Negative US payroll revisions as bad as 2009 add to fears of consumer spending slowdown

We have better ways to know about the present than the past.

That’s the argument for why I have typically refrained from having my world upended by the initial annual benchmark revisions to US nonfarm payrolls data, which just showed that there were 911,000 fewer jobs than previously thought. Economists expected a revision of -700,000.

“On a raw basis, -911K is worse than any figure, preliminary or final, seen since at least 2000,” wrote Omair Sharif, president of Inflation Insights. “On a percentage basis, the revision was -0.6%, in line with the preliminary benchmark revision we saw for 2009, not exactly a great comp.” 

But in a sociopathic macroeconomic sense, we care about jobs because jobs are the major source of income that enables spending.

Job growth has unambiguously slowed, and now, by much more than we thought. Meanwhile, higher-frequency measures of nominal spending have been picking up steam.

The Johnson Redbook Index of weekly same-store sales for US general merchandise retailers is up 6.6% year on year as of September 6, from a post-Liberation Day low of 4.5% year on year in June.

The major question mark around the US economy right now involves reconciling these divergent trends between jobs and spending: what’s signal, and what’s noise? What’s leading and what’s lagging? How will this seeming wedge resolve? Or do income trends mean there’s really not much of a discrepancy at all?

The market’s view on this seems clear: the SPDR S&P Retail ETF, while getting whacked today, posted a record closing high on Monday. That suggests that investors are pleasantly surprised by how well retailers, as a collective, have managed to mitigate negative effects from tariffs and how top-line trends are holding up through the beginning of this shock.

Of course, with tariffs raising prices for imported consumer goods, distinguishing between changes in “nominal” (prices paid) and “real” (volumes sold) spending is key. If Americans were buying less stuff at higher prices, that wouldn’t be sending a good signal for future production.

That isn’t quite what’s happening yet, though tariff-induced price hikes aren’t fully in the rearview mirror.

Less timely measures of real consumer spending, current as of July, are up about 2.1% year on year. That’s down from 2.9% from a year ago, and below the 2012 through February 2020 average of 2.4% that was deemed the “new normal” for marking a period of slower growth following the global financial crisis of 2008. I’d call this a yellow light when it comes to the outlook for consumer spending. 

While yellow lights are not green, they also *checks notes* aren’t red. And, again, higher-frequency data would point to some improvement here from July to August.

Last year, I was able to write, “If 818,000 jobs ‘vanish’ and all the spending one would associate with solid labor market conditions is still there, do they really make a macroeconomic sound?”

This time, it’s more like, “If 911,000 jobs ‘vanish’ and the spending trends one would associate with softening but not alarming labor market conditions are in place, should we be getting a little more concerned?”

And the answer to that is, “Probably, yes.”