Ryan Avent wonders why, with unemployment too high and inflation too low — even by the Federal Reserve’s own previously articulated standards — there is so much talk of “tapering” coming from members of the Open Market Committee (talk of slowly drawing down the Fed’s asset purchases).

Avent mentions the possibility that considerations other than inflation and employment are guiding policymakers’ decisions:  in this case, the concern that the current monetary policy stance is generating financial instability (by blowing up asset bubbles, according to the theory).  Narayana Kocherlakota, head of the Minneapolis Fed, has occasionally been associated with this view, but if his April speech is any indication, his position has much more nuance to it.

For what it’s worth, a paper by James Galbraith, Olivier Giovannoni, and Ann Russo looked back at the Fed’s behavior from 1969 to 2003 to determine what really drives changes in Fed policy.  The paper made waves by revealing an apparent partisan bias in monetary policy during election years, but the main findings were, if anything, more disturbing.  In addition to Fed policy playing a causal role in increasing inequality, the authors found an important behavioral change after 1983:

… contrary to official claims, the Federal Reserve does not target inflation or react to “inflation signals.” Rather, the Fed reacts to the very “real” signal sent by unemployment, in a way that suggests that a baseless fear of full employment is a principal force behind monetary policy. … [A]fter 1983 the Federal Reserve largely ceased reacting to inflation or high unemployment, but continued to react when unemployment fell “too low.”

Assume, for the sake of argument, that the Fed’s unconventional monetary policy does what it’s supposed to do (some economists, heterodox and otherwise, don’t accept that view, but this is about discerning the logic internal to Fed decision-making).  If the Federal Reserve is really targeting an inflation rate of 2 percent, that would suggest the need for continued easing.  This is essentially the position of James Bullard, head of the St. Louis Fed.  At the same April conference where Kocherlakota spoke, Bullard offered some cautionary notes for those who want to see more unemployment targeting from the Fed (you can listen to his speech here).  He went through a New Keynesian model (Ravenna and Walsh) whose upshot is that, in Bullard’s words, “price stability remains the policy advice even in the face of serious labor market inefficiencies.”  In other words, the best way for central bankers to fulfill the Fed’s maximum employment mandate is to . . . focus on price stability.  But oddly enough, that makes Bullard relatively “dovish” right now in terms of continuing the Fed’s asset purchasing program, because inflation is below target.

The Galbraith et al. working paper only looked at the Fed’s “reaction function” up to 2003, but if we take a general model in which Fed policy is motivated primarily by fear of full employment and we plug in the perception of an improving labor market — the perception that unemployment, while still high, is expected to start coming down, perhaps dangerously close to the 6.5 percent unemployment target — this might help “explain,” in a purely behavioral sense, some of the talk about the need to start winding down asset purchases in the near future, regardless of an inflation rate that’s too low.