A National Bureau of Economic Research paper released early this year has caused a small commotion in economic circles. In the paper, titled “The China Shock: Learning from Labor Market Adjustment to Large Changes in Trade,” the authors, David Autor (Massachusetts Institute of Technology), David Dorn (University of Zurich), and Gordon Hanson (University of California, San Diego), calculate that the costs of adjustment to the shock of Chinese imports from the early 1990s and especially since 2001 (when China joined the World Trade Organization) have been enormous. Workers negatively affected by Chinese imports suffered job losses or permanently lower incomes.

Some economic commentators have interpreted the paper as an indictment of free trade. Others—including The Economist—read it as dealing with adjustment costs, not as a statement that trade brings no net benefits. Moreover, Autor et al. note that the China shock was a one-time affair that is now dampened by higher wages in China. Yet, writes Bloomberg​.com columnist Noah Smith, “Economists should still re-evaluate their benchmark theories, and ease up their adamant rhetoric of free trade.”

Is it possible that, contrary to what economists have thought at least since Adam Smith, free trade is bad?

One generally ignored caveat is that trade between China and America is far from free. Trade agreements are as much managed trade as free trade. As Paul Krugman concisely put it in a 1997 article, “The economist’s case for free trade is essentially a unilateral case: a country serves its own interests by pursuing free trade regardless of what other countries do” (quoted by Autor et al.). In America as elsewhere, we are very far from this.

Jobs and protectionism / Jobs have been lost in manufacturing and “offsetting employment gains in other industries have yet to materialize,” states the paper’s abstract. We should be very careful with this focus on jobs. The number of jobs should not be the metric. My rough estimates suggest that the mechanization of American agriculture has destroyed at least 15 million net jobs, but this is a plus, not a minus, since it was accompanied by lower food prices and higher real income. (See my Who Needs Jobs? Palgrave Macmillan, 2014.) Income and ultimately welfare are what matters, not sweating and work.

Standard economic theory proves that free trade creates net benefits. As economics students know, the geometric proof is beautiful and compelling, but the gist of it is not difficult to grasp. Consumers gain because they pay lower prices, and they gain more than domestic producers lose. If this were not the case, domestic producers could bribe consumers, through lower prices, to patronize their own products instead of their foreign competitors’. Moreover and less directly, free trade generates lower prices for many domestic producers’ inputs.

To be unqualified, this proof assumes that the welfare of a foreigner has the same weight as the welfare of a fellow citizen. It is a safe moral assumption to make, if only because one’s foreigner is the fellow citizen of another. You only need to have been born across a political border to be the “other.” Equal human value is also a safe political assumption if we don’t want the clash of national interests (which are the interests of the de facto rulers) to launch prosperity-destroying trade wars or real wars.

Even if one assumes instead that a national government should defend its nationals’ interests against foreigners’ interests, it is only in special circumstances—if it can manipulate the terms of trade—that it can succeed in doing so. And protectionism typically leads to retaliation, which amounts to the retaliator sinking a ship in his own harbor to get even with the trading partner who has done the same in his harbor. Protectionism is not very rational policy.

Mutual prosperity / The ultimate foundation of free trade rests on what economists call the benefits of exchange: both parties to an exchange benefit, otherwise one would have walked out. The same principle applies to a consumer (and his middleman such as Walmart) and his foreign supplier: both parties benefit.

It is true that third parties are sometimes harmed in a non-violent sense. If domestic consumers buy their widgets from a foreign supplier because they are less expensive, domestic competitors lose. But as we have seen, the losses of domestic producers and their employees will normally be more than compensated by the lower prices that consumers get. In a free trade regime, the individuals who lose in the labor (or capital) market will gain from purchasing their own consumption goods at lower prices because foreign suppliers compete with domestic suppliers. Free international trade generates more prosperity, and we would expect few to be net losers.

How do we know that free international trade creates general prosperity? Because the argument is exactly the same for free domestic trade. Domestic trade between, say, Vermont and California can cause disruptions and shocks. Taxi drivers in Vermont are certainly disrupted by Uber. Cheese manufacturers in Vermont are potential disrupters for their California competitors. Creative destruction is necessary for prosperity. This is no less true at the international level.

Autor et al. estimate that one million American manufacturing jobs were lost because of Chinese imports between 1999 and 2011. They add to this loss another 1.4 million jobs (which result, in part, from a questionable aggregate demand effect). These are viewed as short-term distributional and adjustment costs; they are not a net long-term efficiency cost. The authors acknowledge that these costs do not negate the net benefits of free trade. The benefits of free trade come out more clearly in an Econ​talk​.org podcast interview of Autor by Russ Roberts. In short, it is difficult for an economist to oppose free trade.

Market inflexibility / Why have these distributional costs (including adjustment costs) been higher than expected?

The reason is that labor markets have been much less flexible than we would have expected. Of the workers disrupted by Chinese imports, few switched industries or moved to another region. As noted by Autor et al., government assistance programs—Trade Adjustment Assistance and unemployment benefits, but mainly general assistance programs such as health and disability benefits—gave perverse incentives to workers displaced by Chinese imports. I think the regulation of labor markets should also be considered, from occupational licensure to minimum wages and European-flavored regulation of labor markets. Market inflexibility is the problem, not trade as such.

Shocks, creative destruction, and change are the bread and butter of prosperity. Consider again the example of agriculture. From 1950 to 1970 only, 3.7 million jobs disappeared from agriculture, but the labor market adjusted quite easily.

In the United States, as in other developed economies, manufacturing employment has been riding a long-term downward trend. The proportion of (non-farm) Americans occupied in manufacturing has gone from about 30% in the 1950s to 16% in 1990, and to less than 10% from 2006 on. The lost manufacturing jobs moved to services industries. Economists Andrew Bernard, Valerie Smeets, and Frederic Warzynski argue that, in many cases, the conception and distribution activities have stayed in the same firms, which have simply been reclassified in government statistics as non-manufacturing.

“We’ve always known,” writes economist Scott Sumner, “that local labor markets can be hit hard by import competition, and thus [Autor et al.] don’t really change our understanding of trade in any major way.” We must accept that any change, whether from trade or technical change, will generate some losers along with the many winners. We can never be sure that the former will be fully compensated out of the larger gains of the latter. However, both theory and history show that free markets are the best system to assure that net gains will be maximized and the number of net losers minimized.

I think that the policy conclusion to be drawn from Autor et al. and similar estimates is that government interventions—subsidies to disrupted workers and perhaps especially labor market regulations—have created rigidities that prevent the economy from adapting to shocks at minimum cost, as free markets would. This is the problem that needs to be addressed.

Readings

  • “Autor, Dorn, and Hanson on the China Shock,” by Scott Sumner. Econ​log​.org, Feb. 26, 2016.
  • “Rethinking Deindustrialization,” by Andrew B. Bernard, Valerie Smeets, and Frederic Warzynski. National Bureau of Economic Research Working Paper #22114, March 2016.
  • “The China Shock: Learning from Labor Market Adjustment to Large Changes in Trade,” by David Autor, David Dorn, and Gordon Hanson. National Bureau of Economic Research Working Paper #21906. January 2016.
  • “What Should Trade Negotiators Negotiate About? A Review Essay,” by Paul Krugman. Massachusetts Institute of Technology, 1997.