How much does import competition affect U.S. jobs? We at AIER have always argued for the benefits of free trade, but it is still worth understanding what happens to those who may lose their jobs because of international competition. The benefits from trade far outweigh the costs, but as we wrote in February, one reason “free trade remains a topic of debate is that while the benefits diffuse widely through society in the form of lower prices and overall economic growth, the harm is sometimes more directly visible.… [T]hose individuals in the U.S. who have been harmed by free trade have formed a powerful voting bloc.” It is therefore worth knowing exactly how many people lose their jobs because of international competition.
One very influential 2013 study estimates that a quarter of the decline in total manufacturing employment between 1990 and 2007 came from Chinese import competition, which translates to about 1.5 million fewer jobs. This is less than the impact of automation in this time period but still a substantial part of the story. This study has rightfully gotten a lot of attention from admirers and critics alike, and several analysts have challenged their methodology and therefore the estimated size of the employment impact.
One recent example is a March 2017 comment by three researchers from UC-Davis and Tsinghua University. They argue that the 2013 study overstates the employment effect of import competition by a factor of two by ignoring the effects of the housing boom. Essentially, the kind of Rust Belt areas hit hardest by import competition also experienced less of a housing boom, and the lower demand for construction jobs should be held partially responsible for the employment declines previously attributed to import competition. Thus, the job losses attributable to trade are closer to 800,000. The original study’s authors respond to this comment by arguing that import competition is driving the measured differences in housing demand in the first place. When workers lose their jobs, they move away or at least don’t look for a bigger house. Both demand for residential housing and for commercial construction drops, thus cutting off one potential career change for displaced manufacturing workers. Therefore, the comment doesn’t change the original paper’s conclusions. Construction might be one avenue explaining how imports affected employment, but import competition is still the root cause.
I find the original study’s authors’ points more compelling and therefore think the 1.5 million number is the best estimate. Of course, factory workers whose plant closes might not care whether they were laid off because of trade, automation, or bad luck. However, for observers evaluating trade policy, knowing how many workers are affected by enacting or reducing trade restrictions is important.