Branko Milanovic
April 27, 2017

Graduate Center CUNY organized yesterday (April 26) a star-studded discussion of the effects of trade on US jobs and wages. The participants were David Autor and Ann Harrison, both authors of seminal economic papers on the effects of China’s imports on US employment and wages (and professors at respectively MIT and Wharton), Brad DeLong, a polymath and former high Treasury official in Clinton’s administration (and professor at Berkeley), and Paul Krugman, Nobel Prize winner, professor at the Graduate Center CUNY, and probably one of the three most influential economists in the world.

The discussion was started by David Autor who pointed out that although US manufacturing employment was on a downward trend since 1943 (when it accounted for almost 40% of the labor force), what happened in the early 2000s with China’s accession to the WTO was a sharp (“off-the-cliff”) decrease in the number of manufacturing jobs. Between the late 1990s and today some 5 million manufacturing jobs were lost. While Autor thought that technological change was indeed a big factor explaining longer term trends the most recent drop cannot be dissociated from trade (import penetration from China).

Next, in the opening statements, Ann Harrison, referring to two very important papers which she coauthored, argued that the effects of trade are visible when one follows workers by occupation much more than when one looks simply at a given industry. Thus, people displaced by trade, even if they ultimately get reemployed, lose some 25% of their wage. The second paper of Ann Harrison carries a possibly slightly different message: many jobs were lost because capital goods became cheaper, and machines thus substituted for labor. It is a technology story with a twist: technological change responded to the change in relative prices. Harrison said that her position on the technology-trade (TT) debate is probably closer to the presumed position that Brad DeLong (the next speaker) would take than David Autor’s.

Indeed, in his opening statement DeLong supported the technology explanation although, since he made this point after a long detour through his family history, it was not too clear to me whether it still applied to the current situation and China in particular.

Paul Krugman (the statements were made in alphabetical order), referred to the similar TT discussions in the 1990s where the consensus was that some 2/3 or more of the job and wage effects were due to technology. But, Krugman said, economists might have been right on trade’s limited role then; yet keeping the same opinions now was wrong since the boost in trade that has occurred in the past 20 years was much greater than what the US witnessed in the 1990s. He thought that trade today has a massive impact but that it is one-off event, linked to China, and unlikely to be replicated.

If soccer-like I were to summarize the positions, I would say that Autor and Krugman were of the opinion that “trade matters” while Harrison and DeLong tended to favor technological explanation. But that classification is too rough. Autor’s own work, of which he briefly spoke later, shows huge importance of technology, especially in displacing routine labor. (My favorite paper of Autor, Dorn and Hanson is the one which pits the trade vs. technology stories against each other and finds that both…matter.) Similarly, as I mentioned, Harrison does find incontrovertible impact of trade too.

There were at least three areas where the panelists seemed to agree.

(1) Withdrawing from globalization would be extremely costly for the US and the world. DeLong pointed out to the asymmetric effects of NAFTA. While he argued that NAFTA (trade with Mexico) had a miniscule impact on US job loss, withdrawing from NAFTA today would have an enormous negative impact because of the number and density of trade links that have been established in the past two decades. Everybody agreed that it was madness to withdraw from globalization and to go back to protectionism.

(2) Everybody agreed that economists underestimated the impact of trade. As Autor put it, the benefits of trade (cheaper goods) are diffuse, but the costs are concentrated (aka you lose your job). Krugman thought that this was because economists, enamored by the “jewel in the crown of economics”, theory of comparative advantage, tend to look at average effects, not at the heterogeneity of effects. He thought that this was changing now.

On a more philosophical note, Autor added that workers are people (yes) and that even if the compensating mechanisms for job loss were effective (Ann Harrison said they were not), people desire to have meaningful jobs and high wages rather than to subsist on handouts.

(3) The China effect will not reoccur. As I mentioned, the point was made by Krugman in the opening statement, but was expanded on by both DeLong and Autor. DeLong thought that China will be the last example of export-driven development, made possible by the willingness of the US to be open to Asian imports (Japan and Korea before) and by the eagerness of the rest of the world to cover US deficits by squirreling the money in the United States. DeLong thus touched upon the global political economy issues which paradoxically made the richest economy in the world be capital importer rather than exporter. (Yet another example, in my opinion, of how with globalization many of the nostrums of the mid-century neoclassical economics got overturned.) But the one-off effect of China (no Indias, Ethiopias, Burmas waiting in the wings) means that the current trade effects on US labor are not going to be repeated.

It was a great evening. The top economists in perhaps the hottest area of economic policy dispute today duked it out in a very fair way, and came out to a fairly consensual position. We are back to 1817 where trade and technology (the famous Chapter XXXI) played such a great role in Ricardo’s “Principles”.

Branko Milanovic’s original post.