Last Friday, August 5, US Jobs Report for July 2022 surprised even mainstream economists who had forecast a 250,000 increase in jobs created in the official US Labor Department monthly jobs report for the period ending mid-July. The numbers came in at 528,000 in the CES, large corporations survey for the month.
The unexpected large increase in jobs was jumped on by Biden administration and business sources alike who had been arguing publicly in preceding weeks that the US economy was not in recession. How could it be a recession, when the jobs market was so robust, the argument went? Never mind the fact that the US economy measured in GDP terms contracted by -1.6% in the first quarter of 2022, followed by another -0.9% contraction in the second quarter for a combined first half, January thru June, decline of -1.3%. That was just a ‘technical’ recession, not a real one the recession deniers argued. The real declaration of a recession remains with that group of politically well connected-professional economists from elite universities who are members of the quasi-official NBER (National Bureau of Economic Research). It is they who decide whether a recession has occurred, and months after it is virtually over. The NBER experts are yet to say it’s a recession. The NBER looks at more than just GDP and that includes employment, and jobs are being created at the rate of 528,000 this past month.
Of course this argument ignores the fact that jobs are notoriously what’s called a ‘lagging indicator’ and job decline on average begins six months or later after a recession commences. It also ignores the further fact that whenever GDP has contracted two consecutive months, as recently, in all the past US recessions it has been followed by the NBER also declaring well after the fact that a recession has occurred.
But there’s a deeper problem in the view that relies on the last jobs report to argue that there’s no recession yet, notwithstanding the first half GDP contraction in the US. That problem is last week’s jobs numbers—showing 528,000 new jobs—may not be accurate. And even if it is assumed that large corporations created 528,000 jobs, as indicated in the CES (Establishment) survey of the report, the second survey in the report shows something quite opposite. That’s the CPS, or Household survey, the second survey that makes up the monthly Jobs Reports.
There’s a bombshell in the CPS survey that neither the Biden administration, the media, and most mainstream economists are conveniently ignoring—or else aren’t capable of understanding.
Here’s the basic contradiction in last week’s Jobs Report for July:
The CES Survey, which is based on about 450,000 larger enterprises reporting to the Labor Department every month, indicated the 528,000 ‘new’ jobs created in its B-1 Table.
But the second survey, the CPS—sometimes called the Household survey—is obtained by the Labor Department doing a phone survey of 50-60,000 households every month and asking them if they’re working, if unemployed, if out of job are if they’re still looking for one, when was the lasts time they actively searched for a job, etc. The household survey determines the unemployment rate. However, like the CES establishment survey, the CPS survey also determines a monthly level of total employment in the economy.
And the CPS survey reveals something quite different, even contradictory, to the 528,000 jobs gained in July in the CES survey lasts month. The CPS’s Table A-8 shows a decline in total non-agricultural jobs from June to July of -112,000. Moreover, the CPS total employment numbers show an even further fall in total employment since May 2022 thru July 2022 of -181,000.
So what’s going on? What’s the correct number of employment gains in July? Is it the CES establishment survey of 528,000 new jobs in July? Or is it the CPS survey, indicating -112,000 fewer net jobs?
Never has the gap between the two surveys that constitute the monthly Labor Department Jobs Reports been larger. But you’d never know it listening to the mainstream media or the politicians.
There’s a saying that ‘the truth is always in the details’ and that’s never truer than when considering government statistics—especially employment stats but wages and inflation as well.
And there’s a ‘bombshell’ group of stats within the CPS survey that strongly suggest the US labor market has hit a wall since May and is actually beginning to soften quickly—a trend that will likely accelerate in coming months.
And if the CPS is more accurate, and the jobs market is actually not robust but is softening, then perhaps recession is actually here now. And there’s another implication: if the Federal Reserve focuses on the 528,000 number and continues to accelerate its interest rate hikes at 75 basis points again next month (and again thereafter), it will only accelerate the recession that has begun and cause it to go deeper than it already has.
What then is the ‘bombshell’?
In Table A-8 in the CPS survey there’s a category that measures the number of part time jobs created the past month. It shows that no fewer than 800,000 part time jobs were filled during July—300,000 involuntarily (i.e. workers forced to work part time when they wanted full time work) and another 500,000 part time jobs created voluntarily (i.e. workers chose to work part time).
If 800,000 part time jobs were created, how does one get ‘only’ 528,000? It might logically mean that 275,000 or so full time workers ‘lost’ their jobs. Or maybe some of the 275,000 lost their jobs outright but some of that number found their full time jobs reduced to part time. If the latter case, then the average number of hours worked should also show a decline in July. But it didn’t. The average hours worked per week remained at 34.6 as it had in June. And in manufacturing, where typically more overtime hours are worked, it remained stable month to month as well, at 40.4 hours/week. In other words, it doesn’t appear likely that the majority of the 275,000 jobs difference (i.e. 800,000 minus 528,000) is attributable to full time workers previously being reduced to part time hours.
Another possible explanation of the discrepancy is that many of the 528,000 new jobs were actually new part time jobs created in July. The Labor Department does not differentiate between a new job that’s full time and one that’s part time. A new job is a new job. It counts both part time and full time as just ‘a job’.
That many of the 528,000 are likely ‘part time new’ is supported by the job gains in July in those industries that notoriously hire only part timers. There were 74,000 net jobs created in bars and restaurant employment, for example. 22,000 in retail sales. Another 22,000 in hotels, accommodations and entertainment—all notorious for hiring only part timers. Even manufacturing in recent decades has hired an increasing number of part time and temp workers. It too added 30,000 last month.
How many of the 528,000 CES survey July jobs were part time is unfortunately not specifically identified by the monthly Jobs Reports. One must infer from data provided in other Tables in the CPS survey, as we’ve just done. It’s not specific, but it suggests the number is probably quite large.
But it could also be that the 528,000 is composed almost entirely of part time job creation. If so, the full time employment must have risen very little in July. But full time employment not only did not rise in July over June. It actually declined. CPS Table A-9, for example, shows 132,577 full time jobs in July down from 132,648 in June—a decline of 71,000. And the decline is even greater from May: 223,000 fewer full time jobs in July compared to last May.
So full time jobs are actually declining in recent months while part time jobs rose in July by 800,000. It’s therefore likely that the 528,000 CES survey rise in jobs in July is overwhelmingly composed of part time jobs.
And here’s a corroborating further statistic for this assumption in the CPS survey. It involves the number of jobs that are 2nd and even 3rd jobs. The category of ‘Multiple Job Holders’ in CPS Table A-9 shows a consistent sharp rise in multiple job holders in recent months and over the past year for that matter. Multiple jobs mean virtually all part time jobs (except perhaps for a full time job held over a weekend in addition to a M-F full time, but those numbers are low). Multiple jobs rose by 92,000 in July over June and by 331,000 since May. And from July 2021 to July 2022, the increase in multiple (2nd, 3rd) jobs was 549,000.
In other words, hundreds of thousands of the job gains in recent months do not represent formerly unemployed workers returning to the workforce and getting ‘new’ jobs. They represent workers already with jobs—and increasingly those with only part time jobs—taking on new, additional part time jobs. Many of the 800,000 part time jobs in July were thus workers taking on 2nd and 3rd jobs, at least 71,000 per Table A-9.
It’s important to understand that the monthly Jobs Reports do NOT represent workers finding work for the first time—either after being unemployed, or re-entering the labor force, or entering it for the first time. The Jobs Reports report Jobs created, not employment per se.
We can summarize these alternative statistics from the Labor Department’s CPS survey that contradict its CES survey’s 528,000 new jobs assumption in July as follows:
- The CES survey picks up raw jobs data from larger enterprises but misses job trends in the small-medium businesses in which trends are more volatile and downturns (and upturns) in job creation often shift and precede trends in larger establishments
- The CPS survey breaks out diverging trends in part time vs. full time work in more detail as well as part time that reflects multiple jobs vs. single held jobs (either full or part time)
- The CPS shows slowing and evening declining job creation for full time work, which means less total wages for millions of workers compared to if they were full time.
- Slowing to declining full time employment (CPS) amidst rising part time and multiple job holding means employers are hiring more cautiously, preferring part timers in case they have to soon lay off workers as recession deepens
- The part time and multiple job trends are a ‘canary in the employment coal mine’, signaling a slowing in hiring overall that will soon spill over to the CES survey. It does not reflect a robust labor market ‘on fire’ in terms of hiring and economic growth.
- Workers are taking on more part time and 2nd jobs because they probably can’t find decent paying full time jobs offered by companies.
- All the media talk about 11 million jobs out there that workers won’t take does not account for the likelihood these are mostly not full time and are insufficient part time paid jobs.
- All the hype about workers’ quit rates being so high is probably representative of workers quitting poorly paid part time jobs and seeking and getting other better paid part time work (or less dropping out of the workforce because they can’t find something better, which is also rising)
This scenario of the US jobs market does not support the view that the US economy is booming due to the large number of jobs created per the CES survey. The Fed, therefore, by accelerating its rate hikes is doing the opposite of what it should.
Other labor statistics corroborate the view that the labor market is hitting a kind of wall this summer 2022. Look at the labor force participation rate statistic, which is also slowly declining in recent months. Or the rising number of people surveyed who indicate ‘Not in the Labor Force’. Or the numbers of the unincorporated self-employed (i.e mostly independent contractor very small businesses) dropping (taking part time jobs or dropping out?).
It is true that the numbers of employed rose significantly from the spring of 2021 to March 2022. That was due to the economy ‘opening up’ after vaccines became widely available after the worst of Covid in spring 2021. About 6 million jobs were added. But this were not jobs that were ‘created’, as the politicians like to say. These were jobs that were ‘restored’ after the Covid shutdown. How many actual net new created jobs out of that total are likely not many.
Now that the ‘restored’ jobs have been maximized, the US economy appears unable now, going forward, to actual create net new jobs—except perhaps to some extent as part time work, which always rises sharply at the end of a business cycle when it enters a recession period. As the recession deepens, part timers are first laid off. Conversely, in early phases of recovery from recession, businesses typically hire more temps as they test the water whether a recovery is truly underway.
The dynamic of the relationship between full time hiring, part time and temp hiring over a business cycle is poorly understood by most economists. However, one thing is clear: the US economy is already in recession in early stages as the data in the CPS survey Tables, A-8 and A-9, shows. These tables reflect the deeper job trends—not the CES Table B-1.
By the politicians and media—and the Fed—focusing on the CES survey’s 528,000 jobs they are about to miss—and in the case of the Fed exacerbate—the recession that’s already begun and, in turn, fall behind the curve. Of course, that’s nothing new for the Fed.
By the late fourth quarter 2022 the ‘bombshell’ embedded within the CPS data will have ‘gone off’. The jobs market will no longer lag and the recession will be blatantly obvious. The Fed will have to abruptly halt accelerating its rate hike policy. And the NBER will have to agree after the fact that the so-called ‘technical recession’ that arrived in the first half of 2022 did indeed signal the US economy had entered recession.