President's Speeches

Discussion of: “Should the FOMC Have a Ternary Mandate?”

Narayana Kocherlakota | President
Federal Reserve Bank of Minneapolis

Macroprudential Monetary Policy Conference, Federal Reserve Bank of Boston
Boston, MA
October 2, 2015

Narayana Kocherlakota


I want to thank the organizers, and President Eric Rosengren in particular, for inviting me to be part of this conference and for inviting me to discuss this very interesting paper.

The paper’s title poses a normative question: “Should the FOMC have a ternary mandate?” I thought that this was a very interesting question, and it’s what I plan to address in the bulk of my remarks. The paper itself addresses a different question, “Does the FOMC act as if it has a ternary mandate?” I’ll return to this also interesting question at the very end of my discussion.

My basic premise is that, whatever the benefits of a mandate, it is undesirable for that mandate to reduce the level of clarity among policymakers and the public about monetary policy choices and outcomes. Along those lines, I will argue that a mandate is likely to add to public uncertainty about the course of policy and the economy unless:

  • It can be formulated in terms of a long-run quantitative economic objective.
  • There is clarity about symmetry: that is, how the FOMC views positive, as opposed to negative, deviations from the objective.
  • It is clear over what time frame the FOMC should seek to eliminate undesirable transitory deviations from the long-run objective.

I’ll illustrate the usefulness of these three criteria by applying them to the existing two mandates. I’ll then use them to argue that, whatever the benefits of a ternary mandate, such a mandate would have substantial costs in terms of adding to policy and economic uncertainty. I’ll wrap by actually saying a few words about the paper!

All of the views that I express today are my own, and are not necessarily those of others in the Federal Reserve System. As well, I want to be clear that I view the decision about whether to adopt a financial stability mandate for monetary policy as being entirely the province of Congress, not the Federal Reserve. I’m merely offering some thoughts on the costs of such a mandate for the consideration of interested members of the public.

Price stability mandate

Congress has mandated that the FOMC should make policy so as to promote price stability. Over the past few years, the FOMC has taken significant steps to make this verbal mandate a more useful foundation for the making of monetary policy.

In January 2012, the FOMC adopted a description of its long-run goals and strategies that I’ll call the “framework statement.” It translates the words “price stability” into a quantitative goal of a 2 percent annual inflation rate. Here, the term “inflation rate” refers specifically to the personal consumption expenditures (or PCE) inflation rate.

Last year, the FOMC offered further clarity about the price stability mandate by explaining how it views deviations from its inflation goal. The minutes from the October 2014 meeting say that “there was widespread agreement that inflation moderately above the Committee’s 2 percent goal and inflation the same amount below that level were equally costly.” Put briefly, the FOMC views positive and negative deviations from the inflation target symmetrically. I spoke about the benefits of symmetry in a speech last fall,1 and so I was glad to see the FOMC achieve consensus on this important issue.

So, the FOMC now has a quantitative 2 percent goal for annual PCE inflation. It has clarified for itself, and for the public, that this goal is a symmetric one. These are both valuable steps in strengthening the Committee’s making of, and communication about, monetary policy.

However, I do see an area for improvement associated with the FOMC’s current formulation of the price stability mandate: There is no specified time frame for achieving the 2 percent objective. In my view, the FOMC should consider articulating a benchmark two-year time horizon for returning inflation to the 2 percent goal. Two years is a good choice for a benchmark because monetary policy is generally thought to affect inflation with about a two-year lag. It mimics the practice of other leading central banks, like the Bank of England and the Bank of Canada.

Adopting such a time frame would have two related benefits. The first is that it would help anchor longer-term inflation expectations. This kind of anchoring is, of course, always helpful—but it would be especially so right now. If we look at Treasury Inflation-Protected Securities, the current market-based expectation for inflation from 2020 to 2025 (five to 10 years out) has fallen well below 2 percent. And that’s in terms of consumer price index inflation, which generally runs about 0.3 percent higher than the Fed’s PCE inflation gauge. Market participants do not seem convinced that the FOMC intends to return inflation, in a sustainable way, to 2 percent within five years.

The second, related benefit of a time horizon is that it would reduce policy uncertainty for the public. For example, PCE inflation is currently running near zero and, as of June, was not expected by the FOMC’s own staff to return to target before the end of the decade.2 If the FOMC was fully committed to an explicit two-year time horizon, the public could be confident in the current circumstances that the Committee would ease monetary policy so as to facilitate a more rapid return of inflation to the 2 percent target.

Maximum employment mandate

Let me turn now to the employment mandate. Congress has also mandated that the FOMC should make monetary policy so as to promote maximum employment. The FOMC has taken some significant steps toward making this verbal goal operational, but some important challenges remain.

The FOMC’s framework statement provides a fixed goal of 2 percent inflation. However, it quite rightly emphasizes that monetary policy is not a prime determinant of what constitutes maximum employment in the long run. As a result, the framework statement is unable to provide a similar fixed quantitative goal for the employment mandate. Instead, every quarter, as part of the Summary of Economic Projections, FOMC participants announce their forecasts for the long-run unemployment rate, under appropriate monetary policy. These forecasts can be taken as being quantitative measures of the FOMC participants’ long-run goals for the labor market.

These long-run forecasts can and do change greatly over time. And, relatedly, every participant’s forecast has a large amount of (unreported) uncertainty associated with it. I see this kind of variability and uncertainty as being endemic to the maximum employment mandate (or any other mandate couched in terms of real, as opposed to nominal, variables). In my view, this uncertainty about long-run labor market objectives translates into public uncertainty about the FOMC’s likely response to prevailing economic conditions.

Let me turn next to how the Committee views deviations relative to this objective. Earlier, I noted that I was glad that FOMC participants had largely agreed that they viewed positive and negative inflation deviations, relative to the 2 percent goal, symmetrically. But I don’t see symmetry as being appropriate with respect to the employment mandate. The mandate does not ask the FOMC to promote stable employment by seeking to eliminate both positive and negative deviations around the long-run level of employment. Rather, the mandate requires the FOMC to promote maximum employment. This language says to me, as a policymaker, that the FOMC should seek only to eliminate negative deviations of employment relative to its long-run level. Put another way: My interpretation of the employment mandate is that the FOMC should not seek to choke off the creation of jobs in the absence of any inflationary threat.

But it is important to emphasize that my perspective is not held by all of my colleagues within the FOMC. Indeed, the typical formulation of the Committee’s objective (as captured, for example, in the FOMC staff’s optimal control exercises) assumes that the FOMC is equally averse to transitory shortfalls or transitory overruns with respect to its long-run unemployment objective. Indeed, we can see this symmetry in the authors’ formulation of the central bank’s objective function: The central bank is equally averse to output’s being too high or too low relative to its long-run level.

The employment symmetry issue is quite relevant given current conditions. The unemployment rate is at or near its anticipated long-run level and is expected by many to fall further. In this situation, the appropriate stance of monetary policy will be heavily influenced by whether the Committee views unusually high levels of employment as being inappropriate. Lack of clarity on this issue consequently translates into public uncertainty about monetary policy choices and outcomes.

Finally, I’ll note that, just as is true with the price stability mandate, the FOMC has not formulated a time horizon for the employment mandate. To summarize, the FOMC has attempted to address the challenging problem of translating the employment mandate into a long-run quantifiable objective. However, it remains ambiguous whether the FOMC views positive and negative deviations from this objective symmetrically. I see this ambiguity about symmetry as being a potentially important source of policy uncertainty, especially in the current environment.

Should the FOMC have a financial stability mandate?

I’ve used the price stability and employment mandates to illustrate the value of three criteria in assessing the benefits of a monetary policy mandate:

  • Quantifiability
  • Symmetry
  • Time-boundedness

The current formulations of the two mandates don’t satisfy all three of these criteria. I have argued that these shortcomings lead directly to policy and economic uncertainty. To what extent do these kinds of concerns apply to a ternary mandate?

Unfortunately, I am skeptical about being able to formulate a financial stability mandate that would meet any of the three criteria. In terms of quantifiability, it would seem hard to formulate a metric of any kind that will be seen as a reliable gauge of long-run financial stability. To be concrete, many observers have suggested some version of a credit-to-GDP ratio. But what is an appropriate long-run credit-to-GDP ratio for the FOMC to target? As of now, there is little theory or evidence to guide this choice.

Let us suppose, though, that the FOMC were able to formulate a quantifiable target for long-run financial stability. The authors presume that the FOMC’s objective would be symmetric with respect to deviations from this target. My own presumption is that, conditional on output and inflation outcomes, the FOMC would always prefer more stable financial markets. So, I think that it would be more reasonable to model the FOMC’s objective as strictly increasing in financial stability.

But there is still a symmetry issue: Should the FOMC increase or lower short-term interest rates to increase financial stability? It seems clear that on Black Monday, the FOMC could best increase financial stability by easing policy. In contrast, some have argued that the FOMC could have increased financial stability by tightening policy in the mid-2000s.

Finally, I’ve pointed out that the FOMC has not provided a time horizon for either the price stability mandate or the maximum employment mandate. This issue will be even more vexing for a ternary mandate. There is little or no agreement about the lags associated with the impact of monetary policy on financial stability. And the lags might vary dramatically with conditions. Sometimes, the relevant lags might be seen to be a matter of days. At other times, the lags might be seen to be a matter of decades.

To sum up: It would be hard to formulate a quantitative metric of long-run financial stability. It would be hard for policymakers to know how to treat deviations from that metric. Finally, the lags associated with the influence of monetary policy on this metric are highly uncertain. These challenges mean that adding a financial stability mandate would likely generate more public uncertainty about policy choices and economic outcomes. In considering whether to add a third mandate, these potentially large costs would have to be weighed against whatever benefits might be identified.

Does the FOMC act as if it has a ternary mandate?

Let me close by returning—albeit briefly!—to the paper itself. As I noted earlier, the bulk of the paper is actually about a positive question: “Does the FOMC act as if it has a ternary mandate?” The paper uses some very nice empirical work to argue that the answer to this question is yes. More specifically, it documents that the FOMC’s decisions do seem to depend on its readings of financial stability conditions. Of course, even if the FOMC were acting according to its dual mandate, it should take account of financial stability conditions insofar as those conditions affect the baseline outlook for prices and employment. The main point of the authors’ work is that the influence of financial stability conditions on FOMC decisions transcends the impact of those conditions on the baseline outlook for inflation and unemployment. It is in this sense that the FOMC could be said to be acting as if it has a ternary mandate.

At times, the paper suggests that this behavior might, in fact, be desirable. But I find this suggestion a little troubling. Is it appropriate for a group of unelected officials to make up their own, relatively secret, objective for monetary policy? And how can this opaqueness in the formulation of policy be expected to lead to better policy outcomes?

Fortunately, in my view, there is another possible way to interpret the authors’ findings. The manifestation of financial instability is a key source of downside risk to prices and employment. Policymakers may well condition their choices on financial stability considerations so as to reduce the level of risk to their dual mandate outlook. In this case, this conditioning should not be said to represent the pursuit of a ternary mandate. Rather, it represents a different (and I would argue entirely appropriate) approach to the pursuit of the dual mandate.3 As the authors suggest on page 28, I would encourage them to determine if financial stability factors affect FOMC decisions only through the level of risk to the baseline outlook, or if the influence of stability factors transcends their impact on the level of risk.

Thanks for listening and sorry for going a little long.


* I thank Sam Schulhofer-Wohl for helpful comments.


1 See Kocherlakota (2014a).

2 See Board of Governors (2015).

3 I have shown this formally in previous remarks—see Kocherlakota (2014b, c).


Board of Governors of the Federal Reserve System. 2015. Press Release, July 24.

Kocherlakota, Narayana. 2014a. “Clarifying the Objectives of Monetary Policy.” Speech at the Annual Meeting, Helena Branch of the Federal Reserve Bank of Minneapolis, Billings, Mont., Oct. 16.

Kocherlakota, Narayana. 2014b. “Discussion of 2014 USMPF Monetary Policy Report.” U.S. Monetary Policy Forum, University of Chicago Booth School of Business, New York, Feb. 28.

Kocherlakota, Narayana. 2014c. 2014 International Research Forum on Monetary Policy, Washington, D.C., March 21.