In the final stretch of the U.S. presidential race, Donald Trump and Kamala Harris are both touting competing plans to create middle-class jobs for workers without college degrees by revitalizing manufacturing. But the candidates are not only playing on the electorate’s nostalgia for a bygone era; they are ignoring the diminished role that manufacturing can now play as a source of growth and opportunity.
Trump proposes to eliminate the U.S. deficit in manufacturing trade by erecting high tariffs. He blames the long-term decline in American manufacturing employment on bad trade agreements and unfair practices by other countries, especially China. Closing off the economy with trade barriers, he argues, will reverse the trend and generate large increases in. manufacturing jobs.
For her part, Harris wants to double down on the Biden administration’s industrial policies by introducing an additional $100 billion in federal subsidies to the manufacturing industries of the future.
These policy prescriptions are remarkably myopic. They ignore the fact that the decline of manufacturing as a share of total employment—from 30% in the 1970s to barely 8% today—reflects long-term structural forces such as automation, productivity gains, and a shift in demand from goods toward services (which happens naturally as economies develop).
There have been similar declines in other advanced and emerging-market economies, even those with long-standing manufacturing trade surpluses, including Germany, Japan, Singapore, South Korea, and China. Such trends call into question Trump’s belief that a smaller trade deficit is the key to spurring more manufacturing employment.
Moreover, the employment share of manufacturing is declining in mature economies regardless of their overall industrial-policy approaches. The trend is apparent both in economies that have adopted free-market policies (Hong Kong, the United States, the United Kingdom, and Germany) and in those with interventionist policies (Japan, South Korea, and China). Again, such evidence casts doubt on the claim that neoliberal policies are to blame for lost manufacturing jobs, or that more interventionist industrial policies can significantly reverse the decline.
Similar lessons can be drawn from America’s own experience. Trend declines in manufacturing’s employment share were evident in the U.S. as early as the 1950s and 1960s, when international trade played a minimal role in the American economy. Moreover, they continued between 2019 and 2024. Even though the Biden administration maintained Trump’s tariffs and pursued more interventionist industrial policies, jobs in manufacturing grew by a meager 1%.
This trend is likely to continue. Even after the Biden administration’s policies had passed, in 2023 the U.S. Department of Labor continued to forecast a declining U.S. manufacturing employment share over the next decade.
All the evidence points to deep and powerful forces that drive the long-term decline in manufacturing’s share of jobs and GDP as countries become richer. Historically, as an economy evolves, the share of manufacturing employment follows an inverted U pattern: after rising in the initial stages of development, it then declines as development progresses.
Many countries are now moving down the right side of the curve. The trends in their manufacturing employment resemble those that prevailed earlier in agriculture. As farming becomes more productive, fewer farmers are needed, because demand increases only so much in response to cheaper food. Similarly, as manufacturing becomes more productive, goods become cheaper; but since demand for goods does not rise proportionately people spend more of their income on services, and fewer manufacturing workers are needed. In both cases, exports can generate some additional demand, but not enough to sustain employment growth forever.
Meanwhile, technological changes have shifted employment demand in manufacturing toward more educated workers. Older Americans may remember the days (decades ago) when factory jobs provided good pay and benefits to many workers with a secondary education or less. But manufacturing jobs today increasingly require workers with at least a college degree, and this shift seems set to continue.
A central objective of industrial policies in both advanced and developing economies is to promote new manufacturing technologies. Many countries want to master labor-saving digital technologies such as robotics and additive manufacturing (3-D printing), skill-intensive technologies such as nanotechnology and advanced materials, and green technologies such as electric vehicles (which are basically computers on wheels). While the full implications of artificial intelligence are unknown, most of its uses are likely to increase the requirements for more skilled and educated workers.
That is why the next administration should emphasize policies that promote growth and inclusion throughout the economy. This means work-adjustment and job-placement assistance; wage-loss insurance for displaced workers; and grants for skills training and apprenticeship programs to equip workers for jobs and careers offering middle-class incomes. And U.S. structural policies should be expanded and tailored to community needs.
To be sure, manufacturing has a vital role to play in providing the hardware for the digital economy, the semiconductors for AI (where domestic production may be important for national security), and the solar panels, wind turbines, and EVs for decarbonization. But the sector should not be promoted as a vehicle of inclusive growth and employment for low-skilled workers.
Robert Z. Lawrence, a non-resident fellow at the Peterson Institute for International Economics, is a professor at Harvard Kennedy School and the author, most recently, of Behind the Curve: Can Manufacturing Still Provide Inclusive Growth? (The Peterson Institute for International Economics, 2024).