Protectionist sentiment is running high in the US, with both presidential candidates citing the need to shield workers from the alleged harmful effects of foreign trade. So it is perhaps ironic that the nation that has the most complaints against it for violating trade agreements is…
USA USA USA…
Chart: Goldman Sachs
So USA has almost 4 times more complaints against it for violating agreements that Chynaa.. Protectionist That!
As Goldman’s Marina Grushin notes, public opinion seems to stand behind this view, as Pew surveys have found that roughly half of Americans believe trade destroys jobs and lowers wages, compared to only about 20% who think the opposite. Perhaps more surprising, a review of research on trade, jobs, and wage inequality over the last 25 years shows that economists are also increasingly emphasizing the costs that trade can impose on US workers.
Early days: Think tech, not trade
Economic theory predicts that, in aggregate, trade benefits all parties; by importing goods that a trading partner can produce more efficiently, countries increase their consumption and welfare. But economists have also long recognized that trade leads the prices of labor to converge across borders. For developed countries, that pressures less-skilled workers, who find themselves effectively competing with cheaper foreign labor, and who face challenges in transitioning to more competitive parts of the economy.
As trade flourished in the 1970s and 1980s, its effects on labor markets attracted increasing attention. Between 1970 and 1990, goods trade rose from 8% to 15% of US GDP, while the share of manufacturing workers in US employment declined from 25% to 16%. Wage inequality increased, with the “premium” for a college vs. a high school graduate growing from around 45% to 60%. Economists agreed that blue-collar workers were being squeezed; the question was how much of it was due to trade.
For most researchers, the answer at the time was very little. Skill-biased technological change (e.g., the automation of routine tasks) and related productivity gains were generally deemed more important. Robert Lawrence and Matthew Slaughter (1993), for example, concluded from shifts in traded goods prices that trade contributed little to rising wage inequality. Paul Krugman (1994, 1995) similarly assigned trade a “quantitatively minor” role, and estimated that trade with less developed countries accounted for only around 10% of the increase in US wage disparity over the prior 20 years. Effects on employment were also deemed modest. By the estimates of Jeffrey Sachs and Howard Shatz (1994), trade with developing economies between 1978 and 1990 reduced US demand for lower-skilled manufacturing jobs by just 6.2%.
Some researchers did find more substantial effects. Translating the US trade deficit into an effective increase in the supply of less-skilled labor, George Borjas, Richard Freeman, and Lawrence Katz (1992) showed that trade accounted for up to 25% of the widening US wage gap between 1980 and 1985. And Adrian Wood (1994), contending that most studies understated the labor displaced by imports, concluded that trade reduced unskilled manufacturing employment in developed economies by 21.5%, more than three times the Sachs/Shatz estimate. Still, these views were in the minority; among more than 30 studies on wage inequality reviewed by William Cline (1997), most found that the adverse impacts of trade were minimal to modest. In short, the research acknowledged some losses from foreign trade, but emphasized overall gains.
The new view: A bigger role for trade
Between 1990 and 2010, developing economies’ share of world trade roughly doubled, to 38%, driven in large part by EM Asia and China’s 2001 entry into the WTO. In the US, the college wage premium approached 80%, while the trade deficit widened well beyond prior extremes. These shifts prompted economists to revisit the conclusions of the 1990s with new data. Krugman wrote in 2008 that it was “no longer safe” to argue that trade’s impact on inequality in developed economies was insignificant.
Indeed, the work that followed often pointed to greater costs from trade. In an update of Krugman’s 1995 model, Josh Bivens (2013) estimated that trade with less developed countries accounted for one-third of the rise in US wage inequality between 1979 and 2011—and more than 90% of it after 1995. In another example, Michael Elsby, Bart Hobijn, and Aysegul Sahin (2013) found that import exposure could account for 85% of the 3.9pp decline in US workers’ share of national income over the prior 25 years.
Recent research has also highlighted the spillovers from pressure on US manufacturing. In one such study, Avraham Ebenstein and colleagues (2014) found that trade was pushing workers out of generally higher-paying manufacturing jobs and into lower-paying positions in other parts of the economy. Using census data on individual workers across industries, the authors estimated that people who switched occupations due to trade or offshoring saw their real wages fall 12-17% between 1984 and 2002.
China has played an important part in these developments, having increased its share of US imports to 21% in 2015 from only 6% in 1995. MIT’s David Autor and other researchers (2013, 2016) mapped the exposure of 700+ US labor markets to this surge based on initial industry composition, and concluded that workers in more-exposed regions faced lower lifetime earnings, particularly if they were already at the lower end of the pay scale. By Autor’s estimates, Chinese imports cost the US up to 2.4mn jobs between 1999 and 2011, of which nearly 1mn were in manufacturing. Others have found evidence of US jobs effectively “moving” abroad: Ebenstein et al (2012) noted that US job losses have corresponded with Chinese gains in the same sectors. That these shifts occurred even for routine tasks suggests, in their view, that US workers are being displaced by trade rather than technology.
When the facts change, I change my mind
The recent research has not invalidated earlier findings; indeed, the trade landscape has changed considerably since the “first wave” of analysis. More importantly, economists remain proponents of free trade (and, to be sure, some maintain that trade is not an important driver of wage inequality). However, the tone around trade appears to be shifting toward a greater acknowledgment of concentrated losses rather than an affirmation of overall gains. In the words of Harvard University’s Dani Rodrik, “The populist rhetoric on trade may be excessive, but few deny any longer that the underlying grievances are real.”