Sean Rayford/AP Photo
An employee works on the assembly line at the BMW Spartanburg plant in Greer, South Carolina, October 19, 2022.
Since the onset of post-pandemic inflation, there have been two schools of thought about the causes and the cure. One was based on standard macroeconomics: There had been excessive aggregate demand, fueled by excessively generous relief money to those struggling during the pandemic. The cure was standard too: tight monetary policy, until workers felt enough suffering to cry uncle, accepting declines in living standards and lower real wages, thereby reducing wage and price pressures.
Larry Summers most forcefully exemplified this position in a speech at the London School of Economics: “We need five years of unemployment above 5 percent to contain inflation—in other words, we need two years of 7.5 percent unemployment or five years of 6 percent unemployment or one year of 10 percent unemployment.”
This became the conventional wisdom, adopted by central bankers: increase interest rates enough to raise unemployment and lower inflation to the arbitrarily selected threshold of 2 percent. In short: Put the burden of adjustment on workers.
The other school of thought focused on the unique characteristics of the pandemic and the war in Ukraine—the way in which these events induced changes in where people wanted to live and how they worked, as well as the extent of supply chain interruptions. Once the pandemic passed and markets had a chance to adjust, the surge in inflation would pass.
The debate this school had with those arguing for interest rate increases was not about whether normalizing interest rates above zero was a good thing, but whether the Federal Reserve and other central banks should go beyond that, raising interest rates to in excess of 5 percent, which they have now done. Team Transitory, as some referred to the group that expected market forces to tame inflation, argued that the expected benefit of such a rate hike was low, since inflation would come down on its own, and there were marked downside risks. For reasons I’ll explain below, from the moment inflation broke out it was clear that Team Transitory was right.
No one could be sure how fast inflation would come down; some in Team Transitory had more faith that markets would respond quickly to fill in the supply shortages than they should have. It took a little longer than hoped; markets were even more dysfunctional than many thought. But once the predicted responses took hold, inflation came down more quickly, without the alleged necessary increase in unemployment. And no one, no one, should think that there is a risk of an inflationary spiral.
Seldom in history do we see such a quick test of alternative theories. We’re not fully out of the water and we’re still facing new risks, which may (to stay with our metaphor) muddy the waters more: Just as no one could have predicted the Ukraine war in all of its dimensions, so too for Gaza. But we’re at a place where we in Team Transitory can rightly claim victory. And credit for this achievement goes not to the Federal Reserve and central banks around the world. Their misdiagnosis of the problem, egged on by the other school, which for symmetry I will label Team Persistence, imposed enormous, continuing risks on the global economy and those least able to bear the weight. Rather, the credit goes to the market—it eventually responded to the supply shortages.
Where Inflation Came From
Inflation is nothing but the increase in the prices of the millions of goods that make up the consumers’ market basket, so the first approach to inflation should be to determine where the inflation is occurring. Looking carefully at the data quickly revealed that the U.S. inflation was not due to excess aggregate demand but to specific sectoral problems, which, in time, the market could be expected to correct.
The underlying source of the post-pandemic inflation was the myriad of unprecedented supply chain interruptions and demand shifts, exacerbated by the Ukraine war. For example, automobile prices soared early, not because of a sudden surge in demand caused by money burning holes in people’s pockets, but because the car companies hadn’t ordered the computer chips they needed. They had been worried about an insufficiency of demand. We had the knowledge, the factories, everything necessary to produce cars except the chips. To be sure, we could have brought down the demand for cars to meet the limited supply by killing incomes—Summers’s recipe. But the fix would have been far worse than the problem, causing a massive amount of systemic damage.
The cost of housing has also been a persistent problem. But again, it was not a matter of aggregate demand, of macroeconomics, but of pandemic-induced shifts in patterns of demand. After all, a major determinant of the demand for housing is the size of the population, and unfortunately, due to the pandemic more than one million Americans died. Fewer people by itself should have led to lower demand and lower costs of housing. But the pandemic led people to want to live in different places. The advantages of cities seemed diminished, those of resort, rural, and suburban communities increased. It was obviously impossible to move houses from one place to the other, and construction takes time. Subsequently, there were shortages in some places, leading to price increases, and surpluses in others, resulting in price declines. But price adjustments are asymmetric, with the former exceeding (in absolute value) the latter, so that read in aggregate statistics as a surge in the average price of housing. (Of course, there were multiple other longer-term forces affecting prices, varying from place to place: In some places, Airbnb seemed to be driving up prices; in others, construction had been down even before the pandemic for a variety of reasons.)
Tom Williams/CQ Roll Call via AP Images
Former Treasury Secretary Larry Summers on September 7, 2022
Something else happened. Students of the American economy have noted the significant increases in market power in many sectors in the last two decades. Supply shortages increase market power, and predictably, this led to an increase in markups. Again, while one might expect a continuation of the slow secular increase in market power, careful economists predicted that the part of the inflation puzzle due to supply chain bottlenecks would disappear fairly quickly, leading markups to decrease.
One more example: The war in Ukraine, which broke out just as pandemic disruptions were winding down, led to large increases in the price of oil, energy, and food. Again, anyone believing in markets would anticipate over time disinflationary adjustments to the war, indeed deflationary adjustments. (Deflation is a fall in prices; disinflation is a reduction in the rate of inflation.) Not only wouldn’t the rates of increase in prices be sustained, but prices could be expected to decline with a global realignment of the world’s energy system, especially as new energy sources were brought online and as new conservation measures were taken—predicted adjustments which actually occurred.
The anticipation of disinflationary pressures was especially strong, given that the backstop price of renewable energy was so much lower than the market price. As for food, the U.S. and Europe had for half a century paid their farmers not to produce. If there was a real food shortage, they could consider paying their farmers to produce. As it was, the food shortage, like so much else, proved temporary, and there was no need to reverse a long-standing agricultural policy that makes no sense but has become an accepted part of our political reality. Once again, the disinflationary force was predictable.
Differences in Predictions
Remember, inflation is the rate of increase in prices, so for the price of food or cars or oil to continue to be a source of inflation, its price would have to continue to increase—there would have to be greater and greater shortages, not the one-time jump in prices as the pandemic shortage arose. No one on Team Persistence ever explained why, for example, there would be growing shortages of oil so the rate of increase in its price could continue, rising from $80 a barrel, for example, to $120, and then from $120 to $180. As we suggested, economic theory and simple common sense predicted just the opposite—and that is what happened.
Of course, no one then (or now) could be sure. No one knew, or could know, how long or how bad the pandemic or the war would be, or how much government support there would be and how long it would last. No one then could have predicted the war in Gaza, new potential shortages created by difficulties going through the Suez Canal, or food shortages exacerbated by climate change.
Standard economics predicted that individuals and firms would respond to this extreme uncertainty with precautionary behavior: Investment would be postponed and savings would increase. Among responsible macroeconomists, this increased anxiety over an insufficiency of aggregate demand. Responding to sectoral price increases by lowering still further aggregate demand was accordingly particularly problematic.
Barring new disasters, the presumption that price increases would be tamed seemed strong. Ironically, with hindsight, even Summers now agrees, writing in the Financial Times that “one should always have been aware that a substantial part of the increase in inflation was transitory.”
So how to explain the earlier view that to bring down inflation would require the enormous pain? In 2022, as the economy reopened, the labor market experienced a shock. There was a burst of people switching jobs, upgrading their skills and careers, and a major reallocation of workers to go with the sectoral changes of the reopening. This contributed to high nominal wage growth, though wages didn’t keep up with prices. It was also based on the premise that there was still a strong non-transitory element, with insufficient disinflationary pressures able, by themselves, to tame inflation. There was especially a worry of inflation perpetuating itself, through a wage-price spiral or an increase in inflationary expectations. Wage-price dynamics would result in the “sectoral” price increases discussed earlier seeping into non-transitory inflation.
Here again, the evidence was against Team Persistence even early in the post-pandemic era. First, estimates of wage-price dynamics today show that the pass-throughs of wages to prices and prices to wages are so weak that inflation dampens rather than spiraling out of control. This is not a surprise: Workers might like wages to increase when prices increase, or when their expectations of inflation increase, but they have little bargaining power.
Inflationary expectations throughout remained dampened, and rightly so: Those in the market largely belonged to Team Transitory, because they had a grasp of what was going on.
The evidence is now in. Over the past six months, core prices in the U.S. have increased at an annualized rate of just 1.9 percent, below the desired 2 percent target. Overall inflation in November was negative 0.1 percent (using the personal consumption expenditures index). The sources of inflation were transitory, as predicted by Team Transitory. Wage-price dynamics were not explosive, as predicted by Team Transitory.
Differences in Policy
The draconian policy measures suggested by many on Team Persistence were based not only on the premise of insufficient disinflationary sectoral forces, but on a simplistic application of macroeconomics 101. In one standard version, the Phillips Curve states that changes in the rate of inflation increase as the rate of unemployment falls, so disinflation requires an increase in the rate of unemployment. Economists have long been skeptical that this relationship was sufficiently stable to be relied upon, especially when relative prices change dramatically, as is the case here.
Team Persistence worried that the fact that inflation had not increased as unemployment fell in the years after the Great Recession implied that it would take a large increase in unemployment to get inflation down. And they worried too that, as individuals re-examined their relationship to work and their work-life balance, the labor market had fundamentally shifted, significantly increasing the “natural unemployment rate,” the rate at which inflation neither increased nor decreased, with some arguing that it had increased by 1.3 percent. If so, unemployment would have to incrase significantly just to prevent ever-increasing inflation.
In the U.S., there was much talk about the Great Resignation. It was clear that the pandemic was a major perturbation to the labor market, especially because U.S. policy (unlike that of so many other countries) did little to try to maintain workers’ connections with their firms. But no one knew whether there were permanent changes to the labor market, and if not, how long it would take for the market to return to something akin to normal.
All of which meant that, even if there were a substantial non-transitory element (as suggested by Team Persistence), no one knew what it would take to bring down inflation. If labor markets returned to pre-pandemic conditions quickly and worker bargaining power continued to be eviscerated as it had been in the past, then wage disinflation could set in with just small increases in unemployment.
There was still another element of uncertainty: Not only is the Phillips Curve itself unstable, but there is uncertainty about its shape. Even though there was less disinflation than many expected after the Great Recession, if somehow there was a decline in the natural rate of unemployment (the rate above which inflation fell) and disinflation increased rapidly with increases in the gap between the actual unemployment rate and the natural rate, then inflation might be brought down with little cost. (This is referred to as a nonlinear Phillips Curve.)
Again, the evidence is now in, and again, Team Transitory was right. Wage inflation was brought down quickly without the predicted “necessary” increase in unemployment; labor force participation for those 25–54, at 83.3 percent, is approaching a two-decade high. The idea that the natural unemployment rate would have fallen, so that there was scope for a nonlinear Phillips Curve to bring down inflation with little increase in unemployment, seems hard to take seriously. And as Mike Konczal noted, “Inflation has fallen at such a speed that it is no longer in the range predicted by a persistent Phillips Curve model” based on data over the past 50 years.
Wayne Parry/AP Photo
A new home under construction in Brick, New Jersey, July 10, 2023
The Excessive Demand Theory Was Wrong
Team Persistence blamed inflation on excessive aggregate demand caused by pandemic spending, especially from the American Rescue Plan. I’ve explained why we should have expected high levels of precautionary behavior, which would mute consumption and investment. One cannot just pretend that the pandemic was like any other ordinary event. A responsible economist would look to the data to see what was happening to each of the components of aggregate demand (consumption, investment, government expenditures, and net exports) and the total, to see if there was a surge, with a timing and magnitude that could account for the timing and level of inflation.
When one undertakes that exercise, as I did with my Roosevelt Institute colleague Ira Regmi, we showed that aggregate demand was almost consistently below pre-pandemic trends and Congressional Budget Office estimates of aggregate supply. Neither the timing nor magnitude of the short periods when real aggregate demand exceeded putative supply could account for observed inflation. The excess aggregate demand story simply had no legs. And our analysis explained precisely why: One could see in the data the buildup of precautionary balances and a slowdown in investment and other components of aggregate demand.
There were other reasons for being skeptical of the Team Persistence theory. The U.S. had a much larger fiscal package as a percentage of GDP. If one believed in the aggregate demand theory, the U.S. should have seen a much larger increase in inflation, and the inflation should have been particularly larger in non-traded sectors where increases in demand were limited by domestic supply, not by the much larger global supply. Again, a look at the data shows that that was not so. Europe’s inflation was also related to distinctive sectoral problems, particularly related to the pricing of gas and electricity. Countries like Spain that had better electricity price regulatory systems had better inflation performance—reinforcing our claim of the advantages of the sectoral approach to understanding inflation.
Interpreting the Success
Team Persistence has not conceded defeat so easily. They claim that it is only because they were so resolute, because they put so much pressure on central banks to raise interest rates fast and furiously, that this near-victory was obtained. At best, their claims are unproved, and at worst, just wrong.
It was not because of Fed action that the prices of cars, oil, food, or the host of other goods affected by supply-side interruptions came down, but because the underlying shortages were at least partially resolved. Mike Konczal’s Roosevelt Institute paper shows that the overwhelming part of the reduction in prices can be explained by supply-side increases—just as we noted earlier that the increase in inflation was attributable to supply-side interruptions and demand-side shifts.
This is not to deny that increases in interest rates and reductions in credit availability do lower aggregate demand, and in doing so add to the disinflationary pressures. But their actions were not at the center of what happened; they cannot be given credit. Indeed, with U.S. GDP growing in the third quarter at 4.9 percent, it’s hard to believe that it was a weak economy that brought down inflation.
Indeed, in many ways the Fed made things worse. It was more difficult to make the investments required to alleviate shortages. Higher interest rates made housing less affordable and less abundant. But higher interest rates made matters worse for other reasons as well.
Standard models of pricing by imperfectly competitive firms (and as we have already noted, many key markets are far from fully competitive) suggest that raising interest rates leads to higher prices, as firms with market power focus more on the immediate gains that they reap from higher prices.
Earlier, I questioned the importance of expectations as a determinant of inflationary dynamics. But more importantly, inflationary expectations are determined by what is actually going on in the economy, not just by central bankers. Market participants were well aware of the disinflationary forces we’ve discussed—seemingly more aware than Team Persistence at the time—and even before the Fed and other central bankers acted, inflationary expectations were muted. (Of course, in a kind of circular reasoning, Team Persistence might claim that they were muted only because market participants knew that central bankers would act.)
Policymaking Under Deep Uncertainty
I’ve explained the deep uncertainty that policymakers faced as inflation suddenly rose—uncertainty about the depth and duration of all the underlying disturbances to the economy and the shortages to which they give rise. Given the uniqueness of the events sparking post-pandemic inflation, economists advocating draconian measures like mass unemployment should have relied less on hypotheses and assumptions and more on a review of the evidence, and most importantly, a careful analysis of the consequences of what if they were wrong? What if Team Transitory was right, or mostly so? But humility was apparently in short supply.
On the downside, the risks of raising interest rates were considerable, both domestically and internationally, especially coming off a long period of near-zero interest rates. Many countries, firms, households, and financial institutions were deeply in debt, which could easily be managed when interest rates were zero, but not when they are at 5 percent. Within the U.S., we saw inklings of some of the other risks posed by rapid and large changes in interest rates, as the prices of interest-sensitive assets change markedly, with the bankruptcy of Silicon Valley Bank, the third-largest bank bankruptcy in American history, costing an estimated $20 billion. It was overleveraged in long-term, seemingly safe U.S. government bonds, but their price fell as interest rates increased. We managed to avoid this and other bank bankruptcies triggering systemic effects, but only because of strong and unprecedented interventions by government. Globally, a host of countries sit on the edge of default.
Because monetary policy acts with long and variable lags—effects typically continue well beyond a year—if the pandemic lasted longer and households and firms continued to be as cautious in their spending as they had been, tightening credit could well have led to a significant recession. Central bankers wouldn’t know whether they had raised interest rates too high until it was too late.
There are enormous costs to the kinds of unemployment, even if temporary, advocated by the Fed and Summers. Targeting an average unemployment rate of just 5 percent implies an unemployment rate of some disadvantaged groups much, much higher—in recent months, African American youth unemployment has been running at three to six times the average rate. Such high rates would almost certainly leave long-term collateral damage.
Graeme Sloan/Sipa USA via AP Images
Federal Reserve Chair Jerome Powell speaks to the press, December 13, 2023, in Washington.
All of this might be unnecessary if the alternative hypotheses of the sectoral origins of inflation turn out to be correct.
What Team Persistence and financial pundits were worried about was explosive inflation of the kind Argentina is going through now. But such concerns were simply not relevant to the U.S. As we already noted, workers simply don’t have the bargaining power to demand wage increases commensurate with price increases. And, as the economy normalizes, market power should decrease, implying markups decrease, implying, in turn, that wage increases won’t be fully passed through into price increases.
The United States is boasting of a soft landing, and I am optimistic that that will be the case. But if it does occur, it will be because of two offsetting mistakes: an over-tight monetary policy offsetting a more robust economy than was expected. The American Rescue Plan was a great bill that not only protected the vulnerable, but provided a macroeconomic cushion of spending during a difficult recovery and reopening, giving the United States the best GDP numbers of peer countries coming out of the pandemic. Similarly, going forward, the Inflation Reduction Act (IRA) produced not only a needed impetus for the green transition, but strong fiscal support for the economy, with tax credits and other subsidies that were originally estimated at under $400 billion, but are now estimated at two to three times that number.
Other countries are not so lucky: They don’t have the fiscal space to offset the major drag on the economy of tighter monetary policy. The anticipated global slowdown will, of course, have reverberations back in the U.S.
There Were Better Alternative Policies
Some aspects of these economic debates may seem almost theological. As I’ve noted, many pundits, especially in the financial world, believe everything hinges on key macro variables. While inflation couldn’t be blamed on inflationary expectations, these pundits worried that expectations might explode at any day, so resolute action by the central bankers was required.
Of course, tighter monetary policy wouldn’t have solved any underlying problem with the supply of food, oil, housing, or cars; as we’ve noted, quite the reverse. But by killing the economy, there would be no excess demand, and by definition (in their minds) no inflation. To be sure, it might have reduced inflationary pressures in some sectors. But some firms, where there is market power, would still have passed on the price increases in their inputs and some would have even increased markups more because of the higher interest rates. At best, inflation would have been slain but only at the cost of very high levels of unemployment—precisely what Summers was calling for.
The alternative prescription followed naturally from the diagnosis of inflation. It focused on addressing the underlying causes of inflation: if there is a labor shortage, provide better working conditions, child care, family leave policies, and so forth that make work more attractive. If there is an energy shortage, and if firms are hesitant to invest because they fear falling energy prices, government can provide a price guarantee. True, that shifts risk from the private sector to the public, but it’s a justifiable shift if one is worried about the macroeconomic effects of inflation and energy shortages. The IRA seems to be showing that modest subsidies can generate large results. We’ve also seen that at least part of our inflation is a result of the exercise of market power; stronger competition policies, especially in key sectors, could curb market power, thus markups, thus prices and inflation.
Concluding Comments
The events of the past three years have been truly remarkable and unprecedented. There aren’t even case studies to which we could turn to evaluate alternative courses of action. We have to bring to bear what knowledge we have about consumer and firm behavior, not in pandemics—we don’t have such data—but in the presence of uncertainty. This is an area in which there has been too little research, which is just one among many of the failures of modern macroeconomics. But we know that there will be an increase in precautionary behavior, with more savings and less investment.
With extremes of uncertainty, these will have first-order effects. Markets fail to provide the insurance individuals need—whoever heard of pandemic insurance?—so governments have to step into the breach. Governments around the world did, with enormous social benefits. Though we can’t be sure of the counterfactual, I believe there would have been enormous suffering if governments did not act strongly.
We’ve seen the enormous downside risks posed by the excessive tightening of monetary policy. On the other hand, the supply-side measures such as family leave and child care have large social benefits whatever turns out to be the case concerning inflation. If they help dampen inflation, so much the better.
Many economists seem to have strong antibodies toward governments doing what they should, especially through fiscal policy, and look for fault. (They seem to have more confidence in the bankers, former private equity professionals, and others running the Federal Reserve and central banks. Why that should be the case after the disastrous performance leading up to the 2008 crisis is a mystery.) No government does everything perfectly—no human institution is infallible. Money might have been better targeted, but that would have taken time, and in the pandemic, time was of the essence. All in all, the program of the Biden administration was on target. It did not cause the inflation. And those who blame inflation on excessive fiscal spending are not only just wrong; the policies that their unfounded theories have led to have imposed enormous risks on the global economy. As seems so often the case, it is those at the bottom—both in the U.S. and around the world—who were asked to bear those risks.
Team Persistence used standard textbook macroeconomics as the basis for simplistic monetary policies, and gave succor to those who would risk a hard landing and sacrifice the well-being of ordinary workers at the altar of the inflation holy grail of 2 percent, a figure made up out of thin air some 30 years ago. Thankfully, they were quickly proved wrong, before the full force of their misguided policies could take hold. Team Transitory, even if it was not listened to as much as it should have been, provided a needed warning, and may have curbed some of the excesses to which central bankers otherwise might have been prone.
This is not the first time that simplistic macro models misled policy: The disastrous shock therapy as Russia transitioned from communism to a market economy led to a 30 percent decline in GDP and shortening of life spans, not the soaring of living standards that had been promised. A straight line can be drawn between those policies and where Russia is today, once again demonstrating that economic policies can have large political consequences.
Good economic policy requires, as in this inflationary episode, a balance of micro- and macroeconomic policy. Flawed privatizations contributed to the Russian debacle. The deregulation of derivatives and the financial sector more broadly set the stage for the financial crisis of 2008.
Hopefully, in our next dramatic macro episode—and if history is our guide, there will be many of these—less attention will be paid to Team Persistence and those who rely on overly simplistic and unreliable macroeconomic relations, and more attention will be paid to those with more subtle and comprehensive analyses of the economic situation.
