By Greg Ip
Unemployment and inflation are near their lowest levels in decades. Who wouldn't love that?
Janet Yellen, for starters.
What looks like a dream economy could be a nightmare for the Federal Reserve chairwoman. Ms. Yellen's worldview assumes that when unemployment is this low -- 4.4% in August -- inflation should move up to the Fed's target of 2%. Instead, it may have stabilized around 1.5%. That presents the Fed with some unpalatable options: deliberately overheat the economy for years to get inflation back up, then potentially induce a recession to stop it from overshooting; or give up on the 2% target, which could hobble its ability to combat future recessions.
This isn't scaremongering: It's the logical consequence of how central banks believe inflation operates. At the center of their model is the Phillips curve, according to which inflation edges lower when unemployment is above its natural, equilibrium level and putting downward pressure on prices and wages. Below that natural rate, also known as full employment, inflation crawls higher.
This mostly describes how inflation fluctuates around its long-term trend over the course of the business cycle. Trend inflation, which prevails when the economy is right at full employment, is determined largely by public expectations, which are in turn influenced by central banks and their inflation targets. When consumers, businesses and workers expect 2% inflation, they set prices and wages in a way that makes actual inflation hit 2%.
Since the current expansion began in 2009, inflation has persistently fallen short of 2%. Most of the time, that could be chalked up to the ample economic slack left over from the 2007-2009 recession. Today, though, unemployment is around a 16-year low and below the Fed's estimate of its natural rate, 4.6%. Yet using the Fed's preferred gauge of "core" inflation, the price index of personal consumption expenditure minus food and energy, recently slipped to 1.4%.
There are three leading explanations.
One is that the economy actually isn't at full employment; either the natural rate has dropped or many unemployed aren't being counted properly. But history and mounting reports of labor shortages militate against that.
The second, and Ms. Yellen's preferred theory, is noise: One-off drops in prices such as for cellphone plans, prescription drugs and online purchases are masking the underlying trend. But one-off movements can't explain years of undershooting. As Lael Brainard, a Fed governor, noted last week, some one-offs, such as drug-price hikes last year, must be pushing the other way.
That leaves the third explanation: Trend inflation has fallen. Until recently, Fed officials scoffed at the possibility. They noted surveys that suggest the public still expects inflation to return to 2% and credit their oft-repeated promise to hit their 2% target. But are they fooling themselves? Expectations of inflation are determined in great part by what inflation actually has been, and after every recession since 1982, core inflation has averaged less than in the previous business cycle: 4.1% in the 1980s, 2.1% in the 1990s, 1.9% in the 2000s, and 1.5% since 2009.
In 2014, Fed staff slightly revised down its own assessment of trend inflation, according to minutes to the central bank's June meeting that year. Few senior officials were prepared to do the same, but that seems to be changing. Ms. Brainard cited academic research, surveys of the public and the behavior of inflation-indexed bonds as evidence that trend inflation has dropped between a quarter and three-quarters of a percentage point in the past decade.
To get inflation higher, the Fed would have to engineer the opposite of the past 35 years: a prolonged boom that drives unemployment below its natural rate until inflation returns to 2%. As actual inflation rises, so would expectations, locking in the higher trend. To achieve this, Ms. Brainard suggests interest rates shouldn't rise much more, if at all. The Fed is widely expected to leave rates between 1% and 1.25% next week, and doubts about whether it will raise them in December have grown.
Lower trend inflation has much graver implications for the economy than appreciated. In recent decades, it has taken ever bigger swings in unemployment to affect inflation. Some critics say this is proof the Phillips curve is nonsense; central bankers conclude that in fact, it has simply flattened.
Raising inflation half a point could require letting unemployment drop to around 3.5% and keeping it there for five years. Then, to prevent inflation from overshooting, the Fed would have to slow the economy and guide unemployment back over 4%. In theory it could do this gradually enough to avoid a recession; in practice, the number of times since 1948 when unemployment has gone up that much without a recession is zero, according to Goldman Sachs.
This approach could aggravate another worry: financial excess. If stocks and property look bubbly now, imagine what five more years of very low interest rates would do.
The alternative is to ditch the 2% target and accept 1.5% as the new inflation trend. Besides shredding the Fed's credibility, that would mean lower trend interest rates and thus less rate-cutting ammunition to fend off the next recession. If history is any guide, that recession would push trend inflation down further.
Both options are unappetizing, but the second distinctly more so. If Ms. Yellen eventually concludes lower inflation reflects a trend rather than noise, prepare for unemployment to drop much more and interest rates to stay low for a lot longer -- with an attendant rise in financial and economic volatility.
Write to Greg Ip at [email protected]