From EJ Antoni in TownHall yesterday:

…native-born employment rose by more than two million in 2025, while the number of foreign-born workers with jobs increased by less than 400,000. This was the best December on record for the number of native-born Americans with jobs. It’s especially important that we’re seeing this positive change from where we were trending previously.

During nearly the entire final year of the Biden administration, the annual change in native-born employment was negative, meaning Americans were losing jobs. All the supposed job growth during that time was going to foreign-born workers, a category which even the Biden Labor Department had to admit contained an unknown number of illegal aliens.

It’s no exaggeration to say that President Donald Trump is returning the American labor market to the American people. [emphasis added – MDC]

Notice verbal sleight of hand, going from from “native-born” to “Americans”: The emboldened sentence might be right if each and every foreign born worker was a non-American citizen.

As I noted back in August 2024, this is a sort of “Great Replacement” screed (I won’t validate it by calling it a theory) which has no place in proper economic discourse. But even more importantly, Antoni is ignoring the issues with the CPS employment series — in particular the population controls used — which invalidate his assertion regarding foreign born/native born employment trends during 2022-24 (see Edelberg and Watson, 2024).

In another portion of the article, Antoni demonstrates why we should rejoice he was not appointed BLS Commissioner:

In terms of the number of jobs added in December, the monthly survey of businesses showed an increase of just 50,000, but the survey of households showed the number of people employed jumped by 232,000.

If one looked at the variability of the two series, one would know that making m/m comparisons of this sort is just…silly.

Figure 1: Month-on-Month change in NFP (blue), civilian employment, adjusted to NFP concept, smoothed population controls (tan), in 000’s. Source: BLS, and author’s calculations.

Over this period (which excludes the Covid-19 recession), the standard deviation of changes in civilian employment is double that of nonfarm payroll employment. Hence, a month-on-month comparison is simply … inappropriate. If one wanted to do a similar calculation for 12 month differences (as others have done), well then, that might make more sense.