Opinion | Will The U.S. Economy Pull Off a ‘Soft Landing’?

At first, Covid-19 dealt a hammer blow to the U.S. economy. America lost 22 million jobs between February and April of 2020. And many analysts worried that the pandemic might leave “lasting scars” in the form of reduced employment, lower output and more.

At this point, however, both total employment and the share of adults in the labor force are right in line with projections made before the pandemic struck.

Same with the labor force …

… and G.D.P.

In other words, most economic indicators show no scarring at all.

But inflation, which had been quiescent for decades, surged in 2021-22 to levels not seen since the 1980s. It has come down from its peak, but it’s still higher than we’ve come to expect. And trying to get inflation back down has become a priority for policymakers.

The question is, how hard will disinflation be? Will it create anything like the pain many Americans endured in the early 1980s because of the Fed’s brutal decisions on interest rates? Or as people often put it, can we achieve a “soft landing”?

In this article I’ll explain why some economists believe, based on historical experience, that we won’t be able to get inflation down without throwing millions out of work and why others don’t believe that this history is a good guide, arguing that relatively painless disinflation is possible. I’ll also explain why even if smooth disinflation without a major recession is possible, there are major risks that policy will either overshoot or undershoot, so we either get an unnecessary recession or fail to get inflation under control any time soon.

I’m in the camp that believes that bringing inflation down doesn’t have to be very costly, although you shouldn’t trust any economist who expresses great confidence on this issue. But I’m very worried about the problem of sticking the landing in the face of huge uncertainty about the current state of the economy, possible future shocks like debt default or more Covid dislocations, and the often delayed effects of policies designed to fight inflation. For example, are interest rate hikes precipitating a bank crisis?

My goal, however, is not so much to persuade you of the correctness of my own views as to give you a sense of the factors in play and the state of the debate.

As a starting point, let’s ask what we mean by a soft landing.

The soft, the hard and the gray

There isn’t any standard definition of an economic soft landing. But I think most economists would call it a soft landing if we get inflation down to an acceptable rate without a large rise in unemployment.

But what’s an acceptable inflation rate? What’s a large rise in unemployment?

Consider the 1988 presidential election. George H.W. Bush won in a landslide largely because voters had a very favorable view of the late-Reagan economy. Yet in November 1988 the unemployment rate was almost two percentage points higher than it is now, while the rate of inflation was similar to its rate in recent months.

So why can’t we just declare victory? One answer is that during the 1990s the Federal Reserve and its counterparts in other wealthy nations coalesced around the idea that 2 percent, not the 4 percent of the late Reagan years, was the appropriate inflation target. The analytical and empirical basis for that consensus is quite weak, but central bankers have come to view restoring 2 percent as a test of their credibility.

It’s also true that even 4 percent inflation comes as a shock and sudden source of uncertainty after decades during which inflation was low enough that most people didn’t think about it at all. Indeed, conservation of mental effort — simply not having to worry about future prices — may be a significant benefit of low inflation.

On the other side, policymakers used to believe that an unemployment rate below 4 percent was basically unattainable without runaway inflation. But they were wrong: In the late 2010s unemployment fell into the 3s without accelerating inflation.

So at this point policymakers are more or less expected to achieve results that would have seemed wildly unrealistic for most of the past 40 years: 2 percent inflation and unemployment in the mid-3s.

How far would we have to fall short of these goals to say that the attempt at a soft landing failed? Last June, the economist Larry Summers declared that controlling inflation would require 7.5 percent unemployment for two years; that would clearly be a hard landing. Other economists, like Jason Furman and Mohamed El-Erian, have suggested that inflation might remain stuck at or above 4 percent for a long time, which Furman calls the “no landing” scenario.

But less extreme outcomes might fall into a gray area. What if unemployment rises to only 4 point something percent? What if, as Joseph Gagnon suggests, unemployment stays low but inflation levels off at around 3 percent?

In any of these cases, we’ll probably end up arguing about definitions.

But why worry about a hard landing? There are actually two reasons. First, inflation may have a lot of inertia, making it hard to slow. Second, the tools we normally use to control inflation are blunt and imprecise, creating a high probability that we’ll get it wrong one way or another.

The problem of inertia

Whenever inflation becomes an issue, people begin invoking the specter of the 1970s. The standard story about what happened then goes like this: A combination of bad luck (wars and revolutions in the Middle East) and bad policy (printing too much money and ignoring the inflation warning signs) allowed inflation to become “entrenched” in the economy. And purging that entrenched inflation was extremely costly.

What do we mean by entrenched inflation? Some prices, like those of oil or soybeans (or international shipping) fluctuate constantly. Many prices and most wages, however, are revised only at intervals — for example, a typical employer gives its workers contracts that set their pay for the next year. And these price revisions aren’t coordinated. In any given month, some prices and wages will be reset, but most will have been set some time in the past.

What this means in times of sustained inflation is that many of the economy’s players are caught up in a game of leapfrog. Every time they reset prices, they’ll raise them substantially, even if demand for their products is weak. That’s partly to try to catch up with other players’ price increases since their last reset and partly to get ahead of future price hikes by their suppliers and competitors. So inflation becomes self-sustaining unless something breaks the cycle.

One way to break the cycle might be to impose price controls — simply order businesses to stop hiking — or, in a sufficiently cohesive society, to get all the major players to agree to stand down. Such direct approaches have sometimes worked. Price controls did help contain U.S. inflation during World War II. In 1985 Israel engineered a big fall in inflation at relatively low cost by getting major unions and companies to agree to a pact enforcing wage and price restraints.

But it’s hard to come up with other successful examples of imposed or negotiated price restraint. Richard Nixon’s 1971 price controls led to shortages and effectively fell apart.

Last year Viktor Orban of Hungary — yes, the darling of the American right — tried to suppress inflation with selective price controls; his effort also failed, and Hungary currently has the highest inflation rate in the European Union.

What if policymakers can’t legislate or negotiate inflation down? The standard answer — ugly, but time-tested — is to deliberately weaken the economy: use contractionary policies — tax hikes, spending cuts or, usually, higher interest rates — to suppress overall spending. Faced with weaker demand for their products, businesses will raise prices and wages more slowly; as they see other companies doing the same, their price hikes will become even smaller, and inflation will gradually ramp down.

The good news about this approach is that it definitely works. It is, in fact, how the inflation of the 1970s was brought under control. The bad news is that it can be immensely costly, because businesses squeezed by weaker demand may lay off many of their employees before inflation has come down to acceptable levels.

Exhibit A is the story of the 1980s. Inflation came down from around 10 percent at the beginning of the decade to around 4 percent when Bush the elder won his election. But along the way there was a huge bulge in unemployment, which didn’t get back down to its 1979 level until 1987. Here’s a graphic that illustrates the costs:

The shaded area shows the excess unemployment above the 1979 level. In the jargon, a “point-year” of unemployment is one percentage point of excess unemployment for one year; the disinflation of the 1980s appears to have cost 15 point-years. Last year Larry Summers explicitly argued that disinflation this time around might be comparably difficult. Hence his horrifying pronouncements about how much unemployment we’re going to need.

There are two counterarguments. One is that inflation in 2023 isn’t entrenched the way it was on the eve of the ’80s disinflation. Back then, almost everyone expected high inflation for the foreseeable future. You can see these expectations in the wage settlements major employers were making with unions: On average, new contracts granted a 9.8 percent wage hike in the first year and 7.9 percent annually over the life of the contract. Companies wouldn’t have been willing to do that unless they expected rapid growth in both the cost of living and the wages their competitors were paying.

We don’t have comparable numbers today, because private-sector unions have nearly disappeared. But surveys suggest that businesses expect their costs to rise by less than 3 percent over the next year, and workers similarly only expect their earnings to grow around 3 percent.

The other reason to question analogies with the 1980s is that it might be possible to cool the economy without causing big job losses. Some economists argue that other measures, notably unfilled job openings and the rate at which workers are quitting their jobs, are better indicators of economic overheating than the unemployment rate. Both are elevated, but both have come down substantially over the past year without any rise in unemployment.

What does the data say? One key observation is that despite the fact that unemployment hasn’t (yet?) gone up at all, inflation is down a long way from its peak:

On the other hand, there’s still a case to be made that this has been the easy part — that for the past few quarters “underlying” inflation (a slippery concept) has been moving sideways rather than down. That is, while measures of underlying inflation are clearly lower now than they were in early 2022, it’s not clear that they’ve come down since, say, last November.

Why might inflation still be high? Perhaps because the economy still seems to be running hot, for example with a much higher ratio of unfilled job vacancies to unemployed workers than was normal in the past.

I’d say that even if inflation is moving sideways rather than down, that’s enough to refute some of the extreme hard landing stories — Summers’s claim that we needed a bout of 7.5 percent unemployment was based partly on the view that unemployment needed to rise to 5 percent just to keep inflation stable, which doesn’t look plausible at this point. But we don’t know how hard it will be to squeeze out those last two points of inflation. I’d say that data over the next few months should give us a lot more clarity. But I’ve been saying that for many months, and the inflation numbers still keep offering support to both optimists and pessimists.

But suppose, for the sake of argument, that we take the optimistic view that we don’t need a big rise in unemployment to tame inflation. Unfortunately, that’s no guarantee that we won’t have surging unemployment anyway.

The fool in the shower

Policymakers have great power over the economy, at least in the short term. A famous study by David and Christina Romer studied Federal Reserve minutes to identify episodes in which the Fed deliberately sought to slam on the economic brakes or step on the accelerator. They found that what the Fed wants, the Fed gets.

But power isn’t the same as precision. The Fed normally tries to manage the economy by setting targets for short-term interest rates, which it has no trouble achieving. But suppose the Fed raises its interest rate target by one percentage point. This will surely, other things equal, lead to fewer job openings, lower inflation and probably a rise in unemployment. But how big will these effects be? Nobody is really sure.

Fed officials, I’m sure you’ll be reassured to hear, know what they don’t know. That’s why they constantly say that their policy is “data-dependent” — they’ll adjust their actions based on what they see happening.

But keeping a close eye on the data isn’t as helpful as it might seem. For one thing, much economic data lags months behind the actual state of the economy and is subject to frequent revisions. Even more important, changes in policy don’t have immediate, visible effects.

Think about one of the main ways Fed policy affects the economy: via housing construction. A hike in the interest rates the Fed controls may not immediately filter through into a rise in mortgage rates; higher mortgage rates take time to show up in reduced housing starts and even longer to show up in a decline in the number of homes under construction; and the effects of reduced construction on things like retail sales add yet another lag.

Because of all these lags, policy that is too data-dependent — that reacts strongly to the latest numbers — can end up being destabilizing. Milton Friedman is said to have used the metaphor of “the fool in the shower,” who is alternately frozen and scalded because he’s constantly adjusting the taps in response to the current water temperature.

So where are we now? The Fed began raising rates in March 2022; almost every economist I listen to agrees that it was right to do so. But the job market still looks strong, and inflation is still above target. Does this mean that the Fed hasn’t done enough? Maybe. But it’s also possible that the Fed has already done too much but we haven’t yet seen the effects of past rate hikes. I know perfectly reasonable, well-informed economists holding both views.

Here’s where the metaphor of the soft landing actually works pretty well. Imagine a plane that is physically capable of making a smooth descent and touching down gently. But the pilot is trying to navigate through heavy fog with minimal visibility, and the instrument panel gives only an unreliable estimate of the plane’s altitude five minutes ago. Obviously it’s easy to see how the plane might either make a crash landing (i.e., experience a recession) or overshoot the runway entirely (suffer persistent inflation).

So can we manage a soft landing? Between the possibility that inflation will be sticky and hence hard to bring down, and the difficulty policymakers will inevitably face in sticking the landing, it’s hard to give us better than even odds of pulling it off.

But maybe we should step back and take the larger view. Covid-19 was an enormous shock to the economic system, made worse by Russia’s invasion of Ukraine. Yet we made it through the pandemic recession with remarkably little widespread hardship — in the fall of 2021, according to a Federal Reserve survey, 78 percent of Americans reported that their financial situation was at least OK, the highest percentage since the survey began in 2013. (We don’t yet have results for 2022.)

That’s a huge success story and will remain a success story even if our landing is bumpier than we’d like. The job market is so strong that even if we have a temporary rise in unemployment, it won’t create that much hardship. Inflation is above target but nonetheless at a level Americans found quite tolerable in the past, so if it persists longer than the Fed would like, that won’t be a disaster. Unless we have a really, really hard landing, the overall story of the postpandemic economy will be one of remarkable resilience.