Should We Bring Manufacturing Home?

How trade works, where the factory jobs went, who pays for tariffs, and when depending on other countries is a real risk

How Trade Works

“Build it here, bring the jobs home, and stop depending on other countries” is an appealing idea, and people on the left and the right both make it. To test it, start with how economists describe a market economy.

Trade is the same deal across a border. You don’t grow your own food or build your own car. You specialize in what you do well, sell it, and buy the rest. Countries gain the same way: each does best by making what it gives up the least to make, its comparative advantage, and trading for the rest. A country can depend on others the way you depend on your grocery store. The useful question is not “Are we dependent?” but “Dependent on whom, for what, and with what backup?”

Work through five questions whenever someone proposes bringing production home:

The Reshoring Checklist
The Reshoring ChecklistWhat to look forIn the factory debate
1. The goal. Is it more jobs, more output, or more security?Each goal calls for a different policy, and one can work against anotherFactories that add output with robots add few jobs
2. The cause. What actually changed?Productivity growth, foreign competition, and changes in what people buyOutput doubled since 1979 while factory jobs fell
3. Seen and unseen. Who gains, and who pays?Benefits that are easy to see; costs spread across many buyersJobs saved at a steel mill; higher costs for everyone who buys steel
4. Inputs. Is the import a finished good or a part?Imported materials, parts, and machines used by U.S. producersU.S. automakers and machinery makers buy imported steel and parts
5. Responses. What do others do next?Retaliation by other countries; shifts by firms and buyersOther countries’ tariffs on U.S. farm goods and factory exports

Start Here: What Changed?

Factory jobs peaked in 1979. Compare that year with today:

What Changed?: Measure
MeasureThenNowChange
U.S. manufacturing jobs19.6 million (1979)12.7 million (2026)down about 35%
Share of all U.S. jobs22% (1979)8% (2026)down about two-thirds
U.S. factory output (index)47.6 (1979)99.4 (2026)about doubled
Output per factory worker1.0 (1979 = base)3.2× (2026)more than tripled
U.S. share of world manufacturingabout 25% (1995)15% (2023)still second largest
China’s share of world manufacturing3% (1990)32% (2023)now the largest

Output index: 2017 = 100. Output per worker is the change in output divided by the change in jobs.

What changed, and what didn’t? America did not stop making things. U.S. factories produce about twice what they did in 1979, and the United States is still the world’s second-largest manufacturer. What fell was the number of workers it takes. Meanwhile, China grew from a small producer into the world’s largest.

Running the Checklist on Manufacturing

1 The goal: jobs, output, or security?

“Bring manufacturing home” can mean three different things, and they don’t always go together.

  • More jobs. A modern factory uses far more machines per worker than one from 1979, so bringing production back adds fewer jobs than it once did. Factories already had 522,000 open jobs in August 2026, so for many employers the limit is finding trained workers, not finding work for them. A policy good for one goal may be poor for another.
  • More output. Output is already near its record, about 1 percent higher in September 2026 than a year earlier.
  • More security. Some goods matter for defense and emergencies. That goal is about specific products and reliable suppliers, not factory jobs in general.

2 The cause: where did the factory jobs go?

  • Productivity. Better machines, software, and methods let each worker produce more. Since 1987, manufacturing productivity has grown about 2.1 percent a year. When output per worker rises faster than people buy more goods, fewer workers are needed.
  • Trade, especially the “China shock.” After China joined the world trading system, imports surged. One major study estimates that competition from Chinese imports cost 2.0 to 2.4 million U.S. jobs from 1999 to 2011, counting spillovers to suppliers and local businesses. A 2022 correction to the study’s research design found the effect on factory jobs about half as large, though still significant. The economic historian Joseph Francis argues that, once that correction is fully applied, the study’s findings of harm to unemployment, labor force participation, wages, and household incomes disappear. He also argues that the local results can’t be scaled up to a national total. The original authors have been invited to reply, so treat this as a live debate. The losses were concentrated in particular towns and industries rather than spread evenly.
  • What we buy. As people get richer, they spend a larger share of their income on services such as health care, education, and travel, and a smaller share on goods.

Notice the test. If trade alone explained the decline, output would have fallen with jobs. Instead it doubled. Trade explains a real part of the job losses after 2000 and where they landed; productivity explains most of the long decline. Both are true, and both matter.

3 Seen and unseen: do tariffs protect factory jobs?

A tariff is a tax on imports. It is the most common tool for bringing production home, and it produces two kinds of effects:

  • The seen. After the 2018 steel and aluminum tariffs, U.S. steel imports fell 24 percent and U.S. steel production rose about 2 percent, adding about $1.3 billion in output.
  • The unseen. Steel prices rose, and industries that use steel, such as autos and machinery, produced about $3.5 billion less. Federal Reserve economists found that the 2018–19 tariffs were linked to about 2.7 percent fewer manufacturing jobs overall: protection added a little, but higher input costs and other countries’ retaliation subtracted more.
  • Who writes the check. Importers pay the tariff, and studies find that U.S. firms and consumers bore nearly all of it in 2018–19 and about 90 percent of it in 2025. Federal Reserve researchers estimate the 2025 tariffs raised prices of goods other than food and energy about 3.1 percent.

Manufacturing employment fell during 2025 as tariffs rose. It has been rising since December 2025, up 72,000 by September 2026.

4 Inputs: many imports feed U.S. factories

Many imports are not finished products for shoppers. They are parts, materials, and machines for American producers. In 2025, U.S. imports of capital goods, such as machinery and computers, reached a record $1.1 trillion. Census Bureau researchers estimate that about 60 percent of a typical manufacturer’s imports are inputs used in its own production. A tariff on those inputs raises the cost of making things in America, which is the opposite of the goal.

5 Responses: what happens next?

  • Other countries retaliate. When the United States raises tariffs, trading partners often answer with their own tariffs on U.S. exports, including farm goods and manufactured products. Retaliation was one of the main reasons the 2018–19 tariffs cost manufacturing jobs.
  • The trade deficit doesn’t simply shrink. The overall trade deficit reflects a gap between what Americans save and what they invest; when the U.S. invests more than it saves, foreign money fills the gap, and that money comes back as imports. In 2025 the average U.S. tariff rose from about 2.6 to 13 percent, yet the trade deficit was $901.5 billion, almost exactly the same as in 2024.
  • The law responds, too. In February 2026 the Supreme Court ruled that an emergency-powers law does not authorize tariffs. Tariffs under other laws remain, including 50 percent on steel, aluminum, and copper.

WATCH Tariffs and Protectionism and Arguments Against International Trade (Marginal Revolution University)

What Can Government Do?

What Can Government Do?: Tool
ToolHow it worksThe catchWhat the evidence shows
Broad tariffsTaxes imports to make foreign goods more expensive.Raises costs for U.S. buyers and factories that use imports, and invites retaliation.2018–19 tariffs were linked to fewer factory jobs overall
Targeted subsidiesPays firms to build in chosen industries, such as chips.Taxpayers pay, and government must pick which industries and firms to support.Can build capacity for goods tied to security; costly per job
Friend-shoring and stockpilesShifts supply to allies, lines up backup sources, and stores critical goods.Allies are not always the cheapest suppliers, and storage costs money.Spreads risk without making everything at home
Help displaced workersTraining, job search help, and income support for workers who lose jobs to trade.Costs taxpayers and doesn’t save the old job. The federal program for trade-displaced workers stopped taking new workers in 2022.Helps people, not particular jobs
Lower the cost of building hereSpeeds permits, trains workers, and lowers energy and input costs.Slow and unglamorous. U.S. factories had 522,000 open jobs in August 2026.Helps every manufacturer, not a chosen few

Fair Challenges

Decide what you think of each challenge first, then open the response.

“Tariffs protect American manufacturing jobs.”
In the protected industry, they can: after the 2018 tariffs, U.S. steel production rose. But industries that use steel lost about $3.5 billion in output, more than twice what steelmakers gained, and U.S. exporters faced retaliation abroad. Federal Reserve economists found the 2018–19 tariffs were linked to fewer manufacturing jobs overall, and factory employment fell during 2025 as tariffs rose. Supporters argue tariffs give leverage in trade talks and time for industries to rebuild. That is worth testing: did other countries lower their barriers, and did the tariffs come off?
“We can’t depend on rivals for things we need in a crisis.”
This is the strongest argument, and it is fair. Advanced chips are a clear case. But making something at home is not the same as making it reliable. In 2022, the U.S. made about 98 percent of its own baby formula, mostly by a few firms. When one plant closed, shelves emptied nationwide, and tariffs and rules made imports slow to arrive. The cure was more suppliers, including foreign ones. Security comes from many sources, not one.
“China doesn’t play fair.”
Partly true. A 2018 federal investigation found that China pressured U.S. companies to hand over technology and supported hacking to steal trade secrets, at an estimated cost of $50 billion a year. Economists debate the best response: targeted actions aimed at those practices, or broad tariffs that also hit allies and American buyers.
“Government can pick the industries we need.”
Sometimes, for clearly defined security goods. But picking winners takes knowledge no agency has, and protection tends to outlast its purpose. Writer Richard Fulmer points out that sugar quotas, steel tariffs, and the Jones Act, a 1920 shipping law, “live on.” Their benefits go to a few firms and workers who lobby to keep them, while the costs are spread thin across many buyers. Supporters reply that markets underinvest in security, and that clear targets and end dates can limit the risk.
“Trade wrecked whole towns.”
True for many places. The China shock cost many jobs, and the losses fell hard on particular communities. That is a strong case for helping affected workers and places. It is a weaker case for tariffs today, because rebuilt factories tend to be highly automated and may not return to the same towns.
“A trade deficit means we’re losing.”
A deficit is not a scoreboard. It reflects a gap between U.S. saving and investment; foreigners who sell us goods use the dollars to invest here, in factories, stocks, and government bonds. The deficit barely moved in 2025 even though tariffs rose sharply. Some economists worry that steady foreign borrowing carries its own risks.
“Other countries pay the tariffs.”
The evidence says mostly no. U.S. government and Federal Reserve studies found that Americans bore nearly all of the 2018–19 tariffs and about 90 percent of the 2025 tariffs. Federal Reserve research also finds that the burden, as a share of income, is heaviest on low-income households.
“Factory jobs are better jobs.”
Many are good jobs without a four-year degree, and the average factory workweek is about 40.6 hours. But the average factory hourly wage, $36.92 in September 2026, was slightly below the private-sector average of $37.81. And factories already had 522,000 open jobs in August 2026.
“It was robots, not trade.”
Both. Productivity explains why output doubled with fewer workers. But productivity growth slowed to almost zero from 2007 to 2019, and studies of the China shock show trade cost jobs in specific places. Blaming only one makes it easier to dismiss the other.

LISTEN The Economics of Tariffs and Trade (with Doug Irwin), Why Industrial Policy Is (Almost) Always a Bad Idea (with Scott Sumner), How Did America Build the Arsenal of Democracy? (with Brian Potter), and Revisiting the China Shock (with Joseph Francis) on EconTalk, hosted by Russ Roberts, with transcripts

The Normative Question

Economics can show what each policy does to prices, jobs, output, and risk. It can’t decide how much a society should pay for security, whether to protect particular jobs or help the people who hold them, or whether a stable factory town is worth higher prices for everyone else. Buying the cheapest goods leaves us more exposed to some risks; making more at home costs money and freedom of choice. Every choice, including doing nothing, puts the burden on someone.

Key Terms in Plain Words

Specialization
Focusing on what you do well and trading for the rest.
Comparative advantage
Being able to make something while giving up less than others would.
Productivity
How much a worker produces in an hour.
Reshoring
Bringing production back to the home country.
Tariff
A tax on imports, paid by the importer.
Retaliation
Another country’s tariffs in response to ours.
Intermediate inputs
Parts and materials used to make other goods.
Friend-shoring
Shifting supply to allied or friendly countries.
Trade deficit
Buying more from other countries than we sell to them.
Saving–investment gap
When a country invests more than it saves, which matches its trade deficit.
China shock
The surge of Chinese imports after 2001 and its effect on U.S. jobs.
Industrial policy
Government support aimed at growing particular industries.

Sources