The phantom menace

The macroeconomic saga continues

We’re still in macro mode here at The Bellows.

In the month of my birth, the consumer price index stood at a level about 9% above that a year prior. At no point in my adult life has inflation come anywhere close to that mark—not even in mid-2008, at the peak of an oil-price spike every bit as dramatic as that in the late 1970s. And yet, for my entire adult life economists and pundits have warned of the imminent return of high inflation, 1970s-style. The latest example of inflation panic seems on its face a more reasonable iteration than others in the recent past. There is a lot of fiscal stimulus in the pipeline: enough to push American output above what the Congressional Budget Office deems to be the economy’s “potential”. Barring some sort of unforeseen demand-shattering development, nominal GDP growth in 2021 should be pretty darn strong, and inflation is likely to come in high as well.

But people don’t talk about the horrors of inflation in the 1970s because headline CPI jumped up for a year or two and then subsided. Perhaps people forget what exactly the great inflation looked like. Growth in prices accelerated, with various ups and downs but on the whole quite consistently, over the course of two full decades. Now admittedly, sorting out what’s going on in the macroeconomy is hard, and nothing can be taken for granted. Larry Summers—among the most prominent of the current crop of inflation worriers—is right when he says that it would be stupid not to at least consider the risks associated with a jump in inflation. But if we’re going to be warning about these risks, in a way that might potentially deter policymakers from being as generous with economic aid as they’re inclined to be, then we need to really look carefully at the stories we’re telling about what inflation is likely to do and why.

Here’s one story, a particularly scary one, from Summers:

Can and will the Fed control the situation effectively if inflation starts to rise? History is not encouraging. Every past significant inflation acceleration has been quickly followed by recession. Tamping down inflation will require allowing unemployment to rise, and engineering a soft landing is difficult: Unemployment has never risen by half a percentage point without then rising by almost two points, or more. Perhaps policymakers are better equipped to avert a downturn today than they have been in the past, but this kind of reassurance was confidently provided before the 2008 crash. The flatness of the Phillips curve is a double-edged sword: It implies that if inflation does arise, policymakers will have to accept substantial increases in unemployment before it starts to fall. And given the Fed’s guidance about keeping rates low, the economy’s high degree of leverage, and pressures on the dollar, I worry that containing an inflationary outbreak without triggering a recession may be even more difficult now than in the past.

This sounds like the kind of thing someone grappling with hard truths might say. But it looks to me like scaremongering with a reckless disregard for underlying causal mechanisms. Every past significant inflation has been followed by recession. But every past significant inflation has also been associated with either a spike in oil prices or a cycle of monetary tightening or, typically, both. Oil shocks are generally quite contractionary; as James Hamilton has noted, it is hard to find a recent recession in which oil prices didn’t play an enabling role (the pandemic downturn is a notable exception). 

Now, oil prices have been rising. But this increase has essentially unwound the dramatic decline induced by the onset of the pandemic. Year-on-year inflation figures will rise in coming months as increases in energy prices are compared against the depressed baseline of the months after the beginning of the pandemic rather than the more normal levels that prevailed before covid struck. Perhaps Summers anticipates that another sharp jag upward in oil prices is looming, sufficient to induce an economic contraction. But as far as I’m aware that is not the point he’s making in his arguments about inflation, and neither is it a necessary consequence of a stimulus-powered surge in American growth. 

If Summers is not first and foremost worried that an oil shock is about to generate both soaring inflation and a recession, then he must instead be worried that the Fed will induce a downturn as it creates excess unemployment in order to ease price pressures. This, too, would amount to lots of hand-waving with very little attention paid to which mechanics, if any, are generating sustained high inflation.

Energy prices aside, the straightforward story one might tell about rising inflation this year goes something like this. Americans are going to have money burning holes in their pockets. But as they try to spend they will find themselves running up against supply constraints, and so prices will begin to rise. Some of these constraints reflect short-run supply bottlenecks: people want to buy a new car, but there aren’t enough new cars to go around because manufacturers couldn’t obtain enough chips, because the pandemic messed with the chip market, and so for a little while there is too much money chasing too few cars. Other constraints reflect the fact that not everyone can do all the stuff they really want to do at the same time. We’re all dying for a vacation and we all finally have the money to pay for a nice vacation so we all start trying to book vacations. But the supply of hotel rooms and theme parks and everything else hasn’t suddenly jumped, and so the price of all these leisure activities rises.

These dynamics could absolutely push up measures of inflation. But what happens next? Well, short-term supply issues begin to be resolved, for one thing. And Americans work their way through their stimulus checks and accumulated savings (maybe). Then it’s the middle of 2022, chips and cars are no longer in short supply, and American spending is once again constrained by American incomes. Are American incomes accelerating as part of a 1970s-style wage-price spiral? Let’s see. Back at the beginning of 2020 the unemployment rate in America was 3.5%, which was the lowest it had been in half a century. And nominal wage growth was very much not accelerating. Wage growth rose to a respectable but not stellar rate of about 3.5% in 2018, then plateaued. Why didn’t it keep rising? It could be because workers lack bargaining power; here’s a great paper on that subject if you’re looking to read more about it. 

It could also be because there was still plenty of slack in the labor market. The labor-force participation rate had been rising in the run-up to the pandemic but at 63.4% remained three percentage points below the peak reached prior to the global financial crisis, of 66.4%. (The rate currently stands at 61.4%.) Say that full employment means we’re at a labor-force participation rate of 66%. If we are to achieve that level of labor-force participation at an unemployment rate of 3.5%, we’ll need to raise employment from the current level by about 16m jobs. In other words, we need to replace every job lost over the past year and add another 5m jobs or so on top of that. For context, during the best year for jobs growth of the Obama era employment rose by 3m. We’re talking an extraordinary, near-miraculous level of job growth. 

If that were to occur, what we ought to do first is celebrate. Like really celebrate, because the economy hasn’t been in the neighborhood of full employment for more than 20 years. Then after that we could be on the lookout for signs of a sustained acceleration in wage growth, because, again, we cannot take such an acceleration for granted given the big structural changes in the economy since the 1970s and the associated decline in worker power. And if we don’t experience that sort of historical labor-market recovery? In that case, it’s really hard to understand why the Fed would feel the need to raise rates in order to head off inflationary pressures or why, if it did, rates would need to go particularly high particularly fast in order to achieve a level of slack sufficient to slow growth in wages. 

For one thing, Jerome Powell has been clear about the fact that he’s not going to rush the process of rate increases, and the Fed’s new policy framework provides it with more flexibility to overshoot than it’s had in the past. But also, even under the old framework, the Fed did a pretty good job looking through periods of inflation that were pretty obviously short-term in nature. Headline inflation spiked above 5% in 2008 and to near 4% in 2011, and while the Fed probably acted more cautiously than it would have had inflation stayed closer to target it certainly didn’t raise rates in order to deal with those price pressures, for the very good reason that underlying labor-market conditions were not strong enough to turn a brief period of rapid price increases into something more sustained. And as noted above, in the absence of a truly epic and on-the-whole welcome employment recovery this year, that will also be true of whatever inflation we see this year and next.

This is the issue: to tell a really scary story about inflation, you have to tell a story about the Fed unnecessarily overreacting to what will pretty clearly be transitory inflation pressures. Either that, or you need to tell some other kind of story of the sort Summers and others very much aren’t telling: about how Congress will keep passing trillion-dollar stimulus bills even after the labor market recovers, or how American workers will suddenly have corporations over a barrel, or how emergency packages passed in response to a unique challenge will nonetheless persuade Americans that there has been a change in macroeconomic regime thus supercharging inflation expectations to such a degree that prices accelerate upward even in the absence of a truly tight labor market. And ideally, you wouldn’t just tell these stories, but would also do some work explaining why we ought to believe them.

I get that we’ve all had the line that inflation is always and everywhere a monetary phenomenon drilled into our heads. And that dictum seems to have persuaded many people that if rising demand leads to higher inflation and the central bank does not quickly and hawkishly react, then you end up back in the 1970s in short order. But immaculate inflation is not a thing. To argue that dangerous inflation looms ahead, you need some plausible mechanism through which Americans suddenly become convinced that their nominal incomes are going to accelerate ever upwards. It does not seem to me that Americans are about to be convinced of this. Given the way the economy has operated over the past 30 years or so, I find it exceedingly hard to believe that anyone could be confused into believing such a thing by a year or two of durable-goods scarcity and expensive holidays. As far as I can see, the biggest danger we face, inflation-wise, is that the Fed will make a hawkish error because it mistakenly sees temporary inflation as something more worrying. And I don’t really understand how warnings like those aired by Summers help us avoid that outcome.