The Phantom Softening of the Labor Market

The Dallas Fed has shared an intriguing update regarding the labor market’s dynamics. The framework focuses on break-even employment and will be officially published on Tuesday. However, its implications are significant, perhaps even more than the authors anticipate.

Economists Anton Cheremkin, Daniel Wilson, and Xiaoqing Zhou have delved deeper into previous research utilizing immigration court microdata obtained via Freedom of Information Act requests. Their analysis, running through December 2025, presents a narrative that contradicts popular beliefs: the labor market is, in fact, softening.

The break-even employment growth rate is nearly zero. According to the Dallas Fed, this means that approximately 250,000 new jobs per month were added in 2023, particularly during the Biden border crisis, which introduced many new workers into the market. However, by July 2025, this number dropped to about 10,000. Averaging from August to December 2025, the figure even dipped to -3,000. Essentially, this indicates that the economy could lose thousands of jobs monthly while the unemployment rate remains unchanged.

The primary factor here is a sharp decline in unauthorized immigration. Officials in Dallas indicated that net illegal immigration averaged -55,000 per month in the latter half of 2025, leading to a cumulative total of -548,000 for the entire year — significantly above the Congressional Budget Office’s projection of -365,000.

The Department of Homeland Security estimates that the number of illegal immigrants who have exited the U.S. is much higher, approaching nearly 3 million. This includes over 700,000 deportations since January 2025 and an estimated 2.2 million voluntary departures. Even the conservative estimates surrounding illegal immigration, legal exits, and visa expirations point to over 1 million cases.

Participation Rates May Be Haunted by Ghosts

This situation has repercussions that extend beyond mere break-even calculations.

For instance, the labor participation rate fell to 62.0 percent in February from 62.5 percent a year prior. This decline is frequently cited by both pessimists—who view the economy as struggling—and optimists—who argue that monetary policies are overly strict—as proof of a weakening labor market: Americans are exiting the workforce, demand is tepid, and the Federal Reserve should consider lowering interest rates.

Yet, the participation rate fundamentally represents the labor force divided by the private non-institutional population of those older than 16. The denominator is based on Census Bureau estimates, which are updated annually. While adjustments in January 2026 account for these demographic changes, they might still lag and fail to fully capture the number and profile of those leaving the workforce, especially among working-age undocumented immigrants who are vital to the labor force.

Here’s where calculations become intriguing. A substantial portion of those departing the workforce are working-age adults. Many of these individuals arrived to find work, and the assumption is that around 75% of them were indeed employed. If a significant number leave but the population count isn’t accurately adjusted to reflect this change, the participation rate could be artificially lowered by those not currently in the country but still accounted for by the BLS. Even with conservative estimates of outflows, it’s feasible that much of the reported decline in participation might be more statistical blip than genuine weakness. Thus, the drop over the past year could be overstated, if not largely an illusion.

As the saying goes, let’s look at the numbers. If 3 million people exit the labor force with a participation rate of 75%, our estimate indicates that while the actual labor force participation is around 61.9 percent, the perceived rate—impacted by outdated figures—is about 61.2 percent. This discrepancy represents a distortion of roughly 0.7 percentage points. Essentially, much of the noted decrease in participation over the last year is a result of statistical ghosts.

Even considering a lower estimate of 1 million total departures—still below both the DHS and Dallas Fed’s figures—the distortion remains at about 0.2 percentage points, which explains roughly half of the observed decrease.

The Actual Unemployment Rate May Be Lower

In reality, the labor market is tighter than the headline statistics suggest. This implies that the participation rates aren’t falling as drastically as they seem and, when adjusted for, may not indicate any real decline at all. With the break-even employment threshold collapsing to near zero, recent pay statistics that would have suggested weakness two years ago now imply stability or even gradual tightening.

The jobs report from February indicated a drop of 92,000 employed individuals, pushing the unemployment rate to 4.4%. This appears concerning only if we cling to the old standard of 150,000-200,000 new jobs per month from the open borders period. It’s important to note; even the CBO previously underestimated this gap by a significant margin. Models of the labor market constructed on these assumptions have been flawed because the situation is bigger and more rapid than anticipated.

The distortion might also elevate the unemployment rate beyond expectations. The U.S. Bureau of Labor Statistics adjusts household survey outcomes based on demographic segments, such as age and race. When these adjustments exceed the number of prime-age workers who exit the workforce, the estimation becomes skewed. The unemployment rate has only increased from 4.2 percent to 4.4 percent over the past year, a shift that may be even less significant than it appears, especially considering the denominator counts many individuals who are no longer living.

While this analysis may appear to present a more favorable view of the economy under the Trump administration and challenge the belief that the labor market has softened, it simultaneously complicates arguments for significant interest rate reductions. If actual labor supply is lower and effective participation rates are higher than indicated by the distorted numbers, then the anticipated easing of disinflation may not materialize. The Fed may end up implementing actions in the labor market that appear less urgent than they actually are.

Cheremkin’s research is temperate in its conclusions, cautioning against overstatement: “increases in salaries that may have historically indicated economic stagnation are now aligned with a balanced labor market.” This reflects what economists might argue. Essentially, those suggesting a fracture in the labor market need to consider perhaps one of the most significant structural shifts in the U.S. workforce in a generation: a reversal of the Biden-era immigration surge.