We aren’t naturally cynical about economic data, but there are things that don’t add up about the job market.

On the surface, Friday’s report was solid, with nonfarm payrolls up more than 200,000 in June, another good month.  However, downward revisions to the prior two months reduced the net gain in total payrolls to just 95,000, with a net gain of only 50,000 for the private sector.

Now for the strange parts.  Nonfarm payrolls are up 2.6 million versus a year ago.  But civilian employment, an alternative measure of jobs that includes small business start-ups is up a grand total of only 195,000 in the past year!  Not 195,000 per month, but a total of just 195,000 over the past twelve months.  Weird, right?

It's entirely possible that one of the major reasons for this gap is the recent surge in immigration. Immigrants who get jobs at one of the companies included in the payroll survey should be counted because it is filled out by employers.  But the civilian employment figures (the weak one) are based on a survey of individual households and it’s hard to survey households in the US that are brand new or that are skittish about filling out a survey sent by the government, particularly if they are here illegally.

It's also important to point out that a gap between the two surveys this large may be highly unusual, but it has happened before.  As a share of the labor force, the gap was briefly larger in the mid-1980s, the late 1990s, in 2013, and during COVID.

Another oddity is the consistent negative revisions for the past few years.  Back in 2022, the third report for payrolls for a particular month averaged 6,000 less than the initial report for that month.  For 2023, the revisions averaged -30,000.  So far this year they’ve averaged -49,000.  In the past few decades, negative revisions are more likely to happen around recessions than when growth is strong.  So maybe it’s a symptom of weakness to come.

But there’s also a more benign explanation.  Remember, the payroll report is based on a survey of employers.  In the ten years prior to COVID, the government was getting an on-time response rate of 75% from the employers it surveys; but in the past three years the timely response rate has averaged only about 65%.

So maybe the companies that don’t fill out the survey on time for the first payroll report are having more business trouble than their peers (compared to normal).  A struggling company would have more important issues to deal with than filling out a government survey.  Eventually, the statisticians will get used to that pattern and make adjustments, but the data are looking funny in the meantime.

Another oddity is the gap between full-time and part-time jobs.  The civilian employment report shows full-time jobs down 1.6 million in the past year while part-time is up 1.8 million.  That kind of loss of full-time positions is normally linked to a recession and declining payrolls, not continued strong economic growth.

Do these anomalies show the government is cooking the books?  We wouldn’t go that far.  If the Labor Department is cooking the books, presumably for political reasons, then why are they letting the unemployment claims reports show an increase and why don’t they cook the civilian employment report to show more job gains closer to what the payroll report shows?

The problem is that it’s hard to argue at this point that government officials haven’t abused their authority to advance a narrative they’ve found useful, including the “slam dunk” case for Iraq having Weapons of Mass Destruction, or COVIDs “six-foot” rule, school closings, and masks, or even whether the measure of deaths from COVID were “from” COVID or “with” COVID.   It’s not just the CDC and NIH that have lost luster in the eyes of average investors, but other government agencies as well.  More people are skeptical of government reports than we have seen in our careers.

Putting it altogether, we think the job market is poised somewhere between the still strong picture painted by the payroll report and the soft reports on civilian employment.  No recession yet, but some early signs of a slowdown.

Brian S. Wesbury, Chief Economist
Robert Stein, Deputy Chief Economist
Date: 7/8/2024