U.S. Labor Market: Back Against the Wall

Loosening does not mean loose

The economic commentariat stayed true to form on Friday in their reactions to the Employment Situation report for June.  The Wall Street Journal provides a good case-in-point, with headlines declaring that the report “builds the case for a September rate cut”.  This is the same assertion they made in December about March, and in March about July.  And yet, here we sit with inflation pressure rising and hawks on the FOMC bringing rate hikes back on the table.  The commentariat has once again failed to distinguish between a deceleration caused by capacity limits and a deceleration caused by a deterioration of final demand.  This note does make such a distinction and will explore the implications.

Much of the mainstream commentary focused on a “slowdown” in headline employment growth and an increase in the unemployment rate (Chart 1).  There has been a slight slowdown in job growth relative to early 2024 but the change is certainly not dramatic.  Nor is the current rate much different from average employment growth observed since early 2023.  The supposed slowdown falls well within the bounds of statistical error and is therefore a show about nothing. 

To be sure, there has been significant loosening of labor market conditions since 2022 (Charts 2 & 3).  However, loosening does not mean conditions areloose and past trends in economic data do not predict future outcomes.  Trends in economic data provide the starting point in a search for causality, not the end point of statistical analysis.  The question to be asked is not “Will loosening continue as it has been?”, but rather “Will the factors driving the loosening continue to have this effect?”

Many measures of labor market conditions have returned to levels observed in 2019, which has given rise to the narrative that the labor market is poised for downturn.  It certainly could be the case that the labor market is on the edge of a plunge into deflationary territory, but the last place one should look for signs of a future downturn is the labor market.  The labor market is a lagging indicator and its only value in forecasting is in estimating the availability and cost of the scarcest economic resource – human labor.  Studying the labor market shines light on a portion of the stage on which economic activity will play out. 

With that in mind, only a casual assessment of labor conditions is needed to see how primed the economy is for another blast of inflation.  Unlike 2019, the supply of idle labor supply has dropped effectively to zero (Chart 4).  A sudden GDP growth spurt that drives employment growth above labor force growth will run into an inflationary wall almost immediately.  The labor force has, problematically, posted no growth since Q3 2023 (Chart 5).  Indeed, only a surge in prime age participation into boom territory is keeping the size of the labor force from shrinking (Chart 6).  Older workers are taking advantage of their elevated home and stock prices and fleeing the labor market en masse.  Without labor force growth there will soon come a time when any employment growth is inflationary.

Measures of demand for labor show few signs of deterioration.  Layoffs and unemployment claims remain subdued (Charts 7 & 8).  The only indicator flashing a warning sign is the producer price index for employment services (i.e. headhunters).  A downturn in this index is typically associated with recessions.  However, the index remains ten percentage points above its level in December 2019 (Chart 9).  On an absolute scale, the recent pullback in prices does not look like a sign of imminent recession. 

Another sign of a supply shortage is found in employment and participation ratios (Charts 10 & 11).  A falling employment ratio is typically associated with recession.  However, in this case the decline is only occurring with men and is slow and steady, indicative of structural change rather than cyclical shock.  This is confirmed by the contrasting growth in female participation and decline in male participation.  As seen above, the decline in participation is driven by older men.  With wage growth elevated in many industries, it seems unlikely these men are leaving the labor force because of excess slack.  Older men can afford to retire because of the asset bubble, and they have been doing so.

The household survey has been providing many more indications of supply shortage than the establishment survey.  The two sides of the report, logically, reflect the two sides of the labor market, namely supply and demand.  The household survey, representing supply, is reporting that the labor force and employment stagnated in mid-2023 and have not grown since (Chart 12).  In contrast, the establishment survey has reported steady and sustained employment growth, reflecting strong demand for labor (Chart 13).  Indeed, the gap between the two surveys on the question of employment has growth to a level not seen since the historically tight labor market of 1999 (Chart 14). 

Indications of labor market strength can be found at the industry level as well.  Manufacturing employment topped out in late-2022 and growth has been nonexistent since then (Chart 15).  However, far from indicating a lack of demand for new workers, wage growth in good-producing and manufacturing remains elevated (Charts 16 & 17).  Wage growth in service industries has returned to 2019 levels and many industries are showing signs of bottoming out at this level (Chart 18).

Conclusion

The purpose of examining conditions in the labor market is to determine the likely effects on income and inflation of a change in final demand.  The labor market cannot provide a forecast of future conditions, but it can tell you the likely outcome that a change in conditions will produce.  With that in mind, conditions in the labor market unequivocally point towards a lower potential growth rate.  Unless the bureaucrats running the central bank or the politicians running the central government act with Stoic self-restraint the most likely outcome will be inflationary growth.  The employment ratio has been a reliable indicator of long-run real GDP growth, and this measure currently sits significantly below levels observed in 2000 or 2007 (Chart 19). 

By the same token, even with the mediocre growth rates observed in 2023 and 2024, the economy remains at a structural bottom for unemployment.  The stories told by the Beveridge Curve (see Parts One & Two, and follow-up), and the Phillips Curve are one and the same (Charts 20 & 21).  Conditions in the labor market have eased noticeably over the past two years, but nonetheless unemployment remains at its structural bottom.  Instead of a labor market deteriorating into a recession, this writer sees an economy with its back against the inflationary wall.  If real interest rates begin to fall, either because nominal rates come down or inflation begins rising, the economy will rapidly gather pace and the Fed will have an inflationary mess on its hands.