Tomorrow’s Jobs Report Will Finally Capture The Surge In Illegal Aliens, Lead To Another Big Negative Payrolls Revision

Remember last August when the Biden admin finally admitted it had been rigging the jobs number, when as part of its preliminary revision it vaporized 818K jobs in the past year that were never actually added, a historical negative revision (the second biggest on record) which the Fed used as justification for its panicked jumbo 50bps rate cut just a few weeks later?

Well, tomorrow that revision – along with a dramatic increase in the US population estimate by 3.5 million primarily to reflect the surge in illegal immigration – will finally flow through fully into the jobs report, as part of a bigger overhaul of the labor market by the now-Trumpian Bureau of Labor Statistics, and the results could be dramatic.

First, some background.

Each year the US Census Bureau adjusts the weights in the Current Population Survey – the source data for labor-force statistics like the unemployment rate – incorporating updated estimates of the population’s size and composition. In the latest vintage, the Census modified its approach to estimating net international migration – for obvious reasons, the main of which being that for the past 4 years the US effectively had no southern border – which had the effect of substantially boosting its estimate of the US population since 2020.

Bureau of Labor Statistics employment data for January (due for release Feb. 7) will reflect these adjustments, capturing the cumulative undercounting of the population relative to the last vintage of Census estimates. 

That, according to Bloomberg, could lead to a large population adjustment this year which could raise the aggregate unemployment rate by raising the weight of recent immigrants, who tend to have higher unemployment rates than the general population. Overall, Bloomberg expects population adjustments to lift January’s unemployment rate by 5 basis points (which as everyone not named Matt Yglesias knows, is not percent but rather one hundredth of a percent).

Here are the details according to Bloomberg:

• The latest vintage of the Census Bureau’s population estimates raised the level of the civilian noninstitutional population for December 2024 by around 3.6 million relative to the prior vintage a year earlier.

  • The main reason was changes to the methodology for estimating net immigration. Historically, Census has relied on American Community Survey (ACS) data for its estimates of net international migration. But this approach posed two challenges:

    • First, the ACS data is lagged. For each vintage of population estimates, Census had ACS immigration information only for the previous year, so the data failed to capture very recent trends.

    • Second, ACS misses some migrants. This is supported by the divergence in recent years between the ACS immigration estimates and immigration totals from administrative data from other parts of the government.

  • To address this, Census adjusted its methodology to incorporate 75% of the difference between the sum of immigration totals from the administrative data and the ACS data. That resulted in a higher estimate for net immigration, lifting the estimate of the total population from 2020-2024.

• Of that 3.6 million, roughly 3 million are 16 and older – the relevant population for the BLS unemployment statistics.

• Historically, adjustments to the population — even relatively large ones — haven’t had much impact on the unemployment rate.In some cases the adjustments have resulted in significant changes to labor-force participation or employment rates, but usually there have been offsetting effects that dampen any change.

• But for tomorrow’s report there is a risk that the updated immigration estimates could lift the unemployment rate more visibly because the adjustment could increase the survey weight of workers with a higher unemployment rate.

Bottom line: Adjustments to the population controls in January’s household survey to boost the unemployment rate by ~5 bps. Together with Bloomberg’s forecast of a decline in government jobs, the unemployment rate is expected to edge up to 4.16% in January from 4.09% a month earlier.

There’s more: since the increase in the population will also raise the level of the labor force and employment, the updated employment level will imply average monthly job growth of around 150k-170k last year — narrowing the gap between the employment figures from the household survey and the estimates from the BLS’ establishment survey. 

And indeed, Bloomberg cautions that the separate, benchmarking process in the establishment survey is likely to result in significant downward revisions to nonfarm-payrolls data, bringing average monthly job growth last year down to about 150k, from around 180k before revisions. 

As Standard Chartered’s Steve Englander writes (full note available to pro subs), according to the abovementioned BLS estimates as of August 2024, March 2024 NFP is likely to be revised down by 818k after benchmarking to the Quarterly Census of Employment and Wages (QCEW). However, further QCEW revisions since August make a 670k downward NFP revision more likely, so the overall gap between NFP and CPS (as of end-2024) will be reduced to about 2mn from 4.3mn.

The adjustment supports Englander’s – and our – view that the surge in illegal aliens is more likely to be captured in NFP than CPS, as many of these workers were able to obtain employment authorization and work “legally” in the country. The household survey has likely understated annual employment growth by c.0.5mn during the past three years.

One problem the BLS will face is that participation rates and employment-to-population ratios for undocumented immigrants are highly uncertain. The BLS uses administrative data from other sources to bulk up population underestimates from standard Census and American Community Survey sources. However, no survey is likely to capture labor-force participation and employment among undocumented immigrants, so the BLS will probably have to make some assumptions based on legal immigrants and native-born labor-market participants. In addition, according to BLS methodology undocumented immigrants who have not been in contact with US Customs and Border Protection agents are unlikely to show up in either census population counts or BLS household employment. While these fully under-the-radar immigrants are a minority, CBO estimated them at c.800k for 2024.

An additional quirk is that the BLS puts most of its population revisions into each year’s January release (data to be released on 7 February). Its practice has been to introduce an abrupt upward or downward jump in January levels to capture the new population controls. This is arbitrary and does not align with monthly census population estimates. In practice, this means that m/m changes in population, employment, unemployment and labor-force numbers are largely meaningless, as they miss the big January shift. The 12M, 24M and 36M changes embed the population changes so they are relevant to the analysis. Ratios such as the unemployment rate, participation rate and employment to population remain meaningful on a m/m basis.

More to the point, as we first correctly reported and  as the BLS subsequently admitted in August 2024, the March 2024 NFP would be revised down by 818k after benchmarking to the QCEW. However, subsequent QCEW revisions suggest a benchmark NFP reduction of around 670k. The overall gap between NFP and CPS will be reduced to 2.0mn from 4.3mn. 

Of course, that’s not the full story. As we also reported in December, the subsequent QCEW release pointed to en even much weaker Q2-2024 employment growth than signalled by NFP. (see “Biden Lied About Everything: Philly Fed Finds All Jobs “Created” In Q2 Were Fake“). However, the benchmark revision will not correct for the overstatement of payrolls in Q2-2024 until the second quarter of 2026, or more than a year into the future, leaving us with another year of NFP distortions! 

Yes, dear readers, the full extent of the labor market devastation under Biden – and just how fabricated the jobs report truly has been – will continue to be unveiled for years to come, long after Biden himself is gone.

More in the Standard Chartered report available to pro subscribers.