In my experience, economic pundits all too often suffer from S.A.D. — not seasonal affective disorder, a form of depression that afflicts many people during the dark winter months, but seventies analogy disorder, a propensity to react to any hint of inflation with the assertion that the stagflation of the 1970s has made a comeback.
There was a lot of S.A.D. during the early Obama years, with dire pronouncements about the inflationary effects of budget deficits and growth in the money supply. As I wrote at the time, such concerns were foolish in a depressed economy.
But beginning in 2021 we really did have an inflationary surge, and it wasn’t entirely silly to worry that getting inflation back down to an acceptable level would require high unemployment, just as it did after the ’70s. Most famously, Larry Summers, who had gained a lot of credibility by correctly predicting the inflation surge, predicted that disinflation would once again be extremely costly.
That second prediction hasn’t aged well; inflation has come way down without any rise in unemployment. But Summers hasn’t thrown in the towel. The inflation of the 1970s came in two waves, and Summers recently circulated a chart purporting to show that recent disinflation is following the same track as disinflation in the mid-70s, which we know was followed by a major relapse.
Summers soon found himself the target of ferocious criticism, with some accusing him of “chart crime.” But the important issue wasn’t how the chart was constructed, it was the lack of context. As Mike Konczal of the Roosevelt Institute has pointed out, the mechanics of the mid-70s disinflation didn’t look at all like what we’ve seen recently. In the ’70s, disinflation was associated with soaring unemployment, while this time it has been associated with falling unemployment:
This strongly suggests that the mechanism behind the Biden disinflation has been fundamentally different from the mechanism behind the Ford disinflation. The story that most easily fits the facts is “long transitory” — the gradual resolution of economic disruption caused by Covid and its aftermath.
That said, however, does the ’70s resurgence of inflation after what looked like some success on that front have any lessons for today? The first step, surely, is to try to figure out what happened in the ’70s.
Here’s the Fed’s preferred measure of underlying inflation over the course of the ’70s, along with something else that had a huge influence during that period: the real price of energy, which, having surged during the Arab oil embargo of 1973, surged even higher after the Iranian revolution:
You might think that because core inflation specifically excludes the direct cost of energy, it shouldn’t have been much affected by the price of oil. But energy prices have a substantial indirect effect on other prices, because they affect the cost of doing business. Also, in the 1970s many wage contracts included cost-of-living allowances, so that energy prices drove wages. More tenuously, surging oil prices may have driven inflation expectations, which in turn pushed up actual prices.
One way or another, it’s hard to escape the conclusion that the resurgence of inflation after a pause in the mid-70s had a lot to do with a global energy crisis, which doesn’t seem to be part of what we’re expecting now.
That said, inflation did creep up even before the second oil shock. Why? It’s true that unemployment came down substantially after the 1974-75 recession, but it was still higher than it had been in the 1960s. Yet the U.S. economy was behaving as if it were somewhat overheated. Why?
The conventional answer is that the natural rate of unemployment had risen, so that the economy needed higher unemployment to keep inflation down. Indeed, the Congressional Budget Office estimates that the natural rate rose substantially from the 1960s to the mid-70s, before beginning a protracted decline:
The trouble with this story is that it’s more of an assumption than a result: Essentially, the budget office estimates a higher natural rate because inflation was rising, so the reasoning is somewhat circular. And efforts to pin down the reasons the natural rate might have risen aren’t all that persuasive.
Still, it’s the best explanation we have. Are there reasons to believe that something similar might be happening now?
For a while last year it seemed possible. The unemployment rate is only one measure of labor market tightness, and last year other measures — notably the vacancy rate and the rate at which workers were quitting their jobs — were much higher than you might have expected given the unemployment rate, suggesting that, as in the 1970s, controlling inflation might require higher unemployment rates than in the past.
But recently labor markets have been normalizing. Here’s the relationship between unemployment and the quits rate, which has arguably been a better indicator of labor market tightness than vacancies, which have trended upward over time:
For a while quits were unusually high given unemployment, suggesting that any given unemployment rate might be more inflationary than in the past. But that disparity has disappeared over the past few months, presumably reflecting the economy’s recovery from long transitory; we’re now right back at the historical relationship.
So at this point the evidence doesn’t seem to support the view that we’re on track for something like the experience of the later 1970s, in which inflation crept up despite relatively high unemployment, let alone for the kind of inflation explosion that followed the second oil shock.
Does this mean that we’re out of the woods on inflation? Not necessarily. The economy still seems to be running hot, so there is some possibility that inflation will reignite. And anyone who has complete confidence in either current data or current analysis is being foolish: If there’s one thing economists should have learned from the past few years, it’s the need to show some humility. It’s all too possible that the data, the models or both are missing key aspects about what’s going on.
But the idea that the 1970s offer a useful template for where we are now looks quite wrong. We should indeed study the ’70s and learn what we can from them. But making direct analogies between now and then is just S.A.D.
Quick Hits
It’s about the supply side.
S.A.D. during the Obama years.
Britain is looking a lot more like the ’70s than we are.
