The China Shock: Reevaluating the Debate

As the U.S.- China relationship continues to deteriorate, the role of trade has come under increasing scrutiny. Commercial exchanges between the two countries used to be the ballast of the relationship. Trade was understood to benefit both countries economically and helped reduce tensions on political and strategic issues. Over the past decade, this has changed dramatically: not only is economic interdependence now seen in many quarters as a liability, but much of the Washington policy community now believes that the benefits of trade with China are far outweighed by the negative effects. A central element of concern focuses on the negative impact of imports on manufacturing jobs and employment more generally in the United States. This sentiment is particularly held in regions of the country that had strong manufacturing sectors prior to 2000, such as the Midwest and the South. 

The idea that trade with China is inherently disadvantageous for the United States and that it has cost the country jobs continues to shape the political discourse in Washington. This was used as a justification for the trade war and continues to provide grounds for retaining tariffs first imposed by the Trump administration. Although the debate appears to be settled in Washington, it is long from over among economists, who have analyzed this issue utilizing a range of different methodologies and datasets. Whether intentionally or not, Washington has gravitated toward one side of the debate, and it is important that the policy community be aware of the larger ongoing conversation that scholars are still having.  

So, what does a broader review of the data from multiple studies show? Scholars generally find that prior to 2010, imports from China negatively affected manufacturing jobs in the United States. However, there are mixed findings on the net effect on the economy, the final balance of jobs lost in manufacturing, and the growth in service sector jobs. There is also no evidence of trade with China having a significant effect on jobs after 2010, though some economists have found that regions more affected in the 2000s have struggled to recover employment. It is particularly noteworthy that economists have measured the direct effect of imports from China on U.S. employment in the early 2000s and have not found this to be an ongoing trend today. That means that trade measures imposed today would have to have a retroactive effect to fix what happened almost two decades ago.  

There is one other result that all scholars seem to agree on: better-educated, more economically diverse regions of the United States were affected far less by the surge in imports from China. This finding is consistent with data on the impact of globalization in other countries, where better education and opportunities for retraining improved the likelihood that workers would benefit from international trade. Thus, much can and should be done to address the plight of displaced workers through domestic programs. Greater protectionism to reduce imports may not be the best tool to address labor displacement even if imports were the original direct cause of job losses. Thus, current data-driven academic research should inform more directly the policy debates in Washington about tariffs and industrial policy.

The Origins of the China Shock Mythology 

China’s rise as a manufacturing giant originates in 1978 and the announcement of “Reform and Opening,” when the country’s leadership took the first steps to allow foreign investment and to move away from the planned economy. The effects were first felt in earnest in the 1990s when reform policies accelerated, and an increasing number of foreign firms began to relocate production to China.  

David Autor (M.I.T.), David Dorn (University of Zurich), and Gordon Hanson (Harvard Kennedy School) argue that the “shock” to the global economy began in 1992, when foreign trade became a substantial element of China’s economy, only to end around 2010 when China’s share of U.S. imports stabilized.  

The growth in China’s exports was due to domestic reforms that increased productivity and policies to open the country to global trade. Maoist-era economic policies had effectively repressed China’s economic performance, while Deng Xiaoping’s reforms rapidly unleashed China’s economic potential. Entry into the World Trade Organization (WTO) in 2001 amplified China’s already improving productivity. Moreover, the certainty of lower tariff and non-tariff barriers also made China an attractive destination for foreign direct investment.  

Several important circumstances coincided with China’s ascension to the WTO in the early 2000s. The share of manufacturing employment had already been falling in the United States as other sectors of the economy gained importance. Globally, various emerging markets began to grow in a way so remarkable that the acronym BRICS was coined to refer to the breakthrough of Brazil, Russia, India, China, and South Africa. The European Union was established, creating the largest free trade zone and single-currency international bloc. Central and Eastern European economies that had been under Communist planning experienced rapid growth as their productivity and trade increased and they were incorporated into the broader European economy. As this feature shows, there is little consensus on how rising imports from China have affected manufacturing and related employment globally.  

The Debate 

Autor, Dorn, and Hanson are some of the authors that have spent the most time analyzing the impact of trade with China on U.S. employment. The “China Shock” is a term they popularized with their influential paper published in 2016. Their research has been widely cited in newspapers, congressional testimonies, and the policy community. It is not clear, however, that their findings have always been understood in their entirety. And their work has rarely been compared in the media with other scholarly work.  

While Autor, Dorn, and Hanson have published the most on the topic, several other scholars have examined the question by drawing on a range of data sources and methodologies. The findings of various scholars do not always fully align with one another (see Figure 1). 

Among the many scholars who have investigated this topic, this feature focuses on three groups of economists because of the high quality of their work and the clarity of their positions. The first research group includes the aforementioned David Autor, David Dorn, and Gordon Hanson, who are based in the United States and Switzerland. These authors have published widely on the topic and the feature draws on several of their papers. For more details on sources for this feature, please see the methodology box at the end of this feature. 

The second group is composed of U.S.-based scholars Nicholas Bloom (Stanford University), Kyle Handley (UCSD), André Kurmann (Drexel University), and Philip Luck (Department of State). Nicholas Bloom has also looked at the impact of import competition with China on innovation in European firms. The third group includes Zhi Wang (George Mason University), Shang-Jin Wei (Columbia University), Xinding Yu (University of International Business and Economics), and Kunfu Zhu (Renmin University), a team of scholars based in the United States and China. This feature draws largely on research from each of these two teams’ most important papers on the subject (see methodology box).  

The scholarly discussion amongst these groups revolves around four inter-related questions, which this feature will walk readers through. There is agreement on some issues, while on others there are substantial differences — as is common in academic debates. The next section will compare and contrast their positions.  

1. The effect of the ‘China Shock’ on manufacturing employment during 2000-2007 

Autor, Dorn, and Hanson  

This group of scholars have found a geographically concentrated effect of the “China Shock” on manufacturing jobs and wages in the decade leading up to 2010. Autor, Dorn, and Hanson find that increased imports from China hurt employment and wages in regional labor markets that were more exposed to import competition with China. In other words, areas with a high concentration of manufacturing-based industries that were in direct competition with Chinese companies experienced a fall in employment and wages.  

This research team has published extensively over the past decade, and has proposed a range of slightly different estimates. One of their more influential papers published in 2016 (also co-authored with Daron Acemoglu and Brendan Price) indicates that between 1999 and 2011, the jobs lost because of import competition with China in the United States totaled 985,000 in manufacturing and 1.98 million for the entire economy, including both upstream and downstream sectors. Another alternative methodology they used generates a higher estimate of up to 2.4 million jobs lost. A paper from 2013 finds that the loss of manufacturing jobs added up to 1.5 million. 

This group of scholars does not attempt to estimate the positive effects on job creation in other sectors, which differentiates them from the others analyzed here. However, in a 2021 paper, Autor, Dorn, and Hanson find that despite the gains in purchasing power for American consumers benefiting from low-cost Chinese products, 82 out of the 722 regions they analyzed (accounting for 6.3 percent of the U.S. population) still experienced a 1.27 percent reduction in per capita income. 

Bloom, Handley, Kurmann, and Luck  

This group uses a different, confidential, dataset compared to Autor, Dorn and Hanson. However, both datasets come from the United States Census Bureau and the analytical strategy adopted by the scholars is similar (see the methodology box for more details). This research team, which analyzes the period from 1990 to 2015, finds that imports from China had a significant negative effect on U.S. manufacturing jobs before 2010 in areas with lower shares of college educated workers. However, unlike Autor and his co-authors, Bloom, Handley, Kurmann, and Luck try to estimate the effect of growing imports from China on the service sector and find a significant positive effect, discussed in the next section.  

Wang, Wei, Yu, and Zhu 

The third group of experts – Wang, Wei, Yu, and Zhu – also conclude that the direct negative effect of imports from China on manufacturing employment in the United States was substantial and consistent with the findings of the other two groups. Given that their data and methodology are similar to those used by Autor and his co-authors, this is not surprising. However, there are significant differences when it comes to their estimates for service sector jobs and wages, discussed in further detail below.  

2. Did trade with China cause net job losses in the 2000s in the United States? 

Autor, Dorn, and Hanson 

Autor and his co-authors have consistently found that areas that were more exposed to trade with China experienced a net job loss that was not compensated by out-migration or growth in employment in sectors other than manufacturing. In other words, they do not find evidence for the reallocation of labor from less productive industries to more productive ones. 

As mentioned, Autor and his co-authors do not estimate whether imports from China had a positive impact on non-manufacturing jobs—especially on a national scale. On a regional level, they find that the manufacturing job losses translate directly into an increase in the unemployment rate and labor force non-participation rate. This is not necessarily because service jobs did not materialize elsewhere, but rather because the migration rate from these regions is not particularly significant. In other words, those who lost manufacturing jobs in one region stayed in that region and were unable to find alternative forms of employment. 

Bloom, Handley, Kurmann, and Luck  

Bloom and his co-authors take a different approach from that of Autor and his co-authors by trying to account for the effects of trade with China on the service sector. They find that loss of employment in areas with low levels of human capital, measured as the share of the population with a college degree, such as the Midwest and the South, was compensated by an increase in service jobs in areas with high human capital, such as the West Coast and the Northeast. Consequently, they disagree with Autor and his co-authors that trade with China caused a net loss of American jobs.  

This does not mean that there were no significant losses within regions. Like Autor and his co-authors, this second team of economists did not find related migration in response to the changing job opportunities. Thus, Chinese imports may have caused jobs and earnings to move from the heartland of the United States to the coasts.

These scholars also find that most companies affected by the imports from China were multinationals that offshored manufacturing and expanded service jobs in coastal regions in the United States. This means that these companies were not negatively affected, even though their workers were. 

Wang, Wei, Yu, and Zhu 

Wang and his co-authors’ findings on net job losses are the most different from Autor and his co-author’s. This is because of two methodological choices on their part: they distinguish between intermediate and final goods, and they use exporter-specific information to allocate imports from China to the downstream sectors in the United States. Wang and his coauthors claim this helps account for the effects throughout the supply chain. Taking this approach, they find that over the 2000-2007 period the average effect for each region analyzed is a net job increase of 1.27 percent (see the methodology box for more details). They also find a 75 percent wage increase for workers in an average region because of exposure to the China trade.  

This third research team finds an increase in service jobs even in the regions that experienced the sharpest fall in manufacturing jobs due to import competition with China. According to their data, even within areas that were hit the hardest by import competition, trade led to an increase in other job opportunities in the service sector. The authors zoom into five of the hardest-hit counties and find that even in those areas, the overall effect on employment was positive. This finding is different from that of Bloom and his co-authors. The latter found that most service jobs materialized in areas that did not coincide with those where most of the manufacturing job losses took place. Wang and his coauthors conclude that the winners from trade far outstrip the losers, but they do highlight that 25 percent of workers are indeed worse off with lower salaries due to trade with China.

3. What are the effects beyond 2010? 

Autor, Dorn, and Hanson 

Autor and his co-authors find that the broader trade shock on jobs stabilized after 2010, but the effects have lingered and continued to affect communities until at least 2019. In 2019, the regions most affected by import competition had not returned to previous levels of manufacturing jobs or employment rates. It’s important to note that the finding is not that the China shock persists—the economists are clear that it plateaued after 2010 – but rather that the regions affected by import competition with China have struggled to recover a decade later. 

The regions most affected are those with lower shares of college graduates and those narrowly specialized in labor-intensive manufacturing—the most vulnerable industry to Chinese competition. Regions with higher shares of college graduates seem to have recovered more effectively. The advantage of higher education, discussed in the next section, is consistent with the findings of the other scholars analyzed in this feature. 

Bloom, Handley, Kurmann, and Luck  

Bloom and his co-authors find a strong effect of import competition from 2000-2007 on manufacturing and service sector jobs, but they are unable to find any effect at all between 2008 and 2015. They do not evaluate how the regions most affected by import competition with China have performed in the long term. 

Wang, Wei, Yu, and Zhu 

Wang and his co-authors do not address the question, as the focus of the study is from 2000 up to 2007. 

4. Education is significant in reducing the negative effects of trade with China  

All authors find this to be true. Bloom and his co-authors and Autor and his co-authors both find that regions with higher shares of college graduates were less negatively affected by imports from China. Moreover, Wang and his co-authors find that college-educated workers were winners from trade, with an average wage growth of 7.2 percent a year as opposed to non-college-educated workers whose wages declined by 4.3 percent a year on average. This is likely because college graduates are better placed to take advantage of increased opportunities in the service sector. As discussed in the last section, this data suggests that much can be done to improve the country’s resilience to the negative effects of trade by improving access and quality of higher education. 

Not all shocks were alike: The U.S. in comparative perspective 

A full reading of the scholarly debate shows a more ambiguous effect of Chinese imports on American jobs. It is also important to recognize that the experience of the United States may not be equivalent to that of other countries with extensive trade ties to China. Understanding the similarities and differences also has important policy implications.  

The United States is not the only country that experienced rising Chinese imports in the early 2000s. When comparing trade with China and other countries over the past decade, the United States looks fairly average (see Figure 2). Chinese imports grew from around 8 percent of total U.S. imports in 1998 to 18 percent in 2021. Of course, import data does not say much about the composition of those imports and how they compete with domestic industries. Nonetheless, it looks like the United States was not an outlier.  

When turning to manufacturing employment, the data indicate that the United States lost manufacturing jobs over the 2000s. American manufacturing employment fell from roughly 20 million jobs in 1998 to slightly over 15 million in 2021. However, in the timeframe between 2008 and 2019, the country actually experienced growing manufacturing employment in absolute terms (Figure 3). Of course, not all manufacturing jobs were lost due to import competition. Even Autor, Dorn and Hansen estimate that around 59 percent of manufacturing jobs losses in the United States between 2001 and 2019 were caused by import competition with China over the first decade of the 21^st century. 

Another way to look at comparative trends in employment is to examine manufacturing employment as a share of countries’ total employment. In that regard, just almost all countries in the sample, including China, has seen a drop in manufacturing’s share of jobs. This is partly due to an absolute loss of some jobs but is increasingly the result of rapid growth in service sector jobs.  

The most striking variation across countries becomes apparent when the change in imports from China is compared with the change in manufacturing employment in several major economies. As Figure 5 shows, manufacturing job losses in the United States were high relative to the increase in imports from China experienced. Countries “above the line” did a better job of absorbing growing imports from China without shedding a lot of manufacturing jobs.    

Three things stand out from the comparative data about 2000-2010: 1) A substantial number of countries experienced a decline in manufacturing jobs. In our sample of 26 countries, 15 saw a drop in manufacturing jobs over the 2000-2010 decade; 2) The United States, though, stood out as having lost a lot more jobs relative to the average. If the United States had lost jobs at the same rate as the other countries in the sample in figure 5 relative to the growth in Chinese imports, it would have only lost 785,763 jobs. That’s 3,923,252 fewer jobs than those lost in reality; 3) A cursory look at the data shows little indication that more imports from China are correlated with a decrease in manufacturing jobs globally. If anything, it looks like imports may be associated with a rise in jobs. Of course, to answer this question far more factors would need to be taken into consideration, including country-specific and industry-specific conditions. Regardless, the data does suggest that there may be multiple ways of looking at the impact of trade with China.  

After 2010, the picture looks quite different. The United States and several other countries saw an increase in manufacturing employment over the past decade. Manufacturing jobs in the United States increased by 10 percent between 2010 and 2019 even as imports from China rose by 23 percent. This is consistent with the findings of the economists featured here, showing that the impact of Chinese imports on U.S. employment dissipated by around 2010 (see Figure 6). 

The most important takeaway from analyzing the implications for Chinese imports on other countries is that the effect has varied widely. In Figure 5, although one can identify a general pattern, denoted by the upward slanting line, the fact that so many countries are far away from the trend line shows that many other factors are at work. Some countries shed a disproportionate amount of manufacturing jobs during 2000-2010 period relative to the increase in imports from China, while other countries had less labor dislocation and difficulties. The United States falls into the camp of those who appear to have suffered more losses, though the research presented here shows a range of conclusions about the extent of the shock and the net effect on employment.  

There are others who also have suffered, yet in different ways. In Denmark, a country with very liberal firing regulations (like the United States) but strong unions, import competition with China appears to have resulted in a decrease in wages but not a big fall in employment. Employees’ work hours fell, but they were kept on the payroll by companies. One study suggests this may have led to an increase in workers seeking additional education—which subsequently led to higher salaries. At least one paper analyzing Portugal argues that competition with China in markets like Germany led to substantial reductions in employment between 2000 and 2007, after having initially driven companies out of business in the late 1990s. 

By contrast, Germany’s industrial structure seems to have largely spared the country from import competition with China in the 2000s. Rather than undermining German exports, imports simply shifted away from low-wage southern European countries, such as Portugal and Greece, to China. By the 2000s, Germany was already a high-value-added manufacturing hub.

Research suggests that most of the losses in German manufacturing jobs at the time appear to have been caused by competition with Central and Eastern European countries that were integrating into the broader European market. Finally, some decisions may have helped German workers navigate the transition better than the United States. German unions focused on requiring companies to invest in retraining which was crucial in enabling workers to stay on the job as manufacturing upgraded. For companies, the benefit lies in the fact that it is generally cheaper to retrain an existing worker than recruit a new one. 

These various cases point to the fact that investing in training, and education, and encouraging companies to retain workers, when possible, may help increase resilience to trade shocks. 

The Policy Debate: What should not and should be done 

The economists highlighted in this feature do not always agree with each other about how to measure the effect of imports on employment, but there is little debate over what the best solutions to job displacement are. None of the scholars analyzed in this feature or any economists of renown have ever publicly argued that tariffs imposed ex-post can help address the effects of a previous trade shock. It is not even clear they would effectively shield workers if they were in place to begin with. And certainly, tariffs do little to solve the woes of American workers who have lost their jobs.  

Most scholars agree that higher education and worker retraining programs, along with government transfers, would be the most effective way to address the issue. For example, Autor, Dorn, and Hanson discuss the potential of the Trade Adjustment Assistance program, which was too small to ever have a significant impact. This summer, Congress failed to reapprove the program, which had become increasingly small in recent years. A better-funded and comprehensive initiative to help displaced workers—regardless of whether they lost their jobs due to trade or technological change – could be an important approach to consider.

Gordon Hanson has argued convincingly that more local experimentation and investment in local labor market programs for young and disadvantaged workers could be very effective. It may also make these regions more resilient to future changes in trade patterns and may have a more lasting impact than simply trying to reshore manufacturing. Stronger social safety net programs may also help communities confront concentrated job losses.  

All economists who have looked at the issue agree that high shares of college education helped regions in the United States respond better to import competition with China and benefit from trade. This points to the importance of enabling larger portions of the population to access better education. As the future of jobs and import competition changes, it will be important to ensure that workers also have access to the right kind of education to prepare them for a changing job market. Education that enhances flexibility and literacy in STEM, for example, might be particularly valuable.  

The United States is actively investing in advanced manufacturing for semiconductors, clean energy technologies, and more broadly in promoting innovation in frontier technology like AI. This year, the United States Congress and the executive branch have overseen the passage of several bills and initiatives to help boost U.S. infrastructure and manufacturing. This should be welcome, as it can help create jobs, especially in areas that continue to struggle economically. It is not yet clear, however, how many jobs and specifically manufacturing jobs will result from these projects. Production is increasingly automated and factory jobs may no longer look like they did in the 1970s. Moreover, these industries will require skilled workers to succeed and the type of high school education that is common today may not be able to prepare students for the workplace adequately. The factory in the 21^st century is different from what it was in the 20^th, and manufacturing jobs have changed as well.  

Washington is aware of these challenges but solutions are still slow in coming. For example, the “Investing in Tomorrow’s Workforce Act” of 2021 tried to address the challenge posed by automation by providing more training but it has not been passed either by the House or the Senate. The CHIPS and Science Bill which was passed this summer recognizes the importance of human capital. To that end, the Department of Commerce has encouraged project proposals that bring together employers, labor unions, and training providers to develop workforce training programs and apprenticeships to help strengthen the talent pipeline. This is a start, but most of the investment in the bill goes towards loans, and R&D, not education and training. While this investment is crucial, more investment in the educational infrastructure of the country will also be important moving forward. Moreover, the semiconductor industry is not the only one that will need flexible, skilled workers in the coming years. 

There is also a real risk that higher-end service sector jobs, for example, chip or software design, may also be more vulnerable to competition with China and other countries. This raises the question of how to think about employment more broadly. Job creation and destruction will likely follow swifter cycles, as technology advances and will require workers and employers to have an ever-evolving mix of skills. These might include tech literacy and fluency in a foreign language as well as the ability to adapt quickly to changing circumstances over the course of one’s career. Policymaking should adjust accordingly to provide the most flexible solutions that help workers in the short term while enabling long-term economic growth by focusing on education, training, and displaced worker support programs.