JPM with the note:
Just under a year ago, theFedcompleted its framework re-view and officially changed its monetary policy strategy to a flexible average inflation targeting (FAIT) approach. With the acceleration in inflation since then, PCE inflation now averages 2.0% over the past three years. Incoming months aver-age inflation over this three-year lookback period is likely to move even higher. Did the Fed zig just as the economy was zagging? No. The framework review was motivated by the challenges that very low real interest rates present to central bankers. This hasn’t gone away; recently forward real interest rates five-to-10-years ahead returned to negative territory(Figure 1). The specter of the zero lower bound continues to haunt the world’s advanced economies.
The persistence of low real interest rates justifies the Fed’sembrace of FAIT, in our opinion. However, challenges loom further out in the business cycle—challenges that may have been caused or compounded by FAIT. The commitment to overshoot the inflation target(in more than a transitory manner)may require the Fed to push the unemployment rate below its sustainable minimum. The rest of this note discusses whythis raises the risks that the cycle ends with a hard landing. In particular, we examine certain historical properties of the behavior of the unemployment rate. Among those properties is a tendency for the unemployment rate to either move up a lot, or not move up at all. Moreover, when it moves up a lot it also moves up quickly.
We believe these properties highlight the challenge to achieving a soft landing. Moreover, FAIT has likely made this challenge ever greater. However, this does not mean we are fated for a hard landing. The Fed has arguably been more nimble in managing the business cycle since the 1980s. But this re-quired the Fed to take a forward-looking approach.
Shortly after the FOMC moved to a FAIT strategy many Committee members downplayed the importance of economic forecasting in setting monetary policy: “outcomes, not out-look.”We don’t think downplaying of forecasts was a necessary or logical conclusion of the move toward FAIT. Moreover, if the Committee abandons forecasting and is purely reactive we believe it would amplify the risk that FAIT ends with a hard landing.
Most macroeconomic models, including those often used in policymaking institutions, treat economic cycles as symmetric fluctuations around trend growth. In these models the labor market generally inherits these symmetric properties. Figure 2 presents a series familiar to readers, the unemployment rate since WWII. The COVID-19 recession, being sui generis in so many ways, is omitted for ease of seeing the other cycles.
There are a number of plainly visible properties of this series that are inconsistent with the idea that business cycles are symmetric fluctuations around a trend. First, the unemployment rate rarely spends time below its natural rate—i.e.,the rate consistent with maximum employment that doesn’t generate accelerating inflation, usually estimated to be between 4% and 6%—although it often spends time above it. One can also see this in the histogram of the unemployment rate (Fig-ure3), which is clearly skewed to the right.
The second property of this series that is inconsistent with the symmetric model of the business cycle is the speed of unemployment rate moves. It’s been said that the business cycle goes down by the elevator and up by the escalator. The behavior of the unemployment rate does the opposite—up by the elevator, down by the escalator. Table 1 presents the speed of unemployment rate moves from their peaks and troughs associated with the closest NBER-dated contraction and expansion. In recessions the unemployment rate moves up about twice as fast as it moves down in expansions.
The third noteworthy property of unemployment dynamics is the persistence of unemployment rate increases. Returning to the above metaphor, most people don’t take the elevator up just one floor, and when the unemployment rate goes up, ittends to go up a lot. There are a number of ways of seeing this, but perhaps the simplest is in Figure4, which plots the 12-month change in the unemployment rate. Outside of recessions this series is rarely positive. But when this series ex-ceeds 0.5%-pt the economy is always in recession.
The combination of these three features of the unemployment rate spells trouble for the Fed’s economic outlook. In June, the median FOMC participant saw the unemployment rate getting to 3.5% by the end of 2023. This is 0.5%-pt below where the median participant saw the lowest sustainable unemployment rate. And to paraphrase Herbert Stein, something which is unsustainable must stop. For the Fed to get the unemployment rate back to its sustainable 4% rate would apparently imply a tightening in monetary policy to slow growth and push the unemployment rate up 0.5%-pt.
As we saw earlier, when the unemployment rate increases0.5%-pt, it increases much more, and does so in a hurry. This maybe “merely” an empirical observation, but it is a robust one, and so should be taken seriously. Economists haven’t settled on an explanation for this regularity. It may reflect nonlinearities inherent in private sector behavior, perhaps things like downward nominal wage rigidities, external financing constraints, etc. Or it may reflect what’s been called stop-go monetary policy: the tendency—particularlybeforeVolcker—to keep the expansion going until inflation got so high that the Fed had to stop the expansion.
If stop-go policy is the explanation, or part of the explanation, for the asymmetric behavior of the unemployment rate, then there is reason for hope that the expansion won’t end in a premature hard landing. Arguably the Fed successfully achieved soft landings in the mid-90s and in 2019, and part of that success was due to changing course when momentum indicators suggested they had overdone it. But if the FOMC sticks with an “outcomes, not outlook” policy it would keep policymakers’ eyes firmly glued to the rearview mirror, increasing the chances of steering the economy to a hard landing.