Puzzle #1 in economics right now is why inflation, or price growth, has been so slow in recent years. Puzzle #2 is what Puzzle #1 implies for the Federal Reserve’s monetary policy.
Talk to economists about unusually slow inflation, and you’ll get nods and chin strokes. Say it to normal people and, depending on the latest wiggles in the prices of gas or food, they’ll throw something at you. In fact, because food and energy prices are so volatile, the economist’s preferred inflation metric, plotted below, leaves them out. That’s not trickery or sleight of hand, and when we’re thinking about how far your paycheck goes — its real buying power, say, now vs. a year ago — we of course include the price of the full consumption basket, including food and gas. But to get a clearer signal of underlying price momentum, it’s better to look at the so-called core index.
I’ll get to the monster in the graphic in a moment, but if you look toward the end of the figure, you might wonder what it is we in the econ community are scratching our heads about. Core inflation seems to have pretty much just trucked along at around 2 percent for 20-plus years. Where’s the puzzle in that?
It’s that we expect inflation to be more “procyclical” (speeding up in expansions, slowing down in recessions). The story is that tighter job markets generate faster price growth, as employers, facing higher labor costs, bump up prices to maintain profit margins. For reasons that I suspect many readers intuit, however, that story has broken down in recent decades, as the figure suggests. Toward the end of the graph, you do see core inflation falling a bit in the last recession, then recovering, but it hasn’t accelerated at all as the unemployment rate has fallen from 10 percent to the mid-4’s. And lately, at the very end of the figure, it has turned down.
One explanation is that there’s more slack in the job market than that low unemployment rate suggests. I mean, given that price pressures are one measure of slack, this is axiomatic (wage growth, too, is tamer than you might expect). I’m certain this factor is in play (it’s also germane to Puzzle #2), but it doesn’t solve the puzzle. Clearly the job market is tightening, and yet inflation is most recently decelerating.
Perhaps Internet technology solves the puzzle. The fact that one can effortlessly comparison shop must be putting downward pressure on prices, right? Actually, not so much. [Old guy lament: You know what bums me out? Many years ago, when I was a kid, package deliveries were exciting; they invariably contained something fun. Now they’re just cat food and socks.]
First, allow me to blow your mind with a number. Close your eyes and guess what share of retail sales are accounted for by e-commerce. Whoops — hard to read with eyes closed. Anyway, it’s … wait for it … 9 percent! It’s growing fast, of course, but Walmart still accounts for about four times the U.S. retail sales as Amazon (about 8 vs. 2 percent).
As Goldman Sachs economist David Mericle, someone who thinks deeply about these matters, recently put it, the right question isn’t “Would prices be higher if Internet shopping didn’t exist?” but rather, “Is the rise of Internet shopping more disinflationary than the rise of big-box retail was?” And the answer appears to be: maybe a little, but not much.
Certainly globalization and the increase in supply chains is responsible for the low and steady path of prices over the past few decades, but it doesn’t solve the puzzle. For one, it has been going on for decades and, as Mericle shows, the share of low-priced imports in the gross domestic product has stabilized in recent years. Also, most recently it’s price growth in services, not goods, that’s holding down inflation, and the United States has a trade surplus in services.
Bottom line, Puzzle #1 remains a puzzle. Many economists, including Fed Chair Janet L. Yellen, who knows a little something about this stuff, think that eventually the puzzle will solve itself, and inflation will revert to the Fed’s target of 2 percent. That’s certainly possible and broadly consistent with the figure above. But it begs the question posed by Puzzle #2.
Although the Fed’s analysis takes into account all the above and more, the core of its model is the negative correlation between economic slack and inflation. So the breakdown of that correlation poses a big challenge for it. Lately, the Fed has been operating on the assumption that the correlation will reassert itself, and thus it is slowly raising interest rates based on low unemployment, as opposed to standing pat based on decelerating price growth.
I think that’s a mistake, and if I’m right that there’s more slack in the economy than the Fed thinks, it’s a potentially costly one, especially to those who depend on chock-full employment to get ahead.
Instead, I think the Fed should pause or, if it must, raise rates even more slowly than its current plans do until inflationary pressures are very clear, as in tracking above 2 percent and rising.
The cautious central banker would respond: “That’s too late! By then, people’s expectations about inflation will no longer be anchored to our 2 percent target, and that runaway inflationary monster from the ’70s would once again be upon the land!” (ergo, the monster in the picture).
My response: That ’70s monster is long gone. Workers no longer have the bargaining clout to build higher inflation expectations into their contracts. There are no more contracts!
To be clear, neither I nor anyone else knows that “too late!” is wrong. So the Fed should be using its prodigious analytic firepower to try to get a handle on the extent to which rate hikes and balance-sheet runoffs would reel in price growth when and if it showed its face.
I suspect these anti-inflationary tools in tandem with the many forces holding down price growth will be ample to control accelerating inflation, should it appear. If so, it doesn’t matter whether the monster is dead or just sleeping. What matters is squeezing out the slack and ensuring that the benefits of full employment reach those who need them the most.