Immigrant population down; total number of jobs up; how?
We’re informed that immigrants come to the United States to work, not to enjoy our cradle-to-grave welfare state for four generations.
We’re informed by the U.S. Census Bureau that the number of foreign-born residents of the U.S. (a.k.a. “immigrants”) is down by almost 3 million. “Census Bureau Survey Shows Total Foreign-Born Down 2.9 Million” (CIS, yesterday):
since January 2025 … the government’s monthly household survey, officially called the Current Population Survey (CPS), shows an unprecedented 2.9 million decline in the foreign-born population — naturalized citizens, lawful permanent residents, long-term temporary visitors, and illegal immigrants.
Latin American non-citizens who arrived in 1980 or later account for all of the decrease. This population overlaps significantly with illegal immigrants.
Only 44 percent of the decline in the foreign-born are workers; the rest are children, the elderly, and other non-workers. This fact, along with employment gains by the U.S.-born, mitigates the impact on the labor market and may help explain why the survey of businesses does not show a dramatic decline in workers.
The total number of people working in the United States is up (CBS, today):
Here’s a wider picture of steady-then-increased numbers of workers as immigrants departed:
The increase doesn’t seem to have been caused by an increase in the working-age population, which is down slightly over the same time period:
How can this apparent paradox be explained?
One possible explanation is that immigrants displace the native-born in the labor force. When an immigrant with a job leaves, his or her job is sometimes taken over by a native-born American who had been sitting on the couch playing Xbox. If that were true, though, we’d see an increase in the labor force participation rate and/or a decrease in unemployment rate over the same time period, but that hasn’t happened.
Another explanation is that immigrants were working off the books and, therefore, their jobs weren’t counted in most government surveys. If this were true, we’d see a divergence in CPS data, which asks people “did you work for pay?” vs. Current Employment Statistics (CES) data, which asks employers “did you pay someone to work in at least a quasi-legal fashion?” I ran out of brain power and asked ChatGPT to resolve this possibility:
And in the particular January 2025–August 2026 comparison we just calculated, the divergence isn’t enormous once you compare the CPS consistently: CPS total employment is roughly +0.3 million, while CES payroll employment is roughly +0.8 million. So the data could accommodate some loss of informal immigrant employment, but they don’t by themselves demonstrate it.
What’s left as an explanation, given that we can’t abandon the axiom that foreign-born residents of the U.S. are the hardest workers and, certainly, are never on what used to be called “welfare”? We’re living in the Age of Miracles?
Full post, including comments











