There’s an old story about Charles Darwin, which may or may not be true but seems appropriate to our current economic moment. According to the tale, two boys glued together pieces of various insects — a centipede’s body, a butterfly’s wings, a beetle’s head and so on — then, as a gag, presented their creation to the great naturalist for identification. “Did it hum when you caught it?” he asked. When they said yes, he declared that it was a humbug.
That’s kind of where we are in the economy right now. I’m not suggesting anyone is faking the data, but the different pieces of information we have don’t seem to line up — they almost seem to come from different countries. Some data suggest a weakening economy, maybe even on the verge of recession. Some suggest an economy still going strong. Some data suggest very tight labor markets; others, not so much.
Let’s talk about the numbers, and how they don’t add up.
The number we usually use to assess where the economy is going is real gross domestic product — and according to the official estimate, real G.D.P. shrank in this year’s first quarter. We won’t have an official (advance) estimate of second-quarter G.D.P. until later this month, but “nowcasts” that try to estimate G.D.P. based on partial information — like the Atlanta Fed’s widely cited GDPNow — suggest slow growth or even an additional period of shrinkage.
In case you’re wondering, no, two quarters of declining G.D.P. won’t mean we’re officially in a recession; that determination is made by an independent committee that takes a wide variety of information into account. And given the confusing picture right now, it’s unlikely to declare a recession, at least yet.
Among other things, another widely used number — job creation — is telling quite a different story. The official estimate of growth in nonfarm employment in June came in quite strong — 372,000 jobs added — which doesn’t look at all like what you’d expect in a recession.
So do we have a conflict between data on output and data on employment? If only it were that simple. We also have alternative measures of both output and employment — and in each case these are telling different stories than the more widely cited numbers.
We usually track economic growth using gross domestic product — the total value of stuff produced. But the government creates a separate estimate of gross domestic income, the money people get from selling stuff, including additions to inventory. The basic accounting says these numbers must be the same. But they’re estimated using different data, so the estimates never agree exactly. And right now the estimates are diverging a lot: G.D.P. shows a shrinking economy, but G.D.I., well, doesn’t:
What about employment? The Bureau of Labor Statistics carries out two surveys, one of employers — which is where the payroll numbers come from — and one of households, which produces an alternate estimate of the number of Americans working. The payroll number is usually considered more reliable — household data are famously noisy — but for technical reasons (the birth-death model; aren’t you sorry you asked?) the payroll data often seem to miss turning points, when employment growth either surges or plunges.
And right now the two surveys are telling different stories. I use quarterly rather than monthly data to smooth out some of the noise; the household data points to a much bigger slowdown than the payroll data:
But wait, there’s one more puzzle. Everyone says we have an extremely tight labor market, and when you combine that with high rates of consumer price inflation, there are widespread fears that we’re on the verge of entering the dreaded wage-price spiral. But wage growth isn’t accelerating. In fact, it’s falling fast, and at this point may not be much above the level consistent with the Federal Reserve’s long-run target of 2 percent inflation:
Are you confused? You should be. I’ve been in this business a long time, and I can’t remember any period when economic numbers were telling such different stories. On the other hand, we’ve never before faced the kind of shocks we’ve gone through in the past few years: Both the pandemic-induced recession and the recovery from that recession were, to use the technical term, weird, and maybe we shouldn’t be surprised the measures we normally use to track the economy aren’t working too well.
My guess about what’s really happening is that the economy is indeed slowing, but probably not into a recession, at least so far. And a moderate slowdown is actually what we want to see.
At the beginning of 2022, the U.S. economy was almost surely overheated, and this overheating was contributing to (although not the only source of) inflation. We wanted to see the economy cool down before inflation got entrenched in expectations, and that’s an area where all the available data — slowing wage growth, inflation expectations in the financial markets, surveys that ask consumers what inflation rate they expect over the next few years — are telling the same story: Inflation is not, in fact, getting entrenched.
Overall, the picture appears consistent with a “soft landing” — a slowdown that falls short of a full-on recession, or involves a mild recession at worst, together with stabilizing inflation.
But, of course, we don’t know that. In fact, given the wide discrepancies in economic data, economic pundits (including me) have unusual freedom to believe whatever they want to believe. Just pick and choose the numbers that tell you what you want to hear and glue them together.