We expect the unemployment rate in 2020 to be 3½% —a percentage point below the NAIRU—and inflation 2¼%—¼ percentage point above its objective. I call this: “Missing on both mandates, but feeling good”! It would be a remarkable outcome, if sustainable. By sustainable, we mean over a period of two or three years, not indefinitely. To be sustainable, there has to be over this period a tradeoff between inflation and the unemployment rate that is stable. But that would contradict the modern accelerationist version of the Phillips curve. A stable short-run tradeoff would be feasible if inflation expectations remained anchored at 2% over this period, and outcome underpinned by the FOMC’s emphasis on symmetric inflation objective and commitment to 2% average inflation. In addition, monetary policymakers in the past have responded so aggressively to a material undershoot of the unemployment rate that it precipitated a recession. Given the modest overshoot as a result of the flat Phillips curve, the projected 2020 outcome might be sustainable for a few years.
The modern theory of the Phillips curve—sometimes referred to as the expectations-augmented Phillips curve—holds that there is only one unemployment rate consistent with stable inflation: the NAIRU, of course. But that is in the long run. What about over the short or medium term, call it two or three years? A key to sustainability is that inflation expectations remain anchored at 2% for some time even though actual inflation is modestly above 2% over that period. Sustainability, in turn, depends on the credibility of the FOMC’s commitment FOMC’s to the 2% objective over the long run and to the FOMC not responding so aggressively that it precipitates a recession.
The Long-Run Vertical Phillips Curve
Figure 1 shows the well-known vertical long-run Phillips curve and the implicit inflation dynamics associated with it. The NAIRU is shown as 4.5%, participants’ median estimate today. It is the only unemployment rate consistent with a stable inflation rate over the longer run. The arrows capture short-run inflation dynamics, that is, the movement of inflation when the unemployment rate is above or below the NAIRU. The arrows reflect the role of expectations in inflation dynamics. If markets believe that the FOMC is prepared to remain at an unemployment rate below the NAIRU, both inflation and inflation expectations will rise again and again, until the unemployment rate returns to the NAIRU.
The Short- (to Medium-) Term Stable-Tradeoff Phillips Curve
But does that really describe the world we live in today, at least over the medium term? It seems plausible that inflation expectations could move, to and become anchored at, 2% and remain there for some time if: (1) inflation moves only modestly above 2%, especially after a persistent period of being below 2%; (2) it is expected to remain above 2% for a limited period; and (3) the Committee remains credible about its commitment to a symmetric 2% inflation objective, consistent with an interpretation of price stability as 2% on average.1
We show such a short- to medium-run Phillips curve in Figure 2. Along this Phillips curve, long-run inflation expectations are assumed to be constant, at 2%. In this case, the FOMC faces many possible combinations of the unemployment rate and inflation, the so-called menu of choice. It could aim for point A, at the NAIRU and the inflation objective; or for point B where, in return for a substantially lower unemployment rate, it accepts modestly higher rates of inflation.
Anchoring Inflation Expectations at Point B
The first prerequisite is building credibility by successfully stabilizing inflation near 2% over a period of time, reflected in measures of long-term inflation expectations reasonably anchored near 2%. Having an explicit inflation objective helps. Asymmetric inflation objective in the context of undershooting 2% for many years makes the market more likely to accept a modest overshoot for some period without questioning the commitment to a 2% objective, especially the on-average interpretation of symmetric. Keeping the overshoot model, perhaps in the 2 – 2½% range, and having a flat Phillips curve, so a substantial undershoot of the NAIRU produces only a modest overshoot of 2%. With these supporting factors arguably in play, there is a good case for that inflation expectations could remain reasonably anchored at 2% for a few years.
Monetary Policy and Inflation Dynamics
How monetary policy responds to the outcome we project in 2020 will also be a key to sustainability. If the FOMC fears runaway inflation as inflation expectations become quickly unmoored, it will respond more aggressively to slow aggregate demand to a rate below trend to move the unemployment rate back to the NAIRU. In that case, of course, an outcome like we project in 2020 would be unsustainable, even over the medium term.2 On the other hand, as seems to be the case today, if the Committee is comfortable with a period of a modest overshoot of the 2% objective in the context of a well-communicated symmetric inflation
2Indeed, Yellen has said that the defining feature of her balanced approach policy rule is the large coefficient on the unemployment rate gap, 2.3 in her recent presentations of that rule. In that case, the rule prescribes that, if the unemployment rate were to fall one percentage point below the NAIRU, as we project in 2019 and 2020, the FOMC should raise the fund’s rate by 230 basis points above its neutral level. No doubt that such an aggressive response would precipitate a recession. And indeed it did! That rule fit the experience before the last two recessions well! See “Undershooting and Overshooting: Is This Time Different?” November 30, 2017.
objective, and especially following a period of inflation persistently below 2%, such an outcome is more likely to be sustainable at least for a few years.
But What about an Inverted Yield Curve and Overshooting Neutral?
Even if the inflation-unemployment rate tradeoff is sustainable over some medium-term period, there are still reasons to doubt the sustainability of the outcome we project in 2020.
1. There has never been a case where the unemployment rate fell well below the NAIRU that was not followed by a recession, as opposed to a stabilization of the unemployment rate for any period at its trough.
2. There has never been an inversion of the yield curve that has not been followed by a recession. 3. There has never been a case where the fund’s rate overshot its estimated neutral rate without being followed by a recession. And we may be heading toward an inversion of the yield curve and participants’ median rate projections show the fund’s rate moving above its estimated neutral level. Still, an inverted yield curve and overshoot of the neutral fund’s rate simply reflect periods of very aggressive monetary policy tightening that we would expect if policymakers thought that the inflation-unemployment tradeoff as reflected in our 2020 forecast was not sustainable, even in the medium term. And that is the question the Committee is wrestling with: How tolerant to be.
But Heed the Warning in the Vertical Phillips Curve
Blanchard tells a story similar to ours in his “The Phillips Curve: Back to the 1960s.” 3 But he ends with a stern warning: Exploiting the short- to medium-run tradeoff is like playing with fire. Possible, to be sure, under the right conditions and execution, and for a sufficiently limited period, but it carries risks monetary policymakers prefer to avoid. But we expect this framework will influence some policymakers as they decide how aggressive or cautious they believe the FOMC should be during the current tightening cycle.