In 1999, the members of the World Trade Organisation (WTO) gathered in Seattle to launch a new round of trade negotiations that aspired to reduce trade barriers between countries. WTO delegates were met by thousands of protestors who blocked streets, cordoned off the convention centre and clashed with police – an event dubbed the “Battle of Seattle”. The protestors comprised a diffuse coalition of students, NGOs, media activists, anarchists, labour union members, and even farmers and workers from developing countries. United States Trade Representative Charlene Barshefsky and WTO Director-General Mike Moore announced the suspension of the conference. Eventually, the trade talks resumed in Doha in 2001, but failed given the intransigence of various countries. Countries weary of the WTO route quickly switched to bilateral and regional trade agreements as the means to liberalise trade and access markets.
Fast-forward to today and we see the rumblings of a trade war. China and the U.S. are threatening tit-for-tat tariffs, escalating protectionist rhetoric with generous dollops of nationalism, leading to violent fluctuations in global financial markets. President Trump seemingly hopes this protectionist rhetoric will get him in voters’ good books in the run-up to the midterm elections in the U.S. this November. More interestingly, Trump’s coalition and voter base do not resemble the student-dominated protestors in Seattle.
Political preferences and voter preferences
But herein lies a puzzle. Survey data from Gallup show that 70 percent of U.S. adults see foreign trade as an opportunity for growth while only 25 percent perceive a threat from foreign imports. More surprisingly, we observe a strong spike in favour of trade since Trump’s electoral victory in 2016, going from 58 percent in 2016 to 72 percent in 2017 (see figure below from Gallup).
Seventy-eight percent of college graduates surveyed in 2018 view trade as a growth opportunity, while the proportion for college nongraduates was 66 percent. Despite the 12 percentage-point gap, irrespective of education levels, a strong majority of Americans favour free trade.
Adding to this puzzle is the work of David Autor, David Dorn and Gordon Hanson, who, in their 2013 paper highlight what is now termed the “China shock”: Import competition from China led to a loss of 1.5 million to 2.5 million jobs between 1991 and 2007. Their research also show that U.S. regions more affected by rising Chinese imports (see figure below) experienced larger job losses, declining labour participation and a decline in earnings.
Source: The China Trade Shock
In recent work, Autor, Dorn, Hanson and Kaveh Majlesi have found evidence that in the 2016 presidential election, China-affected counties experienced greater increases (or smaller reductions) in the Republican vote share over the 2000 election (that also brought a Republican into the White House). In other words, the 2016 vote share moved towards the candidate with a more protectionist trade agenda.
Resolving the puzzle
So, how do we then explain the pro-trade survey results? More than 80 percent of eligible American voters are ideologically committed, especially on social issues such as abortion, gay rights and gun control. Therefore, any movement in their attitudes towards or against trade will not translate into changes in support for one party relative to the other.
However, in the case of ideologically-uncommitted voters in swing states, attitude towards trade could make the crucial difference. In short, a small segment of voters may have a disproportionate influence on final electoral outcomes. In fact, Trump triumphed by targeting such voters in swing states (e.g. Wisconsin, Pennsylvania, Michigan) who switched from supporting Democrats to supporting Republicans. Therefore a protectionist stance could increase a candidate’s chances of winning despite rising support for international trade. This is consistent with the recent 25 percent tariff on steel imports, whose producers are mainly concentrated in the swing states of Ohio and Pennsylvania.
Recently, Xiangjun Ma and John McLaren have theoretically proven and empirically validated a “swing-state theorem” in a U.S.-type electoral-college set-up. According to this theorem, presidential candidates devise trade policies by focusing on the well-being of swing-state residents over those of other states. This policy platform might not be popular with the majority of voters in a country but can deliver a win in the electoral college.
The divergence between the trade poll results and the election outcomes could also be explained by the relatively high turnout of older eligible voters, who may not bother responding to surveys, especially on issues such as trade. This creates a disconnect between the preferences of eligible voters and those who show up to the polls.
Coming back to the “China shock”, the job-loss estimate of 1.5 million to 2.5 million is small relative to the overall labour force and the total increase in the number of unemployed during the two-decade period studied. As a result, for an average person, this shock should have little impact on protectionist preferences. Moreover, recent research (by Yang Liang and also by Robert Feenstra, Hong Ma, Yuan Xu and Akira Sasahara) adds import and export growth, as well as service-sector linkages, to this empirical analysis. The estimated net negative impact of the China shock then either becomes much smaller or gets reversed, leading to net job creation rather than destruction. In the early 2000s, Christian Broda and David Weinstein showed large consumer gains from trade, through the availability of a greater variety of products. If this is also happening through trade with China, then it is not surprising that, as the Gallup poll shows, even the less educated support trade, despite America’s comparative advantage being in skill-intensive products.
Overall, there is no reason to distrust the Gallup poll results on trade. Clearly, there are grounds to believe that, in the long term, America will remain a champion of free trade. At the same time, the design of the electoral system in the U.S. renders it susceptible to capture by political entrepreneurs.
Devashish Mitra is a Professor of Economics and Cramer Professor of Global Affairs in the Maxwell School of Citizenship and Public Affairs at Syracuse University.
Leave a Comment