There is a new academic paper by David Autor, David Dorn and Gordon Hanson that challenges one of the bedrock beliefs of economics: free trade is good. Titled “The China Shock: Learning from Labor Market Adjustment to Large Changes in Trade,” the paper shows how workers in sectors more exposed to imports from China are suffering relative to other workers. This is not a new perspective at all. Indeed, the bedrock of economics is more like fresh cement compared to the ancient instinct to fear outsiders.
China Shock offers new insights, really tremendous coverage of many issues, and deserves to be taken seriously. It should not, however, be oversold (which I think Noah Smith does a bit here claiming that free trade with China wasn’t such a good idea). Even so, I think the China Shock is flawed, or more accurately, it is short-sighted. Here’s what it is missing, in my opinion:
First, theories against trade need a lot more than anecdotal, short-term empirics. And yes, when it comes to opening trade to 1/6 of the world’s population, there will be a shock before the system restabilizes that will take longer than 15 years to settle. Fifteen years is long enough to ask questions, but not to answer them.
In the parlance of econ 101, a phrase I actually dislike, there will be “winners and losers” from trade. Autor et al. are saying they found more than the expected negative impact on the losers. Nevertheless, my first impression is that their analysis of China’s integration via the WTO to the world economy is just one story, barely half told, and far too early to draw definitive conclusions. We are talking about a core principle based on centuries of history and hundreds of more compelling stories. In short, the anti-trade view tends to be short-sighted.
Second, the ultimate story that substantiates the value of free trade is the historical experience of the United States, specifically the internal growth dynamics after 1789 when the adoption of the U.S. Constitution expressly forbid trade barriers among the states themselves. “No State shall, without the Consent of the Congress, lay any Imposts or Duties on Imports or Exports…” says Article 1, Section 10, Clause 2. This is widely seen as one of the reasons the U.S. is the world’s richest large economy and has been so for over a century. Eugene Melder has a nice paper, published in 1940, on this. More recently, there’s this passage from the 1985 Economic Report of the President:
The power of free trade is amply demonstrated in history, including the early history of the United States. Under the Articles of Confederation, protectionist interests in individual States moved quickly to restrict the flow of competing products from other States. The debilitating effects of this protectionism on the States’ economies convinced the framers of the U.S. Constitution to forbid individual States from levying tariffs (and the Federal Government from levying export duties). Federal courts have guarded the integrity of this prohibition, ruling as recently as 1981, for example, that a Louisiana tax on natural gas passing through the State was unconstitutional. The constitutional ban on State tariffs was crucial to the development of the U.S. economy not only because it established a free-trade area
among the 13 original States, but also because it ensured that the free-trade area would expand automatically as new States joined the Union.
Would Autor et al. expect to find outsize winners if trade between California and the other 49 states were given a trade restricting policy shock? I am confident that a negative California shock would be rather harmful to workers in Arizona, Michigan, Ohio, and the rest. Indeed, until a scholar can convince the public that a single city should declare autarky for its own net benefit, I will look at the debate over trade with real bemusement.
Third, any general analysis of a trade shock has to consider consumption impact, not just production. Not to knock the supply side, but it’s a one-eyed view. The suppression of inflation during this recovery has been an absolute puzzle for the Federal Reserve. I find it hard to believe how the China Shock effect on pricing has been overlooked, not in this paper, just as a general observation. The natural question to ask is how the wage effect balances against the price effect for net well-being. It’s fair to counter that the authors examine real wages, which by definition are inflation adjusted, but is that good enough?
Real wages neglect the broad consumption gains from trade, especially in this era of rapid technological change, and are based on what many economists think is a flawed measure of inflation. The inability of prices to measure quality of life improvements, new products, and more is a major controversy.
Let’s not forget the timeline: 1978, China begins to adopt property rights & free markets. 1977-95, the personal computer is invented/developed/mass-marketed. 2000, China enters the WTO. 2007, the iPhone is invented. China is the manufacturing location for a great deal of modern tech hardware, including the iPhone which is ubiquitous among U.S. consumers of all incomes and did not exist previously.
Fourth, what role does America’s labor rigidity play in all this? The authors note that a standard economics textbook would have predicted a certain kind of labor market response to a trade shock, but that theory has not been realized:
“Labor-market adjustment to trade shocks is stunningly slow, with local labor-force participation rates remaining depressed and local unemployment rates remaining elevated for a full decade or more after a shock commences.”
My second reaction to the paper was not to blame the shock for the slow labor adjustment, rather it was to blame federal and state labor market regulations that hinder flexibility. Increased trade adjustment assistant (TAA) surely slows the adjustment, just like unemployment compensation slows the search for work while causing skills to atrophy. The insight that free foreign trade is incompatible with controlled markets domestically is well documented, and the U.S. case in recent years only confirms that challenge. I think what this paper shows is the weakness of U.S. labor regulations, which were exposed by the size of the shock.
The bottom line is that Autor, Dorn, and Hanson have written a brilliant paper, and it is one that I will be reading more closely and pondering for a long time. But I do want to make it clear that the paper does not change my priors. Free Trade is Good. No exceptions, no limits.