Total nonfarm payroll employment increased by 139,000 in May, and the unemployment rate was unchanged at 4.2 percent, the U.S. Bureau of Labor Statistics reported today. Employment continued to trend up in health care, leisure and hospitality, and social assistance. Federal government continued to lose jobs.
However, the surface is where the good news ends.
Ominously, we are seeing a return of the BLS’ favorite statistical methodology, Lou Costello Labor Math, as the different jobs gauges are not aligning with one another.
At the same time, May did not see any real strengthening in goods-producing wages.
April was a relatively good month for jobs, but the detail underneath show May to be quite a bit less than that, with job growth that is more illusory than real.
Join a growing community that values truth over noise. Become a paid subscriber today!
Lou Costello Labor Math
As readers will recall from last month’s jobs discussion, the hallmark of when the BLS is using the garbage methodology I call Lou Costello Labor Math is when the various job metrics do not trend along similar lines.
While this was the case in April, it is definitely not the case for May. The Employment Level from the Household Survey dropped significantly, and ADP continued to underperform the Establishment Survey All Employees metrics.
Additionally, the BLS unsurprisingly had another set of large corrections to the prior two months data, erasing 95,000 jobs from March and April.
Despite these variances within the data, the All Employees data from the Establishment Survey showed some job growth for May.
As the various job gauges do not align, we should not be surprised to see that the All Employees data by economic sector are printing significant weakness, even as the headline data shows some strength in the overall jobs numbers.
The sector-level weaknesses are another confirmation that the BLS’ methodologies are off, as they do not lend support to the stronger headline numbers.
They are also confirmation that the jobs recession has not eased up yet. The optimism that was guardedly feasible last month has evaporated.
Goods-producing wages did not change at all in real terms.
Real service wages moved up only incrementally for the month.
While we have not seen May’s inflation numbers yet, barring another deflation month wages are still not likely to have fully caught up with price rises following the 2022 hyperinflation cycle.
Wages are certainly close—much closer than they have been—but matching the growth in the CPI since 2021 only means that wages have at least gotten back to where they should be relative to prices. Goods producing wages are close to that point, but service wages still need to rise noticeably to fully recover.
When people do not work a full forty hours per week and wages do not rise, it is mathematically impossible for people’s pay situations to improve. Certainly for the month of May people’s paychecks did not get any better.
Rounding out the grim statistics in the jobs report is the reality that joblessness is still rising in this country. While the unemployment rate technically remained unchanged, the unemployment level did not—it rose for the month, continuing a trend which began in early 2023.
The situation is actually worse than that, because when we incorporate the number of people technically not in the labor force with those who are unemployed, we find that the “real” unemployment rate actually rose for May, and is worse now than in March, before the “good” month of April.
Because those not in the labor force are not factored into the “official” employment calculations, this surge does not show up in the May unemployment rate. It still indicates rising joblessness, however.
A falling employment population ratio means the United States is not producing enough jobs to keep pace with population growth. Even if there was any job growth, which is debatable, it was clearly not enough.
Despite what the headline jobs numbers indicate, the detail refutes the premise that there was any significant job growth in the US for May. Even if the Establishment Survey detail was consistent, the significant variances between the Establishment Survey data, the Household Survey data, and the ADP data make reliance on the Establishment Survey data problematic at best.
That is the consequence of Lou Costello Labor Math—untrustworthy data.
The credibility of the Establishment Survey is further challenged by the job losses shown in Manufacturing and Trade, Transportation, and Utilities. Coupled with weakening job growth in Education and Private Healthcare, a sector ADP indicated had job losses, these losses make job growth at the headline level even less likely.
The Employment Level indicates there was significant job loss for the month, and the falling employment-population ratio is certainly consistent with that.
At best, without speculating on the revisions to job growth that will be recorded for April and May in the June report, there was little to no actual job growth in the US for May. When we consider the magnitude of the revisions that have been made just since the beginning of the year, that little to no actual job growth quickly becomes likely job loss.
It is not possible to establish an improving jobs situation within the US economy based on the BLS May jobs numbers. That paychecks remained weak is further argument the jobs markets are either stagnating or weakening.
What is clear from the May jobs report is that the United States is still mired in the jobs recession it has endured since late 2023. Getting out of that would have required several repeats of last month’s jobs report, and that we did not get for May.
For Donald Trump, the continued loss of manufacturing jobs strikes at the heart of his Agenda 47 pledge to make the US a manufacturing superpower. May’s data moves him farther from fulfilling that promise.
April’s jobs report was guardedly optimistic. May’s report has returned us to the usual posture of resigned pessimism.
The jobs recession continues, and it continues to drag the US economy down.
Thousands of U.S. businesses, large and small, rely on federal jobs numbers when making decisions about company expansions or contractions. It bothers me that they are basing their decisions on highly inaccurate data, as this will affect their levels of business success. This is just so wrong! To what extent are they aware of the deception? Are they all learning the hard way, through bankruptcy court?
It seems to me that a more accurate analysis such as yours, Peter, would be highly valuable to the business world. If somebody else was handling the sales and marketing of this service, and you were left free to just analyze raw data and expose the incompetence of government bureaucrats, you could have a wonderful niche business for yourself. Just sayin’. As in, keep a corner of your brilliant mind open to such opportunities. I would love to see people flocking to your analysis, giving you the remuneration that is your rich reward!
Thousands of U.S. businesses, large and small, rely on federal jobs numbers when making decisions about company expansions or contractions. It bothers me that they are basing their decisions on highly inaccurate data, as this will affect their levels of business success. This is just so wrong! To what extent are they aware of the deception? Are they all learning the hard way, through bankruptcy court?
It seems to me that a more accurate analysis such as yours, Peter, would be highly valuable to the business world. If somebody else was handling the sales and marketing of this service, and you were left free to just analyze raw data and expose the incompetence of government bureaucrats, you could have a wonderful niche business for yourself. Just sayin’. As in, keep a corner of your brilliant mind open to such opportunities. I would love to see people flocking to your analysis, giving you the remuneration that is your rich reward!