Credit: PHOTOMONTAGE WITH PHOTO BY C P GEORGE/CLASSICSTOCK/GETTY IMAGES
Place-based policies offer regions left behind by globalization a path beyond economic populism
GORDON HANSON is the Peter Wertheim Professor in Urban Policy at Harvard University and a research associate at the National Bureau of Economic Research.
Opinions expressed in articles and other materials are those of the authors; they do not necessarily reflect IMF policy.
One of the most firmly held beliefs in economics is that free trade is good for humanity. Yet that confidence in the economic virtue of open markets can blind the profession to the complications of deep global economic ties. When in the 1990s the world leapt into frenzied globalization, policymakers touted the potential efficiency gains but gave short shrift to possible painful distributional consequences. Those consequences have now come back to bite.
The United States’ embrace of higher import tariffs, China’s pursuit of aggressive industrial policies, and the United Kingdom’s exit from the European Union—all of which occurred around 2016—heralded a more fractious global economic order, which, according to Canadian Prime Minister Mark Carney, is steering the world toward geoeconomic rupture.
It took decades to build the international institutions, negotiate the multilateral agreements, and enact the domestic economic reforms that wrought the modern era of globalization. It could be torn down quickly and in unpredictable ways that upend economies. To avoid such a rupture calls for a clear-eyed look at why globalization has become unpopular worldwide and a credible alternative to economic nationalism that addresses the dislocations integration has caused.
The accession of China, India, and other newly liberalizing economies to the World Trade Organization (WTO) in the 1990s and 2000s lifted hundreds of millions of people out of poverty in one of the greatest improvements in human well-being. This much standard economic analysis foretold: We expected efficiency gains and we got them. What economists failed to foresee was the profoundly disruptive impacts of rapid globalization on labor markets in high-income countries (and some middle-income ones, too).
To be sure, emerging technologies and the rise of services were also disruptive. In concert, they probably determined more recent changes in high-income-country living standards than international trade did. But it was signature changes in trade policy—the formation of the EU in 1992, the enactment of the North American Free Trade Agreement in 1994, and China’s accession to the WTO in 2001—that got the public’s attention. Many voters in Europe and the US came to blame these and related policy changes for their economic discontent, which later morphed into support for nationalist and populist political movements that champion economic isolation.
Most economists missed the disruptive impacts of globalization because they misunderstood three things about how labor markets adjust to large shocks. First, they failed to recognize how long-standing industrial specialization patterns exposed manufacturing regions to rising import competition from China and other developing economies.
Since the Industrial Revolution, manufacturing has been highly concentrated geographically, which led Alfred Marshall, the 19th century political economist, to develop his insights on productivity gains from the spatial agglomeration of economic activity. Concentrated first in major urban centers, manufacturing over the course of the 20th century relocated to smaller and medium-sized industry towns and cities.
When China began its spectacular manufacturing export growth in the 1990s, those industry towns were ground zero for the ensuing trade shock. Import-competing regions of France, Germany, the UK, the US, and other high-income nations bore the brunt of manufacturing job loss caused by deepening import penetration. At the same time, superstar cities in those countries, which specialized in business services, finance, high tech, and other knowledge-intensive industries, saw their exports and incomes boom.
Economic models of the time were tuned to capture national impacts of freer trade. Consequently (and consequentially), they missed the highly uneven regional impacts of globalization. We had not expected winners and losers to be so starkly differentiated by place.
Second, economists did not anticipate the magnitude of the scarring effects from manufacturing job loss. The fetishization of manufacturing by those who espouse economic nationalism is not without reason: The sector has long offered a pay premium relative to what people could earn in other lines of work, especially for those without a college degree. When factories closed or laid off large numbers of employees, as occurred in many high-income countries during the China trade shock, industrial workers lost their pay premium. Most, faced with the alternatives of lower-wage jobs in services or exit from the labor force, never replaced their lost earnings.
Although economists first documented the scarring effects of job loss in the early 1990s, they did not appreciate until later that when job loss is regionally concentrated, individual scarring aggregates into large negative local income shocks. Once displaced, former factory workers spent less on nontraded goods and services, paid less for housing, and contributed less in taxes to support local public services—all of which depressed incomes where manufacturing was declining.
Again, because economic models were tuned to account for national-level adjustment to international trade, they projected that workers displaced in import-competing sectors (manufacturing) would simply shift into sectors in which exports were expanding (knowledge-intensive services). Import displacement and export absorption did occur, but among largely disjointed sets of people.
Third, economists overlooked the lack of geographic labor mobility among less-educated and older workers in response to changing economic conditions. Standard economic models posit a spatial equilibrium condition: If real earnings rise or fall in one region, the migration of labor between regions will smoothly arbitrage away spatial differences in pay. In theory, labor mobility transmits localized economic shocks to other regions, dissipating shock impacts and ensuring that regional spikes in joblessness are temporary. In practice, interregional migration works slowly: Spatial equilibration to shocks can take decades.
The slowness of regional migration has been one of the hardest lessons for economists to internalize. Surely, in economies as large as the US and the EU, where millions of jobs are created and millions destroyed each year, job losses in a subset of industrial regions should be easily offset. Such logic is mistaken—first, because it projects frequent job transitions among younger workers to older workers, who are markedly less agile, and second, because it assumes that upward-sloping job ladders available to more educated workers are equally accessible to the less educated. Modern labor markets are undeniably dynamic. But that dynamism has been least evident among workers most exposed to deindustrialization.
It has been painful for countries to learn of globalization’s dark side, which entails widening regional economic disparities and a tendency for former industry towns to get stuck with high joblessness and few well-paying jobs for less-educated workers. In the moment globalization was causing manufacturing job losses, countries had viable policy options to cope with the disruption, including generous and immediate assistance for displaced workers and safeguard tariffs that would have spread import surges over longer periods of time.
Two decades later, such policies are neither available nor relevant. Countries are left to decide whether and how to address regional economic distress caused by globalization, long after the distress materialized. Selecting the right policies calls for a clear understanding of the economic problems countries are trying to solve.
One option to help left-behind regions is simply to let market forces do their thing. The outmigration of labor and the retirement of displaced workers would ultimately help forsaken regions shrink to a smaller, more efficient size. Businesses would close, downtowns would board up buildings, and young labor force entrants would launch their careers elsewhere. If we believe that there are no economic distortions that impede labor market adjustment to adverse shocks, or if we believe that governments cannot implement effective remedies for such distortions, then laissez-faire may make sense. It is important to recognize, however, that although market forces may alleviate spatial differences in economic well-being, they are likely to do so ever so slowly.
Consider the experience of Pittsburgh, which many cite as a successful example of adjustment to deindustrialization. Through the first half of the 20th century, the city was a global center for steel manufacturing. After 1970, import competition, technological change, and other forces contributed to a prolonged industrial decline, during which joblessness and economic hardship were endemic. Although Pittsburgh today is home to health care, life sciences, and robotics, its transformation took more than a full generation. During this period the economic opportunity of local people was circumscribed by distress. For every Pittsburgh, there are several other former industry towns that did not find a path back to prosperity. Long-term adjustment meant diminished incomes, housing prices, and urban amenities.
A second option to help left-behind regions is to target affected individuals through means-tested entitlement programs. Unemployment insurance, income support for low-income households, housing and energy subsidies, and subsidized health care are common ways to help people in hard times. If we don’t have confidence that insurance and credit markets can insulate people against adverse shocks, then expansive social insurance programs may make sense.
Yet such programs condition assistance on individual or household well-being—not on the state of the local labor market. Social insurance may help people avoid sharp declines in consumption during episodes of hardship, but they do not address the causes of regional economic distress. Such programs may make economic adjustment less painful, but they are unlikely to speed adjustment along.
A third option to help left-behind regions is to target the industries whose decline was responsible for economic distress. Recent US import tariffs, for instance, have been justified in part by the claim that they will help bring manufacturing jobs back to communities hollowed out by globalization. On the surface, addressing the negative labor-market consequences of trade by blocking imports may appear sensible. However, the origin of regional distress is not import competition per se but the scarring effects of job loss and regions’ inability to adjust to decline in a major industry.
Import tariffs would not prevent job loss from technological change, artificial intelligence, or other shocks that may be disruptive in the future. Trade protection targets economic distress indirectly and therefore poorly. It is not surprising that US tariffs have done little to restore manufacturing employment or earnings growth in regions harmed by the China trade shock.
A final option is to target left-behind regions through policies that promote local economic development. Place-based policies subsidize investment in human and physical capital with the aim of upgrading labor productivity, earnings, and economic structures in distressed regions. In theory, such policies are justified if the social return on investment is relatively high in communities with elevated joblessness and low wages. In practice, place-based policy has long been controversial among economists, given concerns over rent capture by special interests and informational challenges in program design.
Recent empirical research clarifies where place-based policy works well and where it does not. Less-effective policies—also among the most visible—include the use of tax subsidies to compete for major investment by large firms. Subsidy competitions tend to transfer most of the economic surplus from new investment to the investors themselves, leaving “winning” regions with a winner’s curse of high tax expenditure per job created.
More effective place-based policies condition subsidies on local distress, monitor compliance with program objectives, and design programs for specific contexts via ongoing experimentation. Examples include tax incentives for investing in low-income communities (such as enterprise zone programs with rigorous selection and auditing of participating firms) and sectoral worker training initiatives (active labor market programs, which have been applied in many high- and middle-income countries).
Of these four options, only place-based policy directly targets outcomes that alleviate regional distress: moving out-of-work adults into employment, bringing more well-paying jobs where they are scarce, and making regions more attractive for future investment.
Public opinion has soured on globalization in high-income countries because many middle- and low-wage workers ended up on the losing end of economic openness. The prosperous future they were promised did not arrive. To restore faith in global economic integration we must correct the mistakes of the past and offer a credible alternative to economic nationalism.
