Evan Soltas
Jul 7, 2012

A Single Mandate?

Monetary policy can be a challenging topic, and no part of it more so than interest rates. Looking at the federal funds rate, which has rested at the zero lower bound since December 2008 -- so we're pushing on four years now -- it would be easy to conclude that monetary policy has been extremely loose for some time now.

The Federal Reserve's own statements confirm this, writing of the "highly accommodative" stance of monetary policy and the "exceptional" nature of the accommodation. The Fed's own monetary hawks are looking to rein in what they see as dangerous policy actions. Conservative economists, such at The Economist's Buttonwood, are calling for the normalization -- that is, hiking -- of real interest rates. In politics, Republicans are looking to give the Fed a single mandate of price stability, saying that only then will the Fed achieve phenomenal success like the European Central Bank. (Fed Governor Bullard is actually on the record in support of a single mandate.)

Yet such conclusions and actions would be in grave error.

One of the most common confusions in monetary policy is the failure to differentiate between structurally and cyclically looser or tighter policy. The former implies a structurally higher rate of inflation over the long term with essentially no trade-off in terms of employment. Few want such a thing. The latter implies more accommodation in times of trouble -- i.e., now -- and significantly more restrictive policy in times of growth.

Why then do I dislike the idea of a single price-stability mandate for the Fed? More precisely, I would consider this question as: why do I think there should be a real component to the monetary policy target, in addition to a nominal component? (See here for the Congressional Research Service's discussion of these issues; I follow a different line of reasoning, but our conclusions are both generally against a single mandate.)

When aggregate demand falls, real variables tend to respond before nominal variables -- that is, when the economy goes into recession, we tend to see rising unemployment and falling production long before we see disinflation or deflation, particularly at the level of core prices and wages. This downward nominal rigidity of prices means in practice that a single-mandate central bank routinely puts the economy into price-output disequilibrium, at which point conventional monetary policy becomes a car without four-wheel drive in the middle of a snowstorm -- it loses traction, sending both real and nominal macroeconomic variables swerving helplessly.

That's what happened to the American economy in 2008, and it's where it remains now. Even as real output collapsed and unemployment soared, the headline PCE price index -- whose inflation the Fed targets at 2 percent year-over-year -- dropped only slightly and for a brief moment as fuel prices declined. The core PCE price index has barely budged from its trend path.Nor do I buy the arguments that a dual mandate necessarily precludes the possibility of rules-based monetary policy by introducing the need for a discretionary trade-off between real and nominal variables. My answer is pretty obvious: then specify the trade-off according to a rule, as does nominal GDP, which makes it one-to-one between inflation as measured by the GDP deflator and real GDP, or the set of "H-value" rules I developed here. Such rules make the Fed transparent, predictable, credible, and accountable without a single mandate.

Central banks should target a single variable which has both real and nominal components weighted according to the relative cost of their deviation from a trend path. The only way a price-only mandate makes sense in this framework is if monetary policy has no influence over real variables in the short run, or if there are no economic opportunity costs -- instead of my estimate of them in the trillions of constant dollars -- to high unemployment, low capacity utilization, or other real-variable distortions. Neither is true.