Responsible researcher: Bruno Benevit
Original title: Importing Political Polarization? The Electoral Consequences of Rising Trade Exposure
Authors: David Autor, David Dorn, Gordon Hanson and Kaveh Majlesi
Intervention Location: United States
Sample Size: 432 congressional districts
Sector: Politics
Variable of Main Interest: Voting
Type of Intervention: Import exposure
Methodology: OLS, 2SLS
Summary
The increase in political polarization observed in the United States has raised several questions about its causes. With the aim of identifying one of the factors that contributed to this phenomenon, the authors of this article verified how the increase in competition for imports impacted the positioning of the country's voters. The results indicated that counties exposed to imports became more polarized and a general tilt of the American population to the right. Furthermore, there was an increase in the prevalence of the Republican and Democratic parties according to the racial majority of the county's population, implying a reduction in the share of moderate Democrats elected regardless of the county's demographics.
The increasing rise of ideological division in the United States has been reflected in American political tension. This behavior was manifested through the political orientation of campaign donations and elected politicians from both parties, Republican and Democrat, which has been expanding since the end of the 1970s (Autor et al. , 2020). From the 21st century onwards, the public's media consumption patterns in the country also highlighted this phenomenon. While approximately half of those interviewed in the country supported moderate political positions in the 1990s and 2000s, this share dropped to 40% from the end of the 2000s onwards (Pew Research Center, 2014).
This period coincides with the increasing exposure of American local labor markets to increased foreign competition from China. This exposure directly implied the significant and persistent drop observed in industrial jobs in the United States, in addition to resulting in reduced income for workers in the most affected sectors (Autor et al. , 2020). The authors highlight that the increase in polarization is related to race and level of education in the counties affected by the expansion of commerce. As a consequence, ethno-nationalist movements began to have more relevance in the political positions of the Republican Party.
To the extent that financial crises and income inequality are capable of implying an increase in political polarization, the voting pattern in congress is related to the economic context. Thus, the concentration of the impact of the China shock in specific sectors and regions makes the economic consequences of the expansion of trade between countries politically relevant. The concentrated and well-delineated economic geography of the trade shock with China allows us to establish relationships between economic shocks in the labor market and their political outcomes.
The biggest increase in US imports occurred shortly after China joined the World Trade Organization in 2001: China's share of global manufacturing exports jumped from 4.8% in 2000 to 15.1% in 2010 (Brandt et al. al. , 2017). The increase in Chinese exports was driven by productivity growth induced by reforms in the manufacturing industry (Brandt et al. , 2017). However, the productivity gains associated with these reforms were significantly reduced until the end of the 2000s (Autor et al. , 2020).
In this sense, the authors evaluate the implications of the local labor market for electoral politics in the United States considering displacement zones (ZDs). The study also evaluates the impact of Chinese import exposure by sector and verifies heterogeneous effects arising from demographic characteristics of the respective ZDs.
The article also uses different forms of expression of political engagement. For this purpose, three sources of data were used: (i) data from public opinion surveys by the Pew Research Center on the political beliefs of potential voters; (ii) data from Nielsen Local TV View on the audience of stations with different political alignments (Fox News, with a more conservative tendency, and MSNBC and CNN, more progressive); and (iii) the Database on Ideology, Money in Politics, and Elections (DIME) measures of campaign contributions. These data reveal how the demand for ideology observed in the American population occurred in the period between 2000 and 2016.
Pew Research Center data used to measure changes in voters' ideology over time by periodically conducting questionnaires related to respondents' political beliefs. In each of ten questions, participants choose which of two opposing statements about a topic, one leaning left and one leaning right, best reflects their opinion. Left- and right-skewed responses are computed with values -1 and +1, respectively. Based on the participants' responses, a measure of the distribution of political beliefs from left to right in the range of [-10, +10], the Pew ideology score, is defined. Voters were defined as liberal (score between -10 and -3), moderate (score between -2 and 2), and conservative (score between 3 and 10).
Data from Nielsen Local TV View tracks TV viewing in U.S. households, measuring audience size for particular programs and programming times using electronic monitors connected to home TV sets and viewers' diaries. Nielsen ratings indicate the fraction of all TV households that have a TV tuned to a specific program at a given time. The authors considered the average ratings for the period from 5pm to 11pm, Monday to Friday, covering the period from 2004 to 2016.
Data from DIME is collected through reports required by the United States Federal Election Commission, which tabulate campaign contributions by donor and recipient for all amounts above $200. Based on these data, Bonica (2013) proposes the creation of a campaign financing score (EFC), establishing political alignment based on the net benefit between voting for candidates in general in relation to a specific candidate with antagonistic ideological positions.
To assess the effect of the shock caused by the expansion of imports from China on American electoral positioning, the authors use the average variation in the penetration of Chinese imports in the industries of each DZ. ZDs were established by dividing county cells according to their congressional district, assigning each cell to its respective ZD, and weighting them by the share of the population eligible to vote. Regional trade exposure also considered variation in Chinese import exposure by sector, weighting each sector by its initial share of ZD employment. This measure only considered the portion of growth that is explained by the expansion of China's export supply, disregarding demand shocks in the United States. To eliminate noise caused by redefinitions in electoral districts, the authors performed estimates disregarding observations from county-years that experienced redefinitions two years before a given election.
To understand how the expansion of trade with China affected the political expression of American voters, the authors analyzed the impact on market share for the country's three largest news broadcasters (Fox News, MSNBC, CNN), in addition to the effect on donations to campaigns. To assess the impact on the audiences of the three broadcasters, the authors used the ordinary least squares (OLS) and two-stage least squares (2SLS) methods. Aspects of age, race, district time trends and ZD fixed effects were controlled.
The authors used the 2SLS method to evaluate the effect on campaign donations according to the EFC. Age and racial groups were controlled at the county level, as well as the portions of the population that were female, university educated, born in other countries, and of Hispanic origin.
The study also estimated the effect on electoral participation among registered voters, the probability of electing a Republican candidate, and the share of district votes for Republican candidates, separated according to the degree of competitiveness between the parties (pro-Democrat district, pro -republican or competitive).
Then, the impacts caused on the ideological orientation of the candidates elected in congress are evaluated, estimating the chance of victory of moderate candidates (both parties), liberal Democrats and conservative Republicans. Altogether, the presence of heterogeneous effects was verified according to the racial majority of the counties. Both analyzes controlled demographic aspects.
Finally, the authors adopted the 2SLS method to estimate the impact of trade expansion with China under the presidential elections. The impact on the variation in vote share for Republican presidential candidates between 2000 and 2008 and between 2000 and 2016 was analyzed. Again, demographic aspects were controlled.
The results for analyzes related to voter behavior indicate the presence of significant effects arising from exposure to Chinese import competition. In general, the results regarding the impact of the expansion on the audience of the three largest broadcasters in the country indicated that the Fox News news network increased its market share, showing an increase of around 5 percentage points (pp) over the course of the year. period from 2004 to 2012. In a compensatory way, this increase occurred at the cost of declines for the broadcasters CNN and MSNBC.
Regarding the impact on campaign contributions, the results demonstrated that there was a significant increase of 35% in contributions to liberal campaigns over the years 2002 and 2010, in addition to significant increases of 22.6% in contributions to conservative campaigns. However, no significant results were identified for moderate candidate campaigns, indicating a pattern of polarization in campaign donations.
The evidence found in the analyzes related to the electoral results demonstrated that the Republican candidates were the biggest beneficiaries. Republican party candidates saw significant increases of 2.6 pp in the participation of registered voters and 11.8 pp in their probability of election in districts most exposed to Chinese trade. Although no significant results were identified regarding the vote share in districts, the evidence suggests that the increase in the probability of Republican victory was due to the capture of votes in districts considered competitive between the parties.
Furthermore, the results of the analyzes related to the ideological position of the elected candidates corroborated the relative gains of the Republicans, especially the more conservative ones. Although liberal Democratic candidates showed an increase in the probability of victory between 2002 and 2016, both the increase in liberal Democratic candidates and conservative Republican candidates was mostly explained by a large reduction in the probability of victory of moderate Democratic candidates. When considering the racial profile of counties, these trends are repeated, favoring conservative candidates in non-Hispanic white majority counties and liberal candidates in non-Hispanic white minority counties.
The results of the analyzes on the effects on the presidential elections indicated that the convergence of voter behavior and the preference for congressmen in favor of the Republicans was converted into more votes for the Republican party's presidential candidates. Considering the models with all covariates, the shock from the expansion of Chinese trade caused an effect of approximately 1 pp for the periods from 2000 to 2008 and from 2000 to 2016.
In this article, the authors demonstrated, through a robust methodology, that the districts' contact with the expansion of trade with Chinese imports caused significant political repercussions in the United States. The most conservative wing of the Republican Party was the most favored political group in the congressional elections, a movement that came at the expense of the performance of the moderate wing of the Democratic Party. The electoral performance of the parties was corroborated by the behavior of voters in terms of contributions and audience of news stations, indicating ideological consolidation on the part of voters. Naturally, this behavior was reflected in the presidential elections.
Furthermore, the trade shock experienced resulted in heterogeneous effects with respect to the race of the exposed counties. These results confirm the existence of racial pressures recently observed in the political debate in the United States. According to the authors, this connection is partially motivated by competition for fundraising based on opportunistic groups in the face of such tensions. This heterogeneity arises from demographic and geographic aspects, so that the most labor-intensive regions were the most negatively affected in terms of work opportunities and wages, characteristically dominated by less educated men. The evidence from this study, therefore, suggests that the rise in political polarization in the United States is strongly related to the economic implications of the trade shock of Chinese imports in the country.
References
Autor, D., Dorn, D., Hanson, G., and Majlesi, K. (2020) Importing Political Polarization? The Electoral Consequences of Rising Trade Exposure. American Economic Review , 110 (10), 3139–3183.
Bonica, A. (2013) Ideology and Interests in the Political Marketplace. American Journal of Political Science , 57 (2), 294–311.
Brandt, L., Van Biesebroeck, J., Wang, L., and Zhang, Y. (2017) WTO Accession and Performance of Chinese Manufacturing Firms. American Economic Review , 107 (9), 2784–2820.
Pew Research Center. (2014) Political Polarization in the American Public .
Check the Institution's registration in the e-MEC System here
© 2025 All Rights Reserved
