Research
Asymmetric Preferences for Output
The paper titled “Optimal Monetary Policy with Asymmetric Preferences for Output”
is published in Economics Letters (August 2012, Vol. 117, pp. 654-656). This paper extends the work of Ruge-Murcia
(2003) by recasting that model to examine asymmetric monetary policy preferences toward
output. The asymmetric loss function that faces the policy maker is one where positive
and negative deviations of output from the natural rate are not met with an equal
sense of urgency, an assumption that is made in much of the optimal monetary policy
literature. A linearized, reduced form decision rule for inflation is estimated using
maximum likelihood. Results indicate a rejection of symmetric preferences for output
gap movements. Additionally, evidence indicates that the Federal Reserve targets potential
output rather than some higher output level as would be the case in a similarly extended Barro
and Gordon (1983) type model.
Time Varying Monetary Policy
The paper titled “Using Time Varying Monetary Policy Parameters to Identify Asymmetric
Preferences” is my job market paper. One common drawback of the empirical optimal
monetary policy literature is that inflation and interest rate targets are often assumed
to be constant within the model. In many cases, they are nuisance parameters that
are often unidentified. This paper relaxes this restriction and models these targets
as time varying processes within Surico (2007)'s optimal policy model. This extension
as well as employing a systems equation estimation approach allows for identification
of all of the policy planner's deep parameters. Empirical evidence suggests a rejection
that inflation and interest targets are constant over time. Additionally, the evidence
suggests that inflationary gaps and output gaps are asymmetrically weighted by the
Fed in the full sample estimates. These findings are robust to the possibility that
monetary policy behavior has shifted over time in an extended model. The asymmetry
disappears in the post-Volcker subsample estimation, consistent with Surico (2007).
Does the Fed Follow Speed Limits?
Walsh (2003) suggests preferences for the monetary authority where rapid speed of
adjustment from monetary policy generates loss for the policy maker. This type of
loss function builds persistence into the optimality problem where traditionally there
is none. Walsh also explores a more dynamically rich problem where lagged dependent
variables are incorporated into the structural model equations. Tangentially related
to this, Harvey (2011) argues that the inclusion of lagged values of inflation in
the New-Keynesian Phillips curve is econometrically problematic. He proposes using an
unobserved trend and stochastic cycle in order to better capture the persistence in
the data that the model does not account for. The paper titled “Does the Fed Follow
Speed Limits? Evidence Using Unobserved Components Modeling” synthesizes these two
contributions. A general form “speed limit” loss function is estimated that relaxes
the imposed symmetry assumption that Walsh models and incorporates the unobserved
components al a Harvey. Results indicate the speed of output gap adjustments is not
significant to monetary policy for the Fed. On the other hand, inflation adjustments are
significant.