Trump Immigration Crackdown Cost 668,000 Jobs, New NBER Study Finds, And American Workers Paid the Price

The 668,000 Job Shock, What the Numbers Actually Say

668,000 jobs. Gone.

That's not a projection from a think tank with an agenda. That's what researchers at the National Bureau of Economic Research, the same institution that officially declares recessions, calculated when they ran the numbers on the first nine months of Trump's second-term immigration crackdown.

To put it in perspective: 0.7% of total employment in the United States vanished not because of a recession, not because of a trade war, but because of immigration enforcement. The paper, titled Labor Market Impacts of ICE Activity in Trump 2.0, was authored by Elizabeth Cox and Chloe N. East of the University of Colorado Boulder. It is the first national, causal study of its kind. And the findings are, frankly, staggering.

Where the 668,000 Figure Comes From

Here's how the math works, and it's worth understanding because this isn't guesswork.

The researchers used ICE arrest data from the Deportation Data Project, covering October 2023 through October 2025. They identified geographic areas that experienced sudden, large spikes in immigration arrests (the "treated" areas) and compared them to areas with smaller increases (the "control" areas). This is the gold-standard "difference-in-differences" method, the same approach medical researchers use to prove vaccines work.

In treated areas, they found that employment among likely undocumented immigrants dropped by 4%. For men, who make up over 90% of ICE arrests, the drop was 5%. But here's the kicker: these are immigrants who remained in the United States. They weren't deported. They were scared.

That's the "chilling effect." And it's massive.

0.7% of Total Employment, Putting It in Context

If 0.7% sounds small to you, think of it this way: that's roughly the entire workforce of Wyoming, Vermont, and North Dakota combined. Poof. In nine months.

And it happened quietly. No headlines. No factory closures on the evening news. Just workers not showing up. Projects stalling. Businesses shrinking. And American-born workers, the very people this policy was supposed to help, getting caught in the crossfire.

---

Inside the NBER Study, How Economists Measured the Damage

Let's be honest: a lot of economic studies feel like they're written to be unreadable. Not this one. The methodology is elegant, and understanding it makes the findings impossible to dismiss.

The "Difference-in-Differences" Approach (Treated vs. Control Areas)

Imagine you want to know if a new fertilizer helps corn grow. You don't just look at one field after the fertilizer arrives, you compare it to a nearby field that didn't get the fertilizer. Same soil, same weather, different treatment.

That's what Cox and East did. They compared labor markets that got hit hard by ICE raids to similar labor markets that didn't. Both sets of areas experienced the same national economy, the same interest rates, the same tariff news. The only major difference? The intensity of immigration enforcement.

The result? Areas with sudden ICE surges saw employment crater. Control areas didn't. That gap is the causal effect.

Who Counts as "Likely Undocumented"?

Since no U.S. survey asks directly about immigration status, the researchers used a proxy: foreign-born individuals with a high school education or less, working in sectors where undocumented workers are overrepresented, agriculture, construction, manufacturing, and wholesale.

It's not perfect, but it's the standard approach in peer-reviewed immigration economics. And the results hold up even when they tweak the definition.

Why This Is the First National, Causal Evidence

Plenty of articles have speculated about Trump's immigration crackdown. Local reports from Texas, Minnesota, and Chicago have documented business disruptions. But this is the first study to link national ICE enforcement data to national labor market outcomes with a causal design.

As Chloe East told NPR: "The mass deportations in Trump 2.0 are not helping the labor market overall and not creating more job opportunities for U.S.-born workers."

That's not an opinion. That's what the data shows.

---

The Chilling Effect, Fear Is More Powerful Than Deportation

Here's the part that should make policymakers pause. The 668,000 job losses aren't primarily from deportations. They're from fear of deportation.

For Every 1 ICE Arrest, 6 Workers Stop Showing Up

In treated areas, the researchers found that for every single ICE arrest, roughly six likely undocumented male workers stopped working. Not because they were caught. Because they were terrified.

Think about that ratio for a second. One arrest. Six workers staying home. It's like a multiplier effect in reverse, instead of economic stimulus creating ripples of growth, a single enforcement action creates ripples of economic withdrawal.

These workers didn't quit because they found better jobs. They didn't move to another state. They just... stopped. They pulled out of the labor market entirely, reducing their spending, their tax contributions, their economic footprint.

Why Trump 2.0's Chilling Effect Is 2.6x Worse Than Obama's

This isn't the first time researchers have measured chilling effects. East herself co-authored a study on the Obama administration's enforcement surge. Back then, for every person detained, 2.3 likely undocumented workers stopped working.

Under Trump 2.0? That number jumped to 6.

Why the difference? The researchers suggest it's the increasingly indiscriminate nature of ICE activity. When enforcement feels random, when anyone, anywhere, might be swept up, the rational response is to hide. To stay home. To stop participating in the above-ground economy.

It's the difference between a targeted raid and a neighborhood thunderstorm. You can avoid a raid. You can't avoid thunder.

---

When American Workers Lose Too, The Complementarity Problem

This is where the story gets politically uncomfortable. Because the whole justification for the crackdown, the promise repeated at rallies, in press releases, in congressional hearings, was that removing undocumented workers would open up jobs for Americans.

The data says: not only did that not happen, but the opposite occurred.

The 6-to-1 Ratio Explained

For every six undocumented male workers who stopped working in treated areas, one U.S.-born male worker with a high school education or less also lost employment. That's a direct spillover.

How? Because the labor market isn't a zero-sum game. It's not a pie with fixed slices.

Undocumented immigrants and U.S.-born workers in these sectors are what economists call complements, not substitutes. They don't compete for the same jobs, they make each other's jobs possible.

Construction's -7.5% Collapse

Construction was the hardest-hit industry in the study. Employment among likely undocumented construction workers fell 7.5% in high-enforcement areas.

Here's the metaphor Chloe East uses: when a construction company can't find laborers, the guys who pour concrete, carry drywall, clear sites, they don't just hire American electricians instead. They build fewer homes. Fewer commercial buildings. Which means they need fewer electricians. Fewer roofers. Fewer project managers. Fewer permit expediters.

The whole chain contracts.

Robby Robertson, a construction superintendent in Mobile, Alabama, told Reuters that half his workers stopped showing up after an ICE raid 230 miles away in Florida. His $20 million recreation center project delayed three weeks. Cost: $84,000 in liquidated damages, plus $4,000 per day.

That's not theory. That's a payroll.

No Wage Increases, No Replacement Workers

Here's another finding that should puncture a common myth: employers did not raise wages to attract U.S.-born workers to fill the gaps.

You'd think, if supply suddenly dropped, prices (wages) would rise. They didn't. The study found no evidence of wage increases in affected sectors. What happened instead was a reduction in overall labor demand. Employers didn't pivot; they retrenched.

Why? Because many of these jobs, seasonal agricultural work, dangerous construction labor, irregular-hour manufacturing, aren't just low-paid. They're structurally unattractive to the existing U.S.-born workforce. You can't just raise wages 10% and expect Americans to suddenly want to pick strawberries in 105-degree heat or hang drywall at 6 a.m. for six months straight.

The jobs don't get filled. The work doesn't get done. And the businesses that depend on that work shrink.

---

What the White House Says, And What the Data Answers Back

The Administration's Defense

The White House didn't ignore the study. Spokesperson Abigail Jackson told Axios: "There is no shortage of American minds and hands to grow our labor force. President Trump's agenda to create jobs for American workers represents this Administration's commitment to capitalizing on that untapped potential."

It's a compelling narrative. It feels intuitively right. If there are fewer workers competing for jobs, Americans should benefit. Econ 101, right?

Except labor markets aren't Econ 101. They're Econ 401, with network effects, complementarities, search frictions, and sectoral interdependencies.

Why "Untapped Potential" Doesn't Show Up in the Data

If the White House theory were correct, we'd expect to see:

• Rising labor force participation among U.S.-born workers
• Falling unemployment among native-born men without college degrees
• Wage growth in construction, agriculture, and manufacturing

Instead, the Bureau of Labor Statistics data shows the opposite. The U.S.-born labor force participation rate fell from 61.4% in January 2025 to 61% by February 2026. Unemployment for U.S.-born workers rose from 4.3% to 4.7%.

The "untapped potential" stayed untapped. Because potential isn't the same as prepared, willing, and geographically available workers ready to take specific jobs at specific wages in specific industries.

---

The Bigger Economic Picture, Beyond the 668,000

The NBER paper is devastating in its precision, but it's not the only warning sign flashing red.

Negative Net Migration for the First Time in 50 Years

The Brookings Institution estimates that net migration to the United States was between -10,000 and -295,000 in 2025. Negative. For the first time in at least half a century, more people left the U.S. than entered it.

That's not a policy success. That's a demographic alarm bell. The U.S. population is aging. Births are declining. Deaths are projected to exceed births by 2030. Without immigration, the workforce shrinks. And when the workforce shrinks, so does the economy.

The $1.9 Trillion GDP Warning

The National Foundation for American Policy projected that Trump's combined legal and illegal immigration policies would reduce GDP by $1.9 trillion between 2025 and 2028. By 2035, the cumulative loss hits $12.1 trillion, or $34,369 per person.

That's not a typo. Trillion with a T.

The analysis also projects 6.8 million fewer workers by 2028 and a federal debt increase of $252 billion. All while the administration spends roughly $170 billion on immigration enforcement, including $75 billion for ICE through 2029.

We're paying premium prices to shrink the economy.

What This Means for Housing, Food Prices, and Your Town

Even if you don't work in construction or agriculture, this reaches you.

• Housing: Fewer construction workers = fewer homes built = higher prices. The National Association of Home Builders has already flagged immigration restrictions as a key driver of housing supply constraints.
• Food: The American Farm Bureau Federation reports persistent labor shortages in fruit and vegetable harvesting. Less labor = more spoilage = higher grocery bills.
• Small businesses: In Los Angeles County, a survey of 311 local businesses found 82% reported negative effects from immigration enforcement, with 52% seeing reduced daily sales. Bus ridership in high-vulnerability areas dropped 17,000 monthly riders.

This isn't abstract. It's your contractor delaying your kitchen renovation. It's your restaurant shortening hours. It's your town's new apartment complex sitting half-finished.

---

What Happens Next? Three Scenarios

So where do we go from here? The study ends in October 2025, but the policy continues. Here are three plausible paths:

Scenario A, Enforcement Intensifies

If the administration doubles down, hitting its stated goal of one million deportations per year, the NBER findings suggest the job losses scale proportionally. The EPI projects that deporting 4 million people over four years would eliminate 3.3 million immigrant jobs and 2.6 million U.S.-born jobs. Construction alone would shed 861,000 native-born positions.

The economy doesn't just slow. It contracts in specific sectors, creating regional recessions in immigrant-heavy states like Texas, Florida, California, and New York.

Scenario B, Policy Pivots

Congress could reinstate work permits, revive Temporary Protected Status, or redirect enforcement toward employers rather than workers. The NBER study suggests that even modest reductions in enforcement intensity would restore labor market confidence quickly, the chilling effect reverses when the thunder stops.

Scenario C, Market Adapts (Slowly)

Over five to ten years, businesses might automate some roles, restructure others, or relocate operations abroad. But as the Federal Reserve Bank of San Francisco found, immigration slowdowns have already reduced employment "disproportionately in construction and manufacturing." Adaptation isn't instant. And in the meantime, the 668,000 becomes 1 million, then 2 million.