08% for the U.S. That is, eight-hundredth of 1 percent!…”
  4. “Trade volume impacts were much larger: a doubling of U.S. imports from Mexico…”

That is all true. But in today’s context it seems to me to be… a very partial truth… an even somewhat truthy truth… unhelpful. In my view he should, instead, have written:

  1. “NAFTA did very little to change any of the wages of unskilled labor, the wages of skilled labor, the rent of land, or the profits of capital in the U.S. The aggregate [distributional] effects were rather small…”
  2. “Impacts on directly affected communities were quite severe.” But these communities were quite small, and could have been substantially compensated–and would have been had American voters not elected Republican majorities in Congress in 1994 and thus installed Gingrich as Speaker and Dole as Majority Leader…”
  3. “Overall in an HOV or Ricardian model we calculate net a ‘welfare’ gain of 0.08% for the U.S. But we believe that there are not only “Ricardian” but substantial “Smithian” gains from trade. As a rule of thumb, each $1 of extra imports that the U.S. took in from NAFTA was paid for by products that before NAFTA had sold for some $0.75 to $0.80 cents–a good deal on average…”
  4. “U.S. imports from Mexico doubled from a low base: from about 1% to about 2% of GDP.

And, most important, Dani should not have allowed his readers to keep their illusion that NAFTA is responsible for any appreciable part of the decline in the manufacturing employment share:

Dani Rodrik: What Did NAFTA Really Do?: “The overall efficiency gains are quite small…

…the estimate that these overall gains amount to a “welfare” gain of 0.08% for the U.S. That is, eight-hundredth of 1 percent!… Trade volume impacts were much larger: a doubling of U.S. imports from Mexico…. What about the distributional impacts?… The aggregate effects were rather small (in line with other work), but that impacts on directly affected communities were quite severe…

As I wrote to Dani: The U.S. went from 30% of its nonfarm employees in manufacturing to 12% because of rapid growth in manufacturing productivity and limited demand, yes? The U.S. went from 12% to 9% because of stupid and destructive macro policies–the Reagan deficits, the strong-dollar policy pushed well past its sell-by date, too-tight monetary policy–that diverted it from its proper role as a net exporter of capital and finance to economies that need to be net sinks rather than net sources of the global flow of funds for investment, yes?

The U.S. went from 9% to 8.7% because of the extraordinarily rapid rise of China, yes?

The U.S. went from 8.7% to 8.6% because of NAFTA, yes?

And yet the American political system right now is blaming all, 100%, every piece of that decline from 30% to 8.6% and every problem that can be laid its door on brown people from Mexico.

By not making it clear that you are talking about 0.1%-points of a 21.4%-point phenomenon, I think you are enabling that. I don’t think this is a good thing to do…

And off in email I am right now grousing: And how do we convince Dani Rodrick not to be an idiot to be a positive force here? Smithian gains mean that the net benefits from trade deals are larger than the redistributions as they are calculated in a Ricardian or an HOV framework. And nobody should be allowed to say “NAFTA was a bad thing because the redistributions were huge relative to the net gains” without also saying “abrogating NAFTA would be an even worst thing because you get another batch of large redistributions, and this time the net gains are negative.”