The economy of 2021 had its good points (high job growth) and its bad points (high inflation). I argued in yesterday's column that we couldn't have had the good without the bad, that trying to keep inflation down would have meant sacrificing much of the good employment performance. And how we evaluate last year will depend a lot on what happens next. It will look like a good bargain if, but only if, inflation fades away.
So when will we know enough to make that judgment? Ah, that's the problem.
Economic developments since the pandemic began have taken place in Covid time -- that is, they've moved at a pace that makes past ups and downs look as if they were filmed in slow motion. For example, we used to think that monthly data gave us a quick read on the state of the economy; these days, when you see numbers for, say, December, you now have to ask, "Exactly when in December?" because that can matter a lot for your interpretation. Employment numbers, for example, are for "the pay period including the 12th, which may or may not correspond directly to the calendar week."
And the speed of events creates problems for some of the ways we normally look at economic data -- problems that will create a lot of fog around the inflation picture in the months ahead.
First, when people talk about the rate of inflation, they often mean the percentage rise in prices over the past year. There are good reasons for that convention: By looking at annual rates of change, we both smooth out meaningless wiggles and bypass the problem of seasonal variation (official numbers are "seasonally adjusted" in smart ways, but there are always debates among economists about whether the seasonal adjustments are getting it right).
Right now, however, the one-year rate of change is more or less guaranteed to show continuing high inflation for a while even if actual price pressures are quickly fading away. Look, for example, at the price of gasoline:
Prices at the pump rose steeply for much of 2021 but have recently leveled off and even come down a bit. That leveling off, however, isn't reflected in the one-year rate of change, which is still extremely high and will remain so for months to come:
A second problem involves prices of things people tend to pay for under long-term contracts -- which for consumers especially means housing. The Bureau of Labor Statistics, which produces the Consumer Price Index, measures housing prices using rents -- actual rents on apartments and an estimate of what homeowners would be paying if they were renting their residences. These measures have risen only moderately so far:
But if you look at the rental numbers from companies like Zumper that match renters with landlords, they show much bigger increases:
This doesn't mean that the B.L.S. is getting it wrong: It's measuring what people are currently paying on average, while companies like Zumper are reporting rates on new rentals. And since many housing units are on long-term leases, you expect average rents to lag behind a surge in new-rental prices.
But what this tells us is that there's a lot of measured housing inflation still in the pipeline, inflation that will show up in the official numbers even if new-rental prices level off. Which they seem to be doing:
The implication of these and other measurement issues is that reported inflation is more or less fated to be high for much of this year, no matter what. In particular, even if true inflation pressures recede, which is what many of us expect and hope, that probably won't be obvious in news reports.
Looking at the inflation measurement issue, I was reminded of a line from my former teacher Charles Kindleberger about the balance of payments (a subject nobody discusses anymore, but that's a story for another day): "The existence of a variety of balance-of-payments definitions makes it possible for an observer always to be grave, or optimistic, according to his temperament." Well, for the next few months inflation will be like that -- and I, at least, will try to keep reminding myself not to emphasize only the numbers I like.