Are There Cracks Developing in the Job Market?

The evidence suggests there are, and that's important for everyone from working families to the Federal Reserve.

Here’s my Occam’s Razor explanation of what’s mostly been driving our strong economic expansion until the Trumpies started breaking things. It started with strong fiscal support, but:

…as fiscal support faded, the strong labor market and falling inflation generated real wage gains, supporting real consumer spending.

If that sounds fanciful to you, take a gander at this scatterplot of annual yearly changes in real consumer spending and real aggregate compensation (meaning average real comp times hours worked for the total workforce):¹

That’s a tight fit, with correlation of 0.79, and “correlation” is the right word, because they both cause each other. Strong real comp leads to strong real spending, and strong real spending creates more labor demand. And because consumer spending is 70% of our nominal GDP, this linkage is critical to powering American economic expansions.

That’s all from a macroeconomic perspective, but at the micro-level, labor market conditions are hugely important to working-age families. No question, people really hate inflation, but they hate job loss even more. After all, for all the heated focus on the stock market on one end of the wealth scale and the safety net on the other end, the vast majority of households depend on the job market for their economic well-being.

That’s why folks like me are getting a bit nervous these days: we’re maybe—I’d say probably—seeing some cracks in the labor market:

—Payroll job gains over the past three months average out to 35,000/month. If that sticks, it’s way too low to prevent the unemployment rate from rising (i.e., it’s below anyone’s “breakeven” job-growth number).

—If you’re thinking that very low pace will get revised up, you of course may be right, but I wouldn’t count on it. I recently explained the logic of why payroll revisions tend to be more negative when the economy is slowing.

—Job gains are very highly concentrated in one sector, and it’s a non-cyclical sector: healthcare. As the figure shows, the sector has been a consistently strong contributor, but it’s also looking like the last sector standing. I find this particularly concerning, because we’ve been holding steady in a low-hire, low-fire mode for months now, and Trumpian trade and deportation policies are likely dampening employer and consumer “animal spirits,” meaning this is risky time for job creation to be firing on one cylinder.

—Black workers’ employment rates have trended down in recent months, and that can be an early-warning of weakening labor demand.

—A new Richmond Fed report notes: “…a deeper dive into the unemployment rate's composition reveals signs of broader economic slowdown possibly lurking beneath the current healthy rate.” Their focus is on the difficulty that new labor-force entrants, along with some of the formerly unemployed, are having finding jobs, which they show through elevated measures of time spent in unemployment. This is key point which I’ll get back to in a moment.

On the other hand, a few key indicators are less consistent with the job market losing steam:

—Nominal wage growth is usually somewhat cyclical and it’s not showing obvious weakness. Below is a plot of 6-month annualized wage growth for middle- and lower-wage workers. The pace has decelerated from its pandemic-driven ups and downs, but it’s running at around pre-pan levels.

—The unemployment rate itself is still pretty low at 4.2%, and has been pretty flat for about a year (see lighter line and right axis) in figure below). That said, both unemployment and the more comprehensive (U6; the other line in the figure) measure of labor-market slack have gradually edged up from their full-employment lows of a few years back.

—UI claims remain quite benign. Yes, the stock of continuing claims remains somewhat elevated, but initial claims look fine. This is all consistent with low-hire, low-fire, and the fact of low layoffs means were not in a recession.

But I still think the weight of the evidence is worrisome re where the job market is headed. The problem comes down to the very simple macro hydraulics of d(u)=b(y-y*), which is a fancy way of saying if real GDP growth (y) is below trend or capacity growth (y*), then unemployment (u) will rise (for every point that y is below y*, u goes up maybe a bit < 0.5 ppt; that’s ‘b’ in the equation). Even more straightforwardly, if there’s not enough labor demand to absorb new entrants (see Richmond Fed link), we won’t necessary see spiking layoffs, but u will drift up. And that’s my modal forecast.

Which brings me to my punchline. The Fed still has numerous key reports on prices and jobs to see before their next meeting in mid-Sept. But if conditions are roughly similar to where they are now, a 25 bps rate cut at the Sept meeting makes sense to me. Circling back to the top of this post, keep that tight correlation in mind: the virtuous cycle of a full-employment labor market supporting solid real consumer spending which in turn supports strong labor demand—wash, rinse, repeat—has been foundational in this expansion.

Yes, the Fed and the rest of us have to watch inflation carefully, and the Trumpies are playing with fire in that regard. But I think I’m seeing cracks in the job market, and if that’s right, the sooner the Fed addresses them, the better.

1. The figure omits the pandemic years of ‘20-21 as movements in those months are large outliers.