VoxEU Column International trade Labour Markets

Globalisation and US labour markets

There is some evidence that communities hit hardest by globalisation shifted away from centrist candidates towards ideologically extreme candidates in the most recent US election. This column, taken from a recent Vox eBook, asks what policies those who were elected on a promise of turning the tide of globalisation away will implement, and what the prospects of success for these policies are.

Editor's note: This column first appeared as a chapter in the Vox eBook, Economics and policy in the Age of Trump, available to download here.​

Most Americans agree that some trade is good. It would be a shame, after all, if Alaskans could not eat mangoes just because they happen to have settled in a location where it is impossible to grow them. Public debate on the virtues of trade gets more contentious when the discussion turns to goods that are no longer – or that are increasingly less likely to be – produced in the US, such as textiles, toys, furniture and cars.

Most economists emphasise that trade is a win-win for all countries that participate. Trade, the argument goes, improves welfare, benefiting consumers by lowering prices and by increasing the variety of available goods. Overall productivity improves as each country moves away from producing goods that they are, relatively, not competitive in and towards goods in which they are. Thanks to trade, consumers can buy fuel- efficient Toyotas, sporty BMWs, or large Ford SUVs – and pay a fraction of the cost that they would for ‘Made in America’ toys, furniture and textiles by buying cheaper versions imported from Mexico, China, or Vietnam. Trade frees the economy’s workers, capital, and land to specialise in what they are really good at – in the case of the US, advanced manufacturing goods like robots and planes, or services like banking, insurance, and software.

Economists have long-acknowledged, however, that the benefits of trade are unevenly distributed across different segments of the population, and that the destruction and expansion of industries that accompanies trade bears its own risks and costs. Until recently, however, the economic consensus was that these transition costs were likely small, and the gains were large enough to compensate for the losses that some do suffer.

Then came China, with its large population, solid infrastructure, and extraordinary economic growth. China transformed rapidly, moving from 1% of the world GDP in 1980 to 20% in 2010. China’s accession to the WTO in 2001 locked in access to the US market at low tariff rates, especially in unskilled-labour-intensive industries. Importantly, this status also reduced the threat of future protection levied on imports from China, making firms more comfortable with moving some or all of their production to China.

The rapid entry of a large export-oriented country onto the world scene had more severe consequences for American workers and communities than previous globalisation episodes. The outcome has shifted economists’ focus away from the overall gains from trade, towards quantifying the speed of the adjustment process and distributional consequences. The findings so far suggest a slow transition process involving large and concentrated losses for some workers and their communities.

During the 1990s and 2000s, the American industries competing with China experienced more factory shutdowns and lower employment growth in factories that survived, relative to less-exposed industries (Bernard et al. 2006). Workers employed in these industries saw lower earnings (Autor et al. 2014) and higher uncertainty associated with these earnings (Krishna and Senses 2014), relative to workers employed in industries facing less direct competition. While high-skilled workers in these sectors – such as lawyers, HR specialists, and administrative staff – were able to transition to other industries without much loss of earnings, less educated and lower-wage workers bore the brunt of the impact. Some of these workers received federal social assistance in the form of Unemployment Insurance, Medicaid, Medicare, Supplemental Nutrition Assistance Program, Temporary Assistance for Needy Families or Trade Adjustment Assistance. Nevertheless, these benefits fell far short of making up for the severe losses they experienced.

Where a community lost manufacturing jobs, the impact was not limited to the wages and job security of low-skilled factory workers but also extended to non-manufacturing employment in these areas (Autor et al. 2013). When customers have less money to spend, either because they have lost their job or because their level of economic insecurity has increased, local stores, restaurants, barber shops and other businesses also suffer.

The fortunes of a manufacturing factory are further connected to other industries through its supply chain. If a factory that produces furniture shuts down, the adverse consequences spill over to suppliers of wood, plastic laminate, plywood, metal, iron and machinery, and on to companies that store, transport and sell the furniture.

Local governments can play an important role by investing in public services, such as high-quality education and infrastructure, to ensure the competitiveness of workers and firms in their community. The problem is that funding for these public services is highly localised in the US, with a heavy reliance on property and sales tax revenues. Therefore, a decline in the level of local economic activity depresses tax revenues and restricts the ability of local governments to fund public services, precisely at a time when this support is most needed. The outcome is the deterioration of public services that further exacerbates the negative income shock – less spending on public housing, welfare and public transport, higher (property) crime rates, and lower quality schools (Feler and Senses 2017).

Adverse labour market outcomes faced by individual workers, who are in most cases already at the bottom end of the education and wage distributions, and the decline of economic opportunity in their communities, often brings profound personal consequences. A decline in manufacturing decreases the wages of men relative to women, which in turn reduces their attractiveness as marriage partners – the outcome is a plunge in marriage and fertility rates in the hardest hit communities (Autor et al. 2017). There is also evidence of worse health outcomes and higher mortality rates (Pierce and Schott 2016), and an increase in the number of children born out of wedlock and to teens.

Two very important caveats are missing from this particularly bleak narrative.

  • First, none of these findings suggests that freer trade does more harm than good. None of the studies mentioned above accounts for benefits of trade, not only to consumers in the form of lower prices and more variety, but to firms in services and in advanced manufacturing that benefit from access to the world market by exporting their products and by importing cheap inputs, machinery and raw materials. A decline in prices of tradable goods and a decline in housing prices in exposed localities, at least partially, will curtail the impact on purchasing power.
  • Second, while trade has been a contributor to these negative trends, especially during the 2000s, by no means has it been the primary contributor. Automation of jobs by computers and robots, although less prone to producing zingy one-liners for politicians than trade with China and Mexico, had a much bigger impact on the adverse economic outcomes that factory workers without a college education have experienced. It’s plausible that even jobs whose elimination could directly be linked to China’s WTO accession would have eventually been replaced by machines instead of by Chinese workers. Fast food CEO and former Labor Secretary nominee Andy Puzder recently described why in an interview with Business Insider: ‘[Machines are] always polite, they always upsell, they never take a vacation, they never show up late, there’s never a slip-and-fall, or an age, sex, or race discrimination case’ (Business Insider 2016). It is unlikely that policies to reverse the trend in which manual labour jobs are replaced by cheaper, faster, and more efficient robots, would gain much traction in Washington.

Alas, with such a bleak economic backdrop, nuanced discussion of trade’s pros and cons does not resonate, especially in election years. There is some evidence that communities hit hardest by globalisation have shifted away from centrist candidates towards ideologically extreme candidates in the most recent US election. The shift took place in both ends of the political spectrum: more moderate Republicans were replaced by more conservative candidates and more moderate Democrats by more liberal candidates (Autor et al. 2016). Now, the question is: what policies will these officials – who were elected on a promise of turning the tide of globalisation away – implement? And what is the prospect of success for these policies?

Initial signs are worrying. President Trump consistently promised, on the campaign trail, to restrict trade. This will hurt American consumers, in particular those with low incomes as they spend a greater percentage of their income relative to their richer counterparts on the kind of goods that have declined the most in price, thanks to trade (Fajgelbaum and Khandelwal 2016). They will end up paying more for ‘Made in America’ versions of products previously imported from other countries. This policy will almost certainly trigger retaliation by trading partners, which will result in lower demand for American exports and possibly lead to a global slow-down. Additionally, given the interconnectedness of US and Chinese economies, policies that restrict trade with China will disrupt the supply chains of many American firms which rely heavily on cheap intermediate inputs and raw materials from China.

So far, the only policy implemented on this front is tearing up the Trans-Pacific Partnership (TPP) – a 12-nation trade deal negotiated by President Obama – to which China was not a signatory country. This was an act vigorously supported by the Chinese, as it has the potential to replace Americans with their Chinese counterparts in setting up the rules of the trade relations in Asia.

Importantly, even if Trump’s protectionist policies are implemented and manufacturing companies are successfully incentivised to move back to the US, their new factories are more likely to look like the BMW plant in Spartanburg, SC than a 1970s Ford plant in Detroit, MI – that is, located in a Southern right-to-work state without much regulation, with a factory floor populated not by blue-collar workers but by robots and a few relatively high-skill and high-wage workers who are machine operators, technicians and engineers. This is unlikely to help the plight of workers living in trade-impacted communities elsewhere.

Instead of trying to reverse the trends in globalisation and technological development at the expense of the communities, firms, workers, and consumers who benefit from these trends, a more productive approach would involve reducing the harm to those who lost their livelihoods during this process.

  • First, a well-functioning and adequately funded set of social safety net programmes would buffer against some of the losses faced by these workers and ease their transition to new employment outside their industries.

Expanding eligibility criteria and increasing the generosity of temporary assistance programmes such as Unemployment Benefits, Supplemental Nutrition Assistance Program and Temporary Assistance for Needy Families would help cushion the initial blow of a job loss.

  • Second, while a short-term cushion is important in an environment in which trade shocks are more frequent, unanticipated, and harder to insure against, policies that get people back to work as soon as possible are preferable.

This is not only because of the efficiency losses due to unemployment, but because a job is more than just a paycheck for many workers – it helps create a sense of purpose, provides an often-needed structure, gives the sense of a solid place in one’s community, and is a part of one’s identity.

The cost of losing a job is more than just losing the financial benefits of having a job. Recent research links decline in labour force participation of men, especially of white men without a college degree, to high rates of ‘deaths of despair’ due to suicide and drug and alcohol abuse (Case and Deaton 2017). Retraining subsidies, federal job guarantees, wage insurance programmes, and subsidised loans for displaced workers going back to college are all policies that could help younger workers transition to new sectors that require a very different set of skills than their old job, and help older workers transition into retirement.

  • Third, eliminating local government policies that serve as a barrier to migration and incentivising workers to move to more prosperous communities would be beneficial.

One such barrier is artificial restrictions on housing supply, like zoning laws and building restrictions that increase housing prices. These policies tend to place a heavier burden on low-income households who spend a larger share of their budget on housing.

Another barrier to mobility is occupational licensing laws. Currently, about one third of workers in the US need a state-issued license to work. The time and money cost of acquiring a license in a new state varies widely and serves as a barrier to relocating, especially for low-income households with at least one earner working in an occupation with a licensing requirement, such as a hair-dresser, make-up artist or a child-care provider.

Moving towards standardisation of the eligibility requirements for government assistance programmes, making it less onerous to transfer benefits and health insurance across states lines, and providing relocation vouchers for needy households would also incentivise workers to move away from low-opportunity communities by reducing the risk and pecuniary costs associated with moving.

The current administration has so far shown little appetite for expanding government social safety nets and easing the transition cost for displaced workers. One of the first legislative actions of the Trump administration was an attempt to repeal the Affordable Care Act – a system designed to insure low-income families and provide an important buffer against loss of employer-provided health insurance following a job loss.

Similarly, Trump’s first budget plan involves deep cuts across government agencies and eliminates many federal programmes that assist the poor (Low-Income Home Energy Assistance, Community Development Block Grant Program, Section 4 Capacity Building for Community Development and Affordable Housing Program, to name a few) and programmes that support education and training (21st Century Community Learning Centers Program, Federal Supplemental Educational Opportunity Grant Program, Striving Readers Comprehensive Literacy Program and the Senior Community Service Employment Program). Also, there is a government-wide hiring freeze in place that is likely to make the implementation of policies already in place more difficult.

One area where there is some possibility of reform is the reduction of existing regulations. If some of these reforms serve to reduce barriers to mobility across localities, they have the potential to speed up the adjustment process. However, it is unlikely that this will be enough to end the ‘American carnage’ that President Trump described in his inaugural speech.


Autor, D., D. Dorn, and G. Hanson (2013), “The China Syndrome: Local Labor Market Effects of Import Competition in the United States”, American Economic Review 103(6): 2121–2168.

Autor, D., D. Dorn, and G. Hanson (2017), “When Work Disappears: Manufacturing Decline and the Falling Marriage-Market Value of Men”, NBER Working Paper #23173.

Autor, D., D. Dorn, G. Hanson and J. Song (2014), “Trade Adjustment: Worker Level Evidence”, Quarterly Journal of Economics 129(4): 1799–1860.

Autor, D., D. Dorn, G. Hanson and K. Majlesi (2016), “Importing Political Polarization? The Electoral Consequences of Rising Trade Exposure”, MIT Working Paper, December, 2016.

Bernard, A., J. Bradford Jensen and P. K. Schott (2006), “Survival of the best fit: Exposure to low-wage countries and the (uneven) growth of US manufacturing plants”, Journal of International Economics 68 (1): 219-237.

Business Insider (2016), “Fast-food CEO says he’s investing in machines because the government is making it difficult to afford employees”, March 16, 2016.

Case, A. and A. Deaton (2017), “Mortality and Morbidity in the 21st Century”, Brookings Papers on Economic Activity, Spring 2017.

Fajgelbaum, P.D. and A.K. Khandelwal (2016), “Measuring the Unequal Gains from Trade”, Quarterly Journal of Economics 131 (3): 1113-1180.

Feler, L. and M. Zeynep Senses (2017), “Trade Shocks and the Provision of Local Public Goods”, American Economic Journal: Economic Policy, forthcoming.

Krishna, P. and M. Zeynep Senses (2014), “International Trade and Labor Income Risk in the United States”, Review of Economic Studies 81(1): 186-218.

Pierce, J. and P.K. Schott (2016), “Trade Liberalization and Mortality: Evidence from

U.S. Counties”, NBER Working Paper, 22849.

3,570 Reads