Wonking Out: It’s Getting Harder to Be a Pessimist on Inflation
Thursday’s report on consumer prices was really good news. I mean, really, really good news. The overall Consumer Price Index actually fell slightly for the month of December. We can and should downplay that number, because it was driven in part by special factors like a plunge in gasoline prices and because monthly data are noisy. But even when you try to filter out the noise by excluding more volatile prices and looking at averages over several months, you get a picture of rapidly slowing inflation.
So what’s a pessimist to do? Inflation numbers have been getting better for a while, but the Federal Reserve — which is very unwilling to risk letting up on its inflation fight too soon — has been insisting that inflation is still hot if you look at what it considers more fundamental measures, especially the price of core services, excluding housing, purchased by men with facial hair.
OK, I made up that last part about the facial hair. But the rest is true: The Fed has decided to be pessimistic based on a quite narrow measure. And we won’t have a read on the Fed’s currently preferred inflation measure until a somewhat different set of inflation numbers comes out on Jan. 27.
But the attempt to stay pessimistic on inflation is starting to feel a bit desperate. Also, it represents a strange role reversal from the early days of the recent bout of inflation, when some economists, myself included, formed what many people called Team Transitory, arguing that rising inflation was a temporary result of pandemic-related disruptions and would soon recede.
Actually, at this point inflation is looking somewhat transitory, although it went much higher for much longer than I, at least, considered possible. But one of the ways I and others argued for transitoriness aged poorly. For a while we kept excluding particular parts of inflation that looked idiosyncratic, arguing that what remained looked OK. But the range of goods and services experiencing high inflation kept widening, and we were eventually forced to concede that this was an economywide problem.
Now, however, the upper hand is on the other foot, with inflation pessimists probably making the same mistakes inflation optimists were making a year and a half ago.
This is not to say that it never makes sense to try “looking through” overall inflation to get at some underlying number. Traditional “core” inflation, which excludes volatile food and energy prices, has been an extremely useful gauge over the years. Among other things, it helped the Fed stay the course in 2010-11, when a spike in gasoline prices led to unwarranted demands that it stop its efforts to fight unemployment.
Unfortunately, the traditional core measure hasn’t served us well lately, partly because the pandemic produced huge volatility in prices beyond food and energy, such as those of used cars.
And rising demand for housing, probably driven by the rise in working from home, led to a one-time surge in the rental rates that are used by the Bureau of Labor Statistics to estimate shelter costs, which in turn make up about 40 percent of core prices. Unfortunately, the way the bureau does this means that official shelter inflation lags far behind current developments in the rental market, which means that a lot of current core inflation reflects what was going on in the housing market a year ago and doesn’t capture the huge rent slowdown that has happened since.
So right now we all know that it isn’t safe just to look at traditional core inflation, but we also know that turning to “artisanal” inflation measures (a nice phrase from Justin Wolfers) runs the risk of showing only what you want to see. Still, for what it’s worth, here’s a selection of inflation measures for the past three months:
And here’s what they mean:
Headline is just the official C.P.I. we’ve been discussing.
Traditional core excludes food and energy — but as discussed, we know that it’s exaggerating housing inflation.
Supercore excludes food, energy, used cars — and housing.
Superdupercore — my phrase (I think), but estimated by Jason Furman — uses market rents instead of official shelter inflation.
Wagecore is a new phrase of mine, equal to the recent rate of wage increases minus the gap between wage and price inflation that prevailed in 2018-19.
None of these inflation measures look especially bad! Traditional core and wagecore still suggest inflation running hotter than the Fed’s 2 percent target, but not by all that much.
If you want a broader view, Mike Konczal of the Roosevelt Institute has an incredibly cool graphic showing the distribution of price changes across the economy:
Until mid-2022, inflation just kept getting more widespread, and you had to work ever harder to claim that at some fundamental level it wasn’t that bad. Since then, however, everything has been moving in the opposite direction, and so you need to focus on a handful of bad things to stay pessimistic. In other words, back then optimism required more and more extra work, but now it’s looking like it’s the pessimists who are working overtime.
So what’s the true inflation story? I’d argue that pandemic-related disruptions were, in fact, an important part, and that they have been receding. But I’m also willing to concede that the economy got overheated in 2021, thanks to a burst of federal spending and a relatively slow realization by the Fed that there was a problem. And it seems as if inflation is much more responsive to excessive spending when the economy is running hot than previous experience might have suggested. In economics jargon, the Phillips curve is steep in an economy running at capacity.
The big question then became whether the curve would be equally steep on the way down — whether economic cooling would produce a rapid decline in inflation without the need for a big rise in unemployment.
And the answer appears, at least so far, to be yes. Inflation rose further and faster than almost anyone expected, but now it’s coming down with similar speed. As I said, the inflation news is really, really good.