Richard Drew/AP Photo
A television screen shows the rate decision of the Federal Reserve as traders work on the floor of the New York Stock Exchange, November 1, 2023.
This article appears in the February 2024 issue of The American Prospect magazine. Subscribe here.
Back in early 2021, as President Biden’s $1.9 trillion American Rescue Plan was being drawn up, economist and former Treasury Secretary Larry Summers was concerned. Writing in The Washington Post, he fretted that “there is a chance … [it] will set off inflationary pressures of a kind we have not seen in a generation, with consequences for the value of the dollar and financial stability.”
When inflation did indeed surge in later 2021, Summers claimed vindication, and proposed a solution: mass unemployment. “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,” he said in a speech in June 2022.
This recommendation was based on some classic economic concepts, as Summers explained to Jordan Weissman at Slate the following month. The first is the “non-accelerating inflation rate of unemployment,” or NAIRU. This is supposed to be the Goldilocks unemployment level at which as many people have jobs as possible without prices rising out of control. The second is the “sacrifice ratio,” which is the additional unemployment supposedly needed to bring inflation down.
Summers assumed that NAIRU was 5 percent; raising unemployment up to that level would hold inflation in place, but getting it to actually come down would require more. He estimated that the sacrifice ratio was two percentage points of unemployment (sustained over a year) per point of inflation, and that the Federal Reserve would need to cut inflation by 2.5 points. That was how Summers got the figures quoted above. He figured we needed five point-years of unemployment (2.5 times 2). Starting from the 5 percent unemployment baseline, you could add five point-years with either 10 percent for one year, 7.5 percent over two years, or 6 percent over five years.
In an interview with Bloomberg News, Summers predicted there was only a 15 percent chance that “it’s all going to work out well.” Others were even more pessimistic. Former Obama administration economic adviser Jason Furman estimated the sacrifice ratio was six—three times higher than Summers. If that figure were right, it would have meant we had to add 15 percentage-point years of unemployment to halt inflation. (Even the peak during COVID was below 15 percent!) Arguments like these helped prompt the Federal Reserve to hike interest rates at an extreme pace, going from zero in early 2022 to 5.3 percent at time of writing.
Fast-forward to November 2023 (the latest period for which figures are available), and the Fed’s preferred inflation measure is down to about 2.6 percent. Over that entire period, the unemployment rate never hit 4 percent. Whoops!
Now, I’ll admit that when I took the other side of the argument in February 2021, I certainly downplayed the political risk of inflation. I thought that the population would appreciate a quick return to full employment more than a period of modest price increases, but if President Biden’s approval rating on the economy is any indication, this was badly wrong.
But while Summers and Furman might have been right to predict inflation was coming, they were also completely wrong about what to do about it. It’s a good time to assess what the problems might be with their economics.
THE UNDERLYING THEORY BEHIND NAIRU and the sacrifice ratio is based on something called the Phillips Curve. This is an observed relationship between unemployment and inflation; if unemployment is low, inflation tends to be higher. The theory then assumes that future inflation is based on current inflation, plus expectations of future inflation, plus any deviations from the NAIRU level of unemployment.
This model isn’t entirely implausible, and it worked well enough when applied to the inflation of the 1970s. But it simply did not work today. What gives?
The most convincing argument for what happened I’ve seen comes from Mike Konczal at the Roosevelt Institute. In a recent paper, he theorized that there are two basic possibilities for why inflation decreased: Either demand fell or supply increased. In the first situation, one would expect lower consumption as people bought less stuff, driving the prices down, while in the second one would expect more consumption as production increased, driving down prices and causing people to buy more stuff.
Importantly, lower consumption would be consistent with job loss; there’s no reason to keep workers around if nobody is buying. Higher consumption would mean workers would stay and jobs would be added.
Inflation is calculated by the price of a representative basket of goods and services. Konczal looked at 123 of them in the “core” basket (that is, excluding food and energy, which tend to gyrate a lot) and found that 73 percent of goods and 66 percent of services showed prices falling while supply increased.
In short, we’re mostly looking at the second situation. Inflation was transitory, largely caused by supply chain snarls resulting from the pandemic and Russia’s invasion of Ukraine that have since been sorted out. That fits with the low unemployment of this period. Insofar as high demand was responsible for some inflation, the fact that it fell without any increase in unemployment suggests that we were high on the slope of the Phillips Curve (if there is such a thing)—meaning that prices would fall rapidly in response to only a small change in demand.
Now, orthodox economists have a response to this. According to their theory, a “costless disinflation” is possible if the Fed’s communication is sufficiently credible. If everyone believes their announcement, then expected future inflation will fall to the Fed’s desired rate, thus reducing the need for any unemployment above NAIRU.
But this is a peculiar idea. Fed “credibility” cannot be directly measured (just like NAIRU, by the way). The argument that inflation fell, therefore the Fed’s credible responses must be credited, borders on circular reasoning.
Even if we grant the idea for the sake of argument, the Fed’s actions do not have to be “credible” but merely believed. If every American accepted as a matter of faith that if Jerome Powell were to strip to the waist, anoint himself with sacred oils, and sacrifice a goat to Yaldabaoth, inflation would come down as a result, it would have precisely the same effect on expectations as raising interest rates.
This matters because the Fed’s actual inflation-fighting tool has paradoxical effects on inflation, particularly over the medium term. On the one hand, it does make investment more expensive, hence cutting spending and demand, most notably in the housing sector. But raising rates also directly causes the price of a new home mortgage to increase, and it reduces the supply of future housing—in 2022, new single-family home starts plummeted by about 20 percent—eventually putting upward pressure on rents.
If we accept that the Fed has a role to play in signaling its commitment to price stability, it would be better to have some more direct tools.
ASIDE FROM THE SUPPOSED POWER OF ECONOMIC spells and incantations, there are more practical lessons here. Why were supply chains so tangled by the pandemic and war? First, there was a lost decade of chronic high unemployment and underinvestment after the Great Recession that left the economy brittle and weak. Second was a much longer trend of monopolization, deregulation, and Wall Street–orchestrated transition to “just in time” supply chain management to juice profit margins rather than leaving any slack in the system in case of disaster.
Both these developments made America and the world vulnerable to supply shocks. On the margin, pandemic disruptions and surges in certain types of spending translated into higher prices rather than more production. A decade of underinvestment meant that infrastructure like roads and ports had no space for extra production, while slack-free supply chains meant that new buyers were forced to bid for scarce deliveries rather than bringing new capacity online.
Larry Summers was personally involved in both these developments. As director of the National Economic Council in the early Obama administration, he buffaloed fellow adviser Christy Romer into lowballing her estimate of the needed amount of stimulus based on his gut-check guess of what could be got through Congress. Before that, as Bill Clinton’s Treasury secretary, he was a fervent advocate of the kind of deregulation that produced the top-heavy, financialized economy of the 2000s and 2010s.
Deregulation and the insufficient stimulus created a hugely unequal economy where the benefits of economic growth largely flowed to the top 1 percent. But because the broad mass of consumers did not have much disposable income, the same 1 percent found nowhere to invest their bloated incomes. Investment flowed to ludicrous Silicon Valley boondoggles like Uber, DoorDash, or WeWork, which attempted to capture market share by selling their services far below cost, if not outright frauds like Theranos.
One of the great, albeit painful, virtues of the massive pandemic rescue packages was that they broke through much of that economic crust. Americans paid quite literally for these decades of economic mismanagement, but at this point we can say it was a price worth paying. A persistent increase in demand prompted companies to invest in badly needed new capacity, and they are now delivering more output at stable prices. If Summers had his way, inflation would have been tamed by impoverishing the citizenry so they could not afford to bid up prices, thus locking in a weak supply side once again.
Meanwhile, according to a study by David Autor, Arin Dube, and Annie McGrew, nearly 40 percent of the post-1980 increase in income inequality has been reversed, thanks to the red-hot labor market of the last few years. It is possible once again to invest in companies with a good old “sell above cost” business model. Thanks also to the Inflation Reduction Act’s climate subsidies, investment in green energy and manufacturing is surging. The Biden administration has also taken some belated action to roll back economic consolidation and prevent corporate price-gouging.
With inflation subsiding, it’s now an open question whether we will continue on the trail blazed by Biden’s half-improvised economic program, or gradually fall back into the same economic sandpit of the 2010s. The answer will depend in part on how seriously the complaints of Summers and company are taken. The most recent Fed minutes show a broad consensus around keeping rates high for a long time. But as noted above, the housing market has slowed markedly, and there is still a drastic housing shortage in many cities. So long as inflation remains quiescent, rates should be cut, and soon.