At first blush, it would seem highly inappropriate for policymakers’ policy preferences to be influenced by their individual social values, as opposed to single-mindedly focusing on the dual mandate. However, I came to believe that social values can affect policy preferences when policymakers began to talk frequently about disparities in employment, income, and wealth. For those who value the disproportionate benefits for minority communities of a hot labor market, keeping the unemployment rate below the NAIRU and tolerating the small cost in inflation would be desirable. However, the influence of social values is powerfully contained by the commitment to the dual mandate and to maintaining inflation expectations reasonably anchored at the Committee’s 2% objective. We expect that Chairman Powell, a Republican, does not share the activist social views of Yellen.
Social Values and Policymakers’ Loss Functions
We often model monetary policymakers as seeking to minimize a loss function in which losses are assigned deviations from the two mandates: full employment and price stability. Congress did not dictate the relative weights. So, it is each policymaker’s prerogative. But it is also U.S. President’s prerogative to nominate those he or she prefers. That is why U.S. Presidents usually nominate as Fed Governors those from his or her same party. I was told that I was nominated because President Clinton thought I shared his values.
Hawkish and Dovish Loss Functions
So, how would the loss function differ between Democrats (usually doves) and Republicans (usually hawks)?
Hawks are more likely to prefer a single mandate (inflation) to a dual mandate. Even with a dual mandate, they assign less importance to actively promoting full employment relative to a more single-minded commitment to price stability. Doves certainly assign a high cost on unemployment rates exceeding the NAIRU, but a lower cost (if any at all) to an unemployment rate below the NAIRU. That cost would come in the form of higher inflation, which the loss function already accounts for.
For some time, FOMC median unemployment projections have fallen below the median estimate of the NAIRU, consistent with achieving a “hot” labor market. Some policymakers argue that a hot labor market yields persistent benefits by bringing workers into the labor market who otherwise might never return, raising the longer-run participation rate. So, support for a hot labor market might be motivated by the attempt to improve labor market outcomes or by social values. Observational equivalence here.
The Constraints Imposed by the Mandate, Inflation Expectations, and Theory
It is inevitable that the weights on deviations from full employment and price stability by individual members of the Committee are affected by their social values. However, the room to do so is significantly constrained by an overriding commitment by FOMC participants to the dual mandate and to maintaining inflation expectations anchored at the inflation objective.
Support for the 2% inflation objective, for example, means that participants will not support policies that they worry would result in an increase in inflation beyond some limit–for example, above 2½%–or policies that might result in very persistent increases beyond 2%. However, doves are more likely to be committed to a symmetric inflation objective and see the objective as calling for 2% on average. As inflation has been running below 2% for some time, there is some increasing support of proposals that would likely result in inflation overshooting 2% for some time.
And the overall commitment to anchoring inflation expectations at the inflation objective is reinforced by a belief in the accelerationist theory of inflation dynamics. In that view, inflation does not simply move to a higher level and stop but keeps moving higher as long as the unemployment rate is below the NAIRU.
We expect that participants will also not support a policy expected to unanchor inflation expectations. Those who judge that inflation expectations have already eroded to the downside are more likely to support more aggressive policies that would likely lead to a temporary overshoot of 2% to push inflation expectations back to 2%.
In addition, the uncertainty about the NAIRU will give some policymakers license to test the limits of how far the unemployment rate can be pushed below participants’ estimates of the NAIRU. They will be willing to take the risk of a modest and temporary overshoot of 2% inflation, even if that is not their ex-ante intention.
The Composition of the FOMC and the Chair
Doves are more likely to let their social values influence policy. The Yellen FOMC acted collectively as doves. Now the composition is shifting toward a more neutral central tendency. We expect that Chairman Powell, a Republican, does not share the activist social views of Yellen. That means that the FOMC under Powell will be reluctant to allow the unemployment rate to fall below the NAIRU for social policy ends. But that train has left the station. Inflation will likely overshoot the 2% objective, given how low the unemployment rate is today and where we project it will go over the next year or two. Still, the FOMC will have to navigate a challenging balancing act that reflects the tradeoff between allowing a higher inflation rate versus increased risk of recession.
The relative weights and specification of the loss function influence the policy decisions of some members of the FOMC. This is easiest to identify in cases where someone accepts a greater risk of an overshoot of the 2% objective or indeed believes an overshoot is desirable. But there are limits. Commitment to the dual mandate is one. That may limit the tolerance for an overshoot of more than 2½%, for example. Given the statutory commitment, inflation expectations anchored at 2% also impose a constraint. So, while it is inevitable that social values can affect policy preferences of individual members, the constraints ensure that the Committee as a whole is committed to policies that they believe are compatible with achieving 2% inflation on average.