The Federal Reserve’s New Monetary Policy Framework

I very much appreciate the opportunity to participate in this volume honoring Marvin Goodfriend's contributions.¹ Marvin and I interacted often as we both served in the Federal Reserve System, comparing notes on developments in macroeconomics and monetary policy at numerous conferences and in informal contacts along the margins of Federal Open Market Committee (FOMC) meetings and elsewhere. Marvin had a huge influence on the study and practice of monetary policy within the Federal Reserve and more widely. Even where his ideas took time to filter into actual policy, he often framed the debate, forcing the skeptics to examine their arguments more closely.

For many years, I was one of those skeptics when it came to explicit inflation targeting. In fact, conference organizers liked to position Marvin and me as the pro and con on this topic. While I shared Marvin's objective of anchoring inflation and expectations around 2 percent, I saw advantages in keeping the goal implicit rather than explicit; expectations were becoming anchored at close to 2 percent in any event, and an implicit target might afford greater flexibility to respond to some types of shocks.

These debates were great learning opportunities for me. Marvin marshalled empirical evidence and embedded that evidence in the theory and practice of central banking over history. Marvin was open and honest about his views and the supporting evidence. And however much you might have differed, you couldn't doubt his focus on and devotion to bringing his considerable intelligence and deep learning to serving the interests of the Federal Reserve and the United States.

In the end, Marvin's analysis prevailed. In the aftermath of the financial crisis, I was won over to the view that the benefits of an explicit inflation target would exceed its costs. In 2012, after I left, the Fed adopted a 2 percent target.

My conversion and the explicit target emerged from the threat of the Federal Reserve missing both its employment and inflation targets on the low side, rather than from building the bulwark against high inflation that mostly animated Marvin's advocacy over the years. The major risk to inflation expectations as the country slowly recovered from the global financial crisis of 2007-09 was that they would fall be-low 2 percent, reducing nominal rates and limiting the scope for policy easing in the future. I saw an explicit target as helping gain support for additional monetary policy action from the members of the FOMC who feared that unconventional policies might cause much higher inflation down the road.

But the essence of Marvin's vision has been realized. The Federal Reserve has made an explicit public commitment to achieving 2 percent inflation over time, which should help discipline policy and firm up expectations against deviations from the target in either direction.

Importantly, the inflation target that Marvin advocated for rested on two closely related pillars of his analysis. First, that economic welfare was fostered by effective price stability and by public expectations that prices would remain stable (avoiding "inflation scares"). Second, that those expectations would be more durably anchored, and democratic accountability better served, by central bank transparency about its targets and its plans for meeting them.

Clearly the Federal Reserve has embraced both of those propositions. In addition to the explicit inflation target, policymakers have taken a number of steps in recent decades to be more open about their analysis and rationale for policy. In his essay introducing Marvin's paper on "Monetary Mystique" in this volume, Lars E.O. Svensson out-lines what the Federal Reserve has done to realize Marvin's objective of transparent policymaking. The major actions include: (1) announcing policy decisions immediately (1994); (2) publishing quarterly the projections of FOMC participants for output growth, inflation, the unemployment rate, and the appropriate path of the federal funds rate to achieve the FOMC's legislated objectives for "maximum employment and stable prices" (2008-2012); and (3) adopting and publishing a "Statement on Longer-Run Goals and Monetary Policy Strategy" (2012) that outlines the FOMC's view of its objectives and how it intends to pursue them.

Goals may be largely fixed by legislation, but strategies and the communication around them need to adapt to changing circumstances. In recent years — before the Covid pandemic — the Federal Reserve had been wrestling with how to adapt its targets, strategy, and transparency to a world in which central banks, including the Fed, struggled with getting inflation up to the 2 percent target in an environment of persistent disinflationary pressures. Those disinflationary forces were marked by weak demand, the effects of globalization and technology on costs, and surprisingly muted responses of inflation to low unemployment rates. During this time, very low equilibrium nominal interest rates raised pressing questions about whether existing policy strategies could be consistently successful in achieving the Fed's legislative objectives when the zero lower bound (ZLB) on rates could frequently limit the scope for easing policy in response to negative demand shocks.

The constraint on policy easing presented by the ZLB creates a potential asymmetry toward missing both the inflation and employment goals on the low side on average over time if, as in the strategy adopted in 2012, policy is always aiming just at its 2 percent inflation target. That bias may not be reliably overcome using unconventional policy measures, like asset purchases and forward guidance. In fact, inflation had persistently fallen short of the Fed's goal in the decade from 2009 through 2019, despite interest rates at zero and substantial asset purchases over much of that period. Moreover, inflation misses on the low side occurred even with unemployment rates that had declined to much lower levels than previously thought consistent with maintaining price stability. To be sure, the inflation misses were generally small, but they also were accompanied toward the end of this period by a downward drift in some measures of inflation expectations, raising questions about whether these expectations would continue to be anchored around the 2 percent target.

In response to this experience, the Federal Reserve ran a very public and transparent process to assess how it should alter its monetary policy framework — its monetary policy strategy, tools and communications — to raise the odds on achieving its legislated price stability and maximum employment objectives more consistently in this low natural rate environment. It announced the results in August 2020 in a revised version of its "Statement on Longer-Run Goals and Monetary Policy Strategy" and in a speech by Chair Powell that explained the changes and their rationale.²

Lars E.O. Svensson touches on the new framework at the end of his essay. In this piece I will dig a little deeper, evaluating it through the prism of the two pillars of Marvin's work previously cited — sustaining price stability and being very transparent about how that will be achieved.

The new framework

With respect to the objectives of monetary policy, the new framework retains the critical elements of the old framework. It kept the 2 percent inflation target as its definition of its price stability mandate. With respect to maximum employment, it continued to acknowledge that specifying an explicit numerical goal is unwise because the level of maximum employment consistent with stable prices is not directly measurable, is not under the control of the Federal Reserve, and changes over time for reasons unrelated to monetary policy.

But responding to the 2009-19 experience, the FOMC made some key changes in the specification of the maximum employment goal and the strategy for achieving its objectives. The previous statement hadn't defined maximum employment, but many observers, including many members of the FOMC, gave heavy weight to the unemployment rate and looked at the historical relationship of this variable to changes in inflation to gauge how close the economy was to this goal. Reflecting this approach, the old statement gave the median of FOMC members' recent estimates of the normal long-run rate of unemployment as an example of a measure of maximum employment. Because the history of this relationship had not been a good guide to future inflation in recent years, the new statement dropped this reference to the unemployment rate and added that "the maximum level of employment is a broad-based and inclusive goal." Although the old statement said that the FOMC looked at a "wide range of indicators" of labor market tightness, the new statement and its exposition by a number of FOMC participants has seemed to suggest not only a de-emphasis of the unemployment rate, but also increased attention to a wider array of other indicators, including labor market outcomes across population subgroups.

In addition, the monetary policy strategy for dealing with labor markets was altered in an important way. Policy would take account of shortfalls from maximum employment (e.g., unemployment rates above the estimated normal level), but not necessarily of estimated overshoots (e.g., unemployment rates below the estimated normal level).³ That's because in the 2009-19 period policy had been tightened in low unemployment periods to head off inflation, but experience had been that inflation would be quiescent at much lower levels of unemployment than had been expected. The FOMC continued to acknowledge that policy affected employment and inflation with a lag, but it would not run a tight policy to preempt projected inflation overshoots based only on actual and projected labor market conditions. Tight policy — interest rates being moved above the estimated neutral rate — would depend on already seeing unsatisfactory inflation outcomes.

On the inflation goal itself, the FOMC's new statement emphasized the importance of keeping inflation expectations well-anchored at its goal of 2 percent, but it worried that periods of below-target inflation would be more prevalent than above-target inflation given the ZLB problem, which would tend to pull expectations under 2 percent. To avoid this outcome, it would now seek to achieve inflation that averaged 2 percent over time. That means that when inflation has been running below 2 percent, monetary policy will aim to achieve inflation "moderately above 2 percent for some time." This has been labeled flexible average inflation targeting, or FAIT.

Sadly, we cannot know what Marvin would have thought about the new framework. I suspect he would have been very pleased with how the review was conducted: the process for arriving at the new framework and statement was announced ahead of time; the Fed held public "FedListens" sessions to get input from the public and from academics; it reported on the progress of its deliberations in the minutes of the FOMC; some members of the FOMC used speeches to keep us informed about the evolution of their own thinking; and, simultaneously with the rollout of the new framework, it published the staff analysis that the FOMC had as it considered its options. I also suspect he would have liked the emphasis on keeping inflation expectations anchored at the 2 percent target and the explicit rejection of raising the target, as some academics had been suggesting.⁴

Beyond these key elements I will not try to guess at how Marvin would have reacted. Below, in examining the new strategy, I will channel the principles he imbued in the Federal Reserve for securing price stability and enhancing transparency. But I know that my analysis would have been much stronger had I been able to benefit from the give and take with Marvin that was so important to the evolution of my thinking and that of countless others at the Fed.

Anchoring inflation expectations at 2 percent

As noted, the new framework grew out of a period in which inflation fell short of the 2 percent target and inflation expectations drifted down, despite very low interest rates and much lower unemployment than had previously been thought consistent with low, stable inflation. Keeping expectations from moving below 2 percent is especially important when the real equilibrium interest rate appears also to be quite low, making the ZLB an increasingly salient policy constraint.⁵

The FAIT framework is well designed to counter the disinflationary bias imparted by policy being constrained by the ZLB from time to time. FAIT promises to make up for inflation below 2 percent by aiming to run it "moderately above 2 percent for some time" — a flexible form of price-level targeting. That implies easier policies for longer than if the Fed were simply aiming to return inflation to 2 percent without the makeup. The point of the averaging is to make sure expectations are indeed anchored at 2 percent. In effect, deliberately aiming for inflation to exceed 2 percent for some time and likely allowing the overshoot of maximum employment necessary to achieve that results in an upward inflation bias that offsets the downward bias arising from the ZLB.

But there are other asymmetries in the new framework — beyond the makeup for actual undershoots — that also lean toward taking up-side risks on inflation and raise questions about how well adherence to the framework would anchor expectations in circumstances in which inflation wasn't so quiescent.

One key problem is that the strategy does not address what is to happen if there are persistent overshoots of the 2 percent target. It reaffirms that "the Committee seeks to achieve inflation that averages 2 percent over time" but follows that statement with the aim of making up for undershoots and doesn't address the opposite situation. Making up for overshoots would be required if persistent inflation over 2 percent resulted in longer-term inflation expectations rising above 2 percent. It would involve difficult economic decisions as it implies a need to deliberately run the economy below its sustainable potential for a time to lower inflation and inflation expectations.

One suspects the FOMC might opt for "opportunistic disinflation" in such circumstances — waiting for an external shock, rather than monetary policy, to create slack and lower inflation. That was the view of many members of the FOMC in the 1990s when inflation exceeded 2 percent, but at that time the FOMC was working in the context of an implicit target, not an explicit one, which afforded a greater degree of flexibility. In any event, a framework that addresses making up for undershoots but not overshoots would seem to risk a bias toward inflation over 2 percent if circumstances differ materially for a time from the disinflation pressures of 2009-19.⁶

A second upside inflation bias arises from the asymmetry of the response to deviations of labor markets from estimates of maximum employment. Shortfalls of employment from estimated maximums weigh on the side of accommodative policy, but actual or projected overshoots, by themselves, do not call for tight policy. In the past, the FOMC had increased the federal funds rate by enough to head off possible future inflation when projections suggested that declines in the unemployment rate in the absence of tightening were likely to result in future above-target rates of inflation — even before inflation or inflation expectations had risen into unacceptable territory. Lags in the effects of monetary policy made such preemptive moves desirable to avoid having to impose future output losses to bring inflation back down. In the new framework, the FOMC will continue to get ahead of unwelcome declines in inflation by running accommodative policy when it judges there to be slack in labor markets, but its scope to head off future unwanted increases in inflation would appear to be more constrained, risking overshoots of the target under some circumstances and the resulting greater variability in output.⁷

The choice of this asymmetric reaction function grows out of the experience from 2009-19 when the committee had overestimated the unemployment rate consistent with low stable inflation and in retrospect felt that had it waited to tighten more jobs would have been created more quickly with inflation still contained. If the relationship between slack and inflation is as attenuated as it seemed to be from 2009-19, then waiting to see actual inflation rise might not be very costly in terms of unanchoring expectations on the upside, because any rise in inflation would be small. But a steeper Phillips curve, say because of greater uncertainty and eroding credibility around the 2 percent target or a larger decline in unemployment relative to the natural rate, would impart a more definite inflation bias to policy.

The forward guidance on interest rates provided during the pandemic implemented this asymmetric response to labor markets in a particularly aggressive way. In September 2020, the FOMC adopted language that promised to keep interest rates close to zero "until labor market conditions have reached levels consistent with the Commit-tee's assessments of maximum employment and inflation has risen to 2 percent and is on track to moderately exceed 2 percent for some time." This meant that short-term rates would remain deeply negative even in the run up to full employment with inflation at the target and predicted to rise further, leading to a likely overshoot of maximum employment and continued upward pressure on inflation. This was more aggressively easy guidance than required by the new framework, which only called for policy to be accommodative before the economy had reached maximum employment, but it reflected the new emphasis on reaching full employment and not projecting increases in inflation.

As I write in early 2022, the economy has rebounded from the initial shock of shutdowns in the wake of the onset of the Covid-19 pandemic, inflation has surged to levels well in excess of the 2 percent target, and the labor market has tightened considerably faster than expected. That surge has reflected both strength in demand and Covid-related constraints on supply.⁸ The situation is complex and unprecedented, containing elements of adverse supply shocks that are always difficult for monetary policy to navigate.

To the credit of the Federal Reserve, as the persistence of inflationary pressures became increasingly evident, it pivoted quite rapidly toward an accelerated removal of the extreme accommodation it had put in place in the initial stages of the pandemic. It began to reduce its purchases of securities sooner and more rapidly than it had previously expected, and it is clearing the way for beginning to raise its target interest rate in March 2022, also much sooner than it or many observers had expected as recently as summer 2021. Although measures of short- and medium-term inflation expectations have risen to well above 2 percent, longer-term expectations remain anchored at levels consistent with 2 percent, suggesting the Federal Reserve retains credibility for achieving its long-run objective over time.

Still, this experience suggests that the asymmetries of the new framework could risk an upward drift from the 2 percent inflation anchor when the surrounding macroeconomic circumstances deviate from the damped demand, low-inflation environment of 2009-19. The Federal Reserve needs to consider and then spell out its approach to achieving its price stability goal if the prior disinflationary forces do not reemerge. To be sure, the FOMC can never anticipate and discuss all the situations it might face in the future. But it should be able to describe in a general way how it would expect to react if, for example, inflation persisted above its target and expectations began to drift higher despite estimated slack in labor markets or if wages and prices suggest that labor markets are approaching a level of "maximum employment" that is lower than previously expected.

Transparency

As indicated by the previous discussion of asymmetries and changing circumstances, the new framework is more complex than the old one, in which monetary policy was always targeted at 2 percent inflation and goal conflicts posed by supply shocks were subject to a "balanced approach" to both goals based on deviations on either side of inflation from target and of employment from the estimate of maximum.

Complexity challenges transparency. As Marvin emphasized, transparency is critical for policy effectiveness — the more accurately the public can predict what the central bank is going to do, the more rapidly the economy is likely to move toward central bank objectives. Transparency is also critical in a democratic society for accountability and preserving central bank independence. The previous regime could be approximated by relatively simple policy rules that could serve as benchmarks and guides for the FOMC and its observers. In the new framework, policy responses depend on how long the policy rate has been pinned at zero and inflation has fallen short of target and whether "maximum employment" is being approached from above or below, greatly complicating the explanation of strategy.

Although the process of coming up with and rolling out the new framework was very open, in some respects the new framework is a step back in transparency. To an extent, this is an understandable consequence of the inherent complexity and lack of experience with the new framework. And some of the lack of transparency results from a deliberate attempt to preserve flexibility.

Yet, I am convinced that the Federal Reserve can improve the transparency of policy under the new framework.

First, as I've already discussed, it needs to outline how it will deal with circumstances that differ materially from those that prompted the rethink — for example, adverse supply shocks, persistent and material inflation overshoots, and the waning effects of the disinflationary forces of the 2009-19 period. A crucial question is how monetary policy strategy would evolve in the framework if the problem becomes one of reducing inflation rather than getting it up to target.

Second, the new framework has redefined its "maximum employment" goal in ways that are less transparent. Both the old and new statements note that maximum employment isn't under Fed control but rather is "defined by nonmonetary factors that affect the structure and dynamics of the labor market." The previous statement, however, gave the most recent estimate of participants for the long-term unemployment rate consistent with its mandates as a reference point. To be sure, that estimate had to be interpreted along with other data on labor markets, wages, and prices to get a sense of whether the unemployment rate consistent with stable prices was shifting, but the long-term unemployment rate provided a guidepost.

The current statement characterizes maximum employment as a "broad-based and inclusive goal that is not directly measurable and changes over time" and omits the reference to participants' projections of the sustainable unemployment rate. The FOMC has yet to define what measures are encompassed by "broad-based and inclusive," with some participants emphasizing that they will be paying close attention to the labor market experience of low-wage and minority groups with-out discussing how that view intersects with the FOMC's price stability mandate. If economic agents are to understand and accurately anticipate monetary policy, the FOMC needs to spell out more clearly what it means by broad-based and inclusive. 

Third, another aspect of the new strategy that would benefit from additional explanation is how flexible average inflation targeting will work in the real world. I have considerable sympathy with the "flexible" piece of FAIT — it's in line with my position on inflation targeting in my discussions with Marvin that an implicit target gave the FOMC greater scope to deal with unexpected and unusual shocks arising, for example, from financial market developments. But the FOMC has an explicit target — 2 percent inflation over the longer run — and emphasizes the gains for both parts of the dual mandate from having expectations anchored at that level. In the circumstances that the framework was designed for — avoiding a downward drift in expectations from persistent undershoots of the target — the committee has said that if inflation did fall persistently short of target, it would "aim to achieve inflation moderately above 2 percent for some time." 

As inflation came to exceed 2 percent over the course of 2021, market participants were struggling to gauge the FOMC's definitions of "moderately" and "for some time." I wouldn't expect precise definitions of those words — that would take the F out of FAIT — but some guidance would be a useful enhancement of transparency. Are there ranges around the level of inflation or the length of its persistence over 2 percent that would stretch the definition of these words too far? Also, after "some time" of "moderate overshoots" would the FOMC resume targeting 2 percent inflation, or would it be content with inflation moderately over 2 percent and count on future shortfalls at the ZLB to bring the average back to 2 percent? How would FAIT work over time — what's the end game once the average has been secured at 2 percent?

Conclusion

The new framework is well adapted to the circumstances that faced monetary policy from 2009 through 2019. But it is complex and incomplete, with a deliberate lean toward taking upside inflation risks to offset the downside risks inherent in very low interest and inflation rates. It has been implemented initially in a global pandemic that is unprecedented in modern times, which has had complex and difficult-to-predict effects on aggregate demand and supply and prompted extraordinarily expansionary fiscal and monetary policies. These forces — the virus and the policy response — have produced very high and persistent inflationary pressures, much higher and more persistent than the Federal Reserve or most mainstream economists had predicted or had been contemplated in the new statement. To the Fed's credit, the FOMC has reacted relatively quickly to the emergence of those pressures by beginning to dial back its highly accommodative policy.

The Federal Reserve has said that it will review the new framework after five years. That review and the opportunities to clarify the framework in the next few years should be used to address the work that still needs to be done to meet Marvin's objectives. They were to assure that policy would act to keep inflation expectations anchored at 2 percent under a wide variety of circumstances and that the Fed would gain the full benefits of transparency for policy effectiveness and democratic accountability. We honor Marvin's memory by trying to live up to his high standards for analysis and policy. We can best contribute to the public welfare by applying his ideas and identifying ways to narrow the inevitable gap between Marvin's standards and actual policy practice.

Cite as: Kohn, Donald. 2022. "The Federal Reserve's New Monetary Policy Framework." In Essays in Honor of Marvin Goodfriend: Economist and Central Banker, edited by Robert G. King and Alexander L. Wolman. Richmond, Va.: Federal Reserve Bank of Richmond.