The Labor Force Participation Rate Is Probably Wrong
The labor force participation rate has been sliding for over year, falling from 62.6 percent in January 2025 to 61.9 percent in March 2026, according to the official figures from the Bureau of Labor Statistics. That looks like a labor market in retreat, Americans pulling back from work, perhaps discouraged by uncertainty or softening demand.
But there is a good reason to think the decline is substantially an illusion produced by the way the statistic is constructed, and that the real story is not about workers at all. It is about how immense shifts in immigration and border policies threw off the population estimate in the denominator.
The participation rate is a simple fraction: the labor force divided by the civilian noninstitutional population. The numerator—how many people are working or actively looking for work—comes from the monthly household survey, which contacts about 60,000 households. The denominator, however, does not come from the survey. It comes from Census Bureau population estimates that are updated annually. So, the BLS’s 2024 participation rates are based on the Census Bureau’s 2023 estimate. The 2026 BLS numbers, starting in February, are based on the Census estimate released at the end of 2025.
And those annual estimates have been struggling to keep up with what may be the fastest swing in migration policy the country has experienced in modern history. What’s more, while the Census Bureau does revise the estimates for past years, the BLS figures are frozen in time and never updated to reflect the new population estimates. The time series for the participation rate in 2025 will always reflect the Census estimate released in December of 2024.
Just How Big Is the U.S. Population?
In other words, once a participation rate is published, it tends to stay published with whatever population estimate was in place at the time, even if that estimate later turns out to have been significantly wrong. In a period of stable population growth, this creates only modest noise. In a period when net migration swings from nearly three million people a year to something close to zero, it can create readings that are badly misleading in both directions.
Let’s start with the Biden years. In the original 2023 estimate for the July 2022 through June 2023 period, the Census Bureau said net international migration contributed 1,135,220 to U.S. population growth. In the estimate released at the end of 2024, the Census revised that same period up to 2,289,938. Similarly, the population estimate for July 1, 2023, that had been projected originally turned out to be about 1.8 million too low.
For 2024, the gap was even larger. The original projection had the total U.S. population on July 1, 2024, at 336,673,595. The follow-up, released in January 2025, reported the actual figure at 340,110,988, meaning the projection was an undershoot of about 3.4 million people.
As a result, the denominator used in the household survey had been too small relative to the true population. But the 2025 update also raised the measured labor force sharply, so the headline participation rate itself did not change much. The lesson is not that the entire rise in participation was fake. It is that the labor market picture during the immigration surge was more uncertain than it appeared, because the underlying population estimate had not yet caught up with reality.
Many Americans did not like the Biden border policies, which helped Trump return to the White House. Not surprisingly, the Trump administration very quickly moved to reduce the number of unauthorized arrivals, deport people who had taken jobs and residency in the U.S. without authorization, and convince many more to voluntarily return to their homelands.
What was a surprise was the scale. The Census Bureau’s estimate has net international migration falling to 1.3 million in 2025 and projects it will fall further to roughly 321,000 in 2026. But this is likely an underestimation of the impact of the Trump administration’s immigration policies. The Dallas Fed separately estimates that net unauthorized immigration was negative 548,000 in 2025—meaning over a half a million more unauthorized immigrants left the country than arrived. If the denominator has not fully absorbed that reversal, the population estimate may now be biased in the opposite direction: too high relative to the people actually living in the country. A denominator that is too high depresses the participation rate, making the labor market look weaker than it is.
In other words, the potential for error we saw introduced to the participation rate by Biden’s open borders policies—making the population denominator too small—has likely been turned upside down by Trump’s border security regime. And the policy reversal has been so strong that the participation rate has become unreliable. What we’re seeing as a decline is likely a statistical phantom.
A recent Federal Reserve Board note makes the problem concrete. The 2026 population projection currently embedded in the estimates assumes that immigration still adds about 320,000 to population growth. But the Fed’s own updated analysis suggests the actual change will be very different, ranging from roughly negative 925,000 to positive 185,000. When the assumption built into your denominator is 320,000 and the plausible range runs from nearly a million below zero to barely positive, you have what amounts to a textbook warning that the real-time number may be wrong.
Have You Ever Seen a Phantom Marching?
Walk through the recent readings with the denominator problem in mind. In January 2025, participation stood at 62.6 percent. By March 2025, it had ticked down to 62.5 percent. By December 2025, it was 62.4 percent. Then came the annual population update. The January 2026 reading was initially reported at 62.5 percent, but after BLS incorporated the updated Census population estimate, it was revised down to 62.1 percent, a four-tenths drop that had nothing to do with anyone losing a job or leaving the labor force. February 2026 came in at 62.0 percent, and the March reading fell further to 61.9 percent.
The headline trajectory looks like a labor market losing momentum. But a substantial share of the decline—particularly the sharp step-down in the January revision—is almost certainly the denominator catching up with population changes that had been accumulating for months or years.
The Unemployment Rate Is a Better Guide
There is a useful contrast here. The unemployment rate is likely to be a more reliable indicator of labor-market conditions than the participation rate in this environment, precisely because of the measurement structure. Unemployment is calculated within the labor force itself—the number of unemployed people divided by the total labor force. Both the numerator and the denominator come from the same household survey, so population-estimate errors affect both sides of the fraction in roughly the same way. The participation rate, by contrast, is directly exposed to denominator errors because it divides a survey-based measure of the labor force by an independently estimated population figure. When the population estimate is wrong, the participation rate moves even if nothing in the labor market has changed.
We may have spent a year understating the participation rate as net migration collapsed. The labor market did not necessarily change as much as the denominator did. Anyone reading the participation rate as a straightforward signal of labor-market health right now should understand that they may be reading the migration-estimation error as much as they are reading the economy.
