Economists have long argued that the benefits of trade between countries -- which include lower prices and higher living standards -- outweigh the costs, which include a higher risk of displacement for workers in competing industries.
Recently, however, the costs of trade have dominated the discussion, in the wake of large job losses in the manufacturing sector from 2000 to 2011. While running for president, both Donald Trump and Bernie Sanders denounced trade agreements as disastrous for the U.S. and for the nation's middle class because manufacturing-sector wages tend to be relatively high. Meanwhile, Hillary Clinton distanced herself from her earlier pro-trade positions. The argument Trump and others make is that trade with China and other low-income countries harms Americans by depriving them of good-paying jobs.
My new research concludes that the reality is more complex, and the anti-trade argument has serious weaknesses. First of all, trade has large benefits that have been ignored in recent political debates. For example, the buying power of all Americans, particularly those with lower incomes, would be much weaker if not for imports. Secondly, the employment challenges resulting from trade competition have been exaggerated. During the period when imports were rising most rapidly -- 2001 to present -- a manufacturing worker's risk of becoming unemployed did not increase relative to other sectors, and manufacturing layoff rates held steady and were consistently below those in other sectors. I also revisit research finding that the average person living in a local area exposed to Chinese imports is harmed, and I conclude that those results do not withstand greater scrutiny.
Evidence From the Economics Literature
Despite the economics profession's pro-trade orientation, recent work from a number of economists highlights losses from trade, bolstering -- even if inadvertently -- the intellectual foundations for opposition to trade agreements. One recent paper concludes that Chinese import competition resulted in the loss of approximately 2 million U.S. manufacturing jobs from 1999 to 2011, roughly one-third of the sector's total losses. If these losses are accompanied by a shift of employment to other sectors, then a decline in manufacturing employment is not necessarily a problem. An important paper from some of the same authors (David Autor, David Dorn, Gordon Hanson and Jae Song) sheds some light on how well workers are able to transition to other jobs when faced with rising import competition. After tracking U.S. workers from 1992 to 2007, they found that workers exposed to greater import competition were as likely as others to have a paid job in a given year, but cumulative earnings were lower as a result of imports. Interestingly, those who were observed working outside of their initial firm did not experience earnings losses unless they were among the lowest-paid workers initially. One cannot know for sure why most workers who switch industries -- sometimes even within manufacturing -- are able to mitigate earnings losses, but the results provide evidence that this is harder for lower-paid workers.
These nuanced results are somewhat at odds with the more pessimistic conclusions from a famous and widely cited paper called The China Syndrome, also by Autor, Dorn and Hanson. They conclude that the local areas exposed to Chinese imports do not experience offsetting benefits. As a result, they argue, residents in these areas experience an increase in unemployment and a large decline in average wages and labor force participation.
With this and other publications, these authors have made great contributions to social science. Nonetheless, I've re-examined their data and concluded that their results are sensitive to a specific way in which the authors choose to analyze their data. My own interpretation of the data finds that local areas exposed to import competition from China became no worse off, except that employment shifted away from manufacturing to other sectors. The manufacturing workers who remained in the sector experienced higher wage growth in the areas facing more import competition, and total employment shifted toward more highly educated workers outside of manufacturing.
To summarize these findings, there is strong evidence that import competition resulted in net employment losses in the manufacturing sector, but displaced manufacturing workers have found jobs with other employers just as readily as workers displaced from other sectors, like construction and retail. Those who change jobs, meanwhile, are no more likely to see income gains or losses than other workers. Finally, at the local level, there is no robust evidence that the areas most exposed to import competition -- which include prosperous areas like San Jose, California, and Austin, Texas -- experience adverse effects; indeed, during the 1990s, the more exposed areas performed significantly better on several measures. In short, competition between businesses located in the U.S. and those located abroad looks a lot like domestic competition.
Where Have the Good Jobs Gone?
Returning to the argument that import competition has disproportionately destroyed good jobs and eroded decently paying job opportunities, there are two additional considerations. First of all, net job growth from 2000 to 2015 has not been concentrated in low-paying industries. If anything, the distribution of jobs by industry has shifted toward slightly higher-paying industries. Among the high-paying industries that have seen large net job gains are healthcare, computer systems design services, consulting and management services, insurance, and various professional services. Meanwhile, many of the manufacturing job losses were concentrated in lower-paying industries like apparel and textiles. Shifting employment across industries, therefore, cannot explain low average wage growth.
Secondly, while the manufacturing sector does pay well -- even after years of exposure to import competition -- most good-paying jobs for workers without a college degree are outside of manufacturing, many in occupations requiring a high level of technical knowledge. Using 2015 Census data from IPUMS-USA, I estimate that 84% of workers with less than a bachelor's degree who earn $50,000 or more (which is higher than the median manufacturing worker) work outside of manufacturing. Thirteen percent work in professional services, finance or insurance; 12% work in retail and wholesale trading companies; 11% work in construction; 10% in healthcare; 9% in transportation and warehousing; 8% in government; and the rest in other sectors.
Globalization has become a convenient scapegoat for rising inequality and weak economic performance because the blame falls on foreigners. An alternative explanation is that the U.S. and perhaps other rich countries are hampered by weak competition in financial markets and rising inefficiency in purely domestic markets like healthcare, housing and education, which consume larger and larger shares of national spending with little to show for them.
[Editor's Note: This post was updated to clarify the summary of Autor, Dorn, Hanson and Song's findings.]
