top of page


The New (Anti) Free-Trade Consensus

June 25, 2024

By Rob Lester

Earlier this month, Republican National Committee spokeswoman, Anna Kelly, raised eyebrows when she told Bloomberg’s Nancy Cook that “the notion that tariffs are a tax on US consumers is a lie pushed by outsourcers and the Chinese Communist Party.” While it’s true that tariffs can lower the prices received by Chinese exporters, as an empirical matter the 2018 tariffs on China applied by President Trump have mostly been passed through to US importers. And, over the long run in competitive industries, higher import prices ultimately get passed onto consumers. A full accounting of the costs and benefits of tariffs requires nuance and the quantitative magnitudes may be historically contingent. But setting aside complicated questions about optimal unilateral tariff rates, Ms. Kelly’s statements reflect a sentiment increasingly shared among the electorate and politicians. Namely, that international trade is a net negative for American workers.  

  Although Ms. Kelly may have been intentionally hyperbolic, she is in good intellectual company in questioning whether the US gains from free trade. Indeed, the eminent economist and Nobel Laureate, Angus Deaton, recently expressed skepticism of globalization’s benefits to American workers. Meanwhile, earlier this spring, Treasury Secretary Janet Yellen spoke of the need to diversify away from China. Although she was careful to say that President Biden’s policies, specifically those in the Inflation Reduction Act, were not a turn towards “American protectionism,” the President’s recent tariffs on Chinese imports imply otherwise. It’s one thing for politicians and political operatives to disparage globalization and promote tariffs. But how did we arrive at an intellectual place where economists of Deaton and Yellen’s stature question free trade orthodoxy?

  A good place to start is the trade relationship between China and the US. A combination of Chinese economic reforms and reductions in US tariff rates, culminating in China joining the World Trade Organization in 2001, caused the share of US imports from China to increase from three percent in 1990 to 19 percent in 2020. While the inflow of Chinese goods into the US benefited consumers through lower prices, it had uneven employment effects. In a series of papers, economists David Autor, David Dorn, and Gordon Hanson showed that communities heavily exposed to Chinese imports experienced reduced wages and employment rates, higher levels of disability claims, more political polarization, and lower marriage rates. Workers in these communities were very slow to reallocate geographically or change occupations. The now infamous China Shock confirmed that trade can have large distributional consequences, leaving some workers worse off for a sustained period of time.

  Increased data availability and advances in computational techniques continue to improve our understanding of globalization. Economists studying international trade have developed highly sophisticated models that address the aggregate and distributional consequences of trade with China. A one sentence summary is that despite the adverse distributional consequences, the increase in trade with China raised the living standards of the average American. I focus, however, on what gets taught at the level of an economics principles class. My reasoning is that many undergraduates are exposed to at least a small dose of classroom economics and they likely take the soundbites from the classroom (e.g. the minimum wage causes unemployment or inflation comes from excess money creation) into their professional lives. I’m not saying these soundbites are always correct or even correctly recalled, but they do form the prejudices and instincts of someone reading the news, writing the news, or, in Ms. Kelly’s case, making the news.

  Does introductory economics explain the China shock? The conventional model students learn in that class is attributed to the 19th century British economist, David Ricardo. In the simplest version of Ricardo’s model, there are two countries (say, the US and China), two goods (such as software and textiles), and one factor of production (labor). In a world without trade, each country divides its labor force between software development and textile production. China and the US can increase their incomes through trade by specializing in the good they produce at a lower relative cost. The subtlety of the Ricardian model is that even if the US is more efficient than China at producing textiles and software, the US gains from trade by allocating their labor force into the industry for which they have a comparative advantage. To take a specific example, suppose a worker in the US can produce either ten shirts or two software programs in a single day. The “cost” of a software program is five shirts. In China, a worker can produce seven shirts or one piece of software in a given day. The cost of a piece of software in China is seven shirts. In this example, the US has a comparative advantage in software and China has a comparative advantage in shirts. Consequently, US workers gain from trade by specializing in software development and Chinese workers gain by specializing in shirt production.

  To answer the question posed at the beginning of the last paragraph, no, the Ricardian model does not explain the distributional effects of the China shock. The reason is that there is only one factor of production—labor—that is perfectly mobile across industries. To say that the “US gains from trade” is equivalent, in the textbook Ricardian model, to the statement that “all US workers gain from trade.” Without heterogeneity, there is no way to assess the distributional consequences of trade.

  Despite the inadequacy of the Ricardian model, it carries a powerful and non-intuitive insight how countries gain from trade. Distribution considerations notwithstanding, the professional consensus overwhelming supports the conclusion that countries gain from trade at the aggregate level, with the only debate being the quantitative size of the gains. Accordingly, my recommendation is that economic instruction at the principles level should be refined rather than totally overhauled. The source of this refinement is to incorporate the specific factors model, which is usually taught in international economics electives.

  The key addition to the specific factors model is a second factor of production and the assumption that one factor is immobile. When a country opens to trade the immobile input in the import-competing industry experiences a decline in income. The application to the China shock is clear. If one interprets blue-collar labor as the immobile input, which is consistent with Autor, Dorn, and Hanson’s empirical findings, then free trade will lower the incomes of blue-collar workers in import-competing industries. Instructors could discuss how the import competing industries were geographically concentrated and what factors (e.g. cultural or political) might inhibit these workers from changing occupations or moving to different cities. The specific-factor model doesn’t add much analytic complexity once students have already learned the Ricardian model and the basics of supply and demand and it also allows them to address distributional issues of international trade, which are central to evaluating the China shock.

  As I mentioned at the beginning, trade theory also helps us make sense of recent empirical findings on the US-China trade war. President Trump’s 2018 tariffs raised import prices and reduced aggregate income. Meanwhile, Autor, Dorn, and Hanson, the original expositors of the China shock (this time with Anne Beck), have shown that Chinese retaliatory tariffs reduced US employment, primarily in agriculture. Looking ahead to the 2024 election, economists at the Peterson Institute forecast that President Trump’s new round of proposed tariffs would reduce median household income by at least $1,700 annually. Whatever were the distributional costs to the China Shocks, reversing it appears to be even more disruptive.  

  Notably, my revision to introductory economics comes from within the undergraduate economics cannon. At the risk of being reductive, economics textbooks tend to be neoliberal in their exposition and policy prescriptions. Authors are generally sanguine about the allocative role played by market prices and cautious about government intervention. This isn’t to say that textbook authors are free-market ideologs. To the contrary, approximately 80 percent of registered voters in the American Economic Association (the discipline’s main professional organization) are Democrats. It’s common for economists, including textbook authors, to support greater levels of income redistribution, government provision of public goods, and taxes on activities conferring negative externalities such as carbon emission. However, policy advisors in both the Biden and Trump administrations appear to be much more skeptical of neoliberalism. Moreover, as New York Times writer David Leonhardt has argued, there may be a “neopopulist” consensus emerging in Congress.

  In my view, the turn away from neoliberalism is a mistake. And because the epicenter appears to be the China Shock, adapting the way we teach international trade at the introductory level is crucial to preserving a minority supportive of free trade. Key policy institutions from the Council of Economic Advisers to the Congressional Budget Office are populated by PhD economists or people with undergraduate training in economics. Modest curricula reforms could, if not forestall the turn towards protectionism, provide sound policy advice to populist sympathetic politicians while limiting some of their misguided ambitions. 

Rob Lester is an Associate Professor of Economics at Colby College.


bottom of page