There are two surveys put out every month: The Payroll or Establishment Survey, which is the one that gets the headlines, and the Household Survey. The Payroll Survey only surveys large and medium-sized businesses and counts the number of jobs. The Household Survey calls on households and only asks if people in the household are employed. A person holding two part-time jobs is counted as two in the Payroll Survey (number of jobs) but only as one in the Household Survey.
The Payroll Survey shows employment growth of nearly 2.7 million jobs between March and November. The Household Survey, over the same period, shows 12,000. Something is seriously wrong! The headline number that is broadcast in the media is the Payroll Survey, so if it is incorrect and the Household Survey is accurate, then the “hot” jobs market, which the Fed uses to justify its rate increases, is really a “cold” jobs market.
Looking at the recent past, the two surveys were in synch until March, when the Payroll Survey took off. Some of the issues could revolve around the small business birth/death assumptions. Since the Payroll Survey doesn’t sample small businesses, the Bureau of Labor Statistics (BLS) adds a number based on a time trend, and they even seasonally adjust this data. Over that March to November period, the Birth/Death model added 1.3 million jobs. This appears strange to any observer of economic trends, which tells us that business is contracting, not expanding. Even if the Birth/Death add-on is eliminated, the discrepancy is still 1.36 million jobs between the two surveys.
Every year, BLS issues revisions to its monthly Payroll Survey. These revisions occur in February for the preceding year. They are rarely discussed in the media because the data being revised is ancient history. However, such revisions are actually available quarterly, and the Federal Reserve Bank of Philadelphia does a quarterly update on a much timelier basis. On December 13, the Philly Fed released its “Early Benchmark Revisions…” for the March to June period (yes, still a lag of nearly six months as it is now late December). According to the Philly Fed, “In the aggregate, 10,500 new jobs were added during the period rather than … the U.S. CES estimated net growth of 1,047,000 jobs for the period.” Note: 10,000 is a lot less than a million!
Since June, economic conditions have weakened significantly.
The latest data we have for the two Surveys is from November. The Household Survey (blue bars in the chart above) shows a net growth of 12,000 over the March to November period (and 2,000 in July to November period). The Payroll Survey says 2.69 million (1.36 million excluding the Birth/Death add-on)! Given the revisions to the Payroll Survey over the March to June period (Philly Fed), it appears that the more reliable job numbers come from the Household Survey.
Worse, full-time job growth has been negative. Any growth that there is has occurred in the part-time jobs area, and we see that multiple job holders are on the rise. These are not healthy signs for the labor market.
Now that its own Regional Bank has thrown cold water on the Fed Chairman’s argument that rates must rise because of the “hot” jobs market, we wonder if that Chairman and the other FOMC members are listening.
Going back in history, we find that this isn’t the first time there has been such a divergence. Large divergences occurred two other times this century, both during election cycles (2012 and 2016).
The latest housing data revolves around housing starts and building permits, both of which have important implications for future GDP. Single-Family starts fell -0.5% M/M in November and are down -16% Y/Y. Single-Family new building permits fell a huge -7.1% M/M and are down nine months in a row. This bodes ill for housing and the economy.
The chart above shows the rapid deterioration of the Home Builders Market Index, a sentiment survey of the home builders themselves. The index stands at 33. 50 is the cut-off between expansion and contraction. The index is now nearly to the level briefly reached in the lockdown month of April 2020 (the index then was 30). As noted from the chart, this index was in the teens during the housing debacle in the Great Recession. It does appear that such depths will be reached again, perhaps as soon as late Q1.
We say this because of the downbeat housing starts and permits data in the Single-Family sector. The chart (below) on the left shows the rapidity of the fall in starts, and because permits are tanking too (right side), future starts will continue at low levels. This is a huge negative sign for the economy in the short and intermediate term.
The Multi-Family sector paints a slightly different picture. As shown in the chart below, starts (left-hand side) rose in November (+4.1% M/M) and are up on a Y/Y basis (+23%) to the third highest level in 35 years.
Such starts are up because of rising rents in the recent past (new data says rents are now falling). This is good news, from an inflation point of view, in that supply will continue to come online for a few more months which will continue to lower rents. However, as seen on the right-hand side of the chart, Multi-Family permits are falling rapidly. Thus, the rise in new starts is destined to be short-lived.
As we have discussed in past blogs, signs of economic weakness have been emerging for months. The labor market seemed to be an exception, but upon further review, it turns out that it was also weakening. It is clearly much cooler than the headline Payroll data has led markets, business leaders, and the Fed to believe. The housing sector is the most interest-sensitive one. Both the starts and permits data are screaming further sluggishness.
Retailers hired fewer associates for the holiday season as they correctly foresaw a falloff in sales relative to last year. And since July, CPI’s annualized growth has been less than 2.5%. Even the equity markets have thrown in the towel as hopes for a soft landing have been dashed. The Recession is here; its signs are everywhere. The new year promises deflation too.
We have one final thought as the new year approaches: Out with the old (TINA – There Is No Alternative) and in with the new (BAAA – Bonds Are An Alternative).
Robert Barone, Ph.D.
(Joshua Barone contributed to this blog)