When a short period of stagflation is the right policy
Monetary policy acts with a lag, so how can policymakers evaluate the appropriateness of their policy? In this commentary, Narayana Kocherlakota argues that policymakers need to follow the suggestion of Jan Qvigstad, a Norwegian central banker, according to which a central bank should plan to miss both its targets but in opposite directions. A decreasing inflation coupled with unemployment that remains high in an indication of a policy that has been too tight. On the other hand:
If the Fed is doing its job right, it will push up the unemployment rate before inflation declines, and both will remain elevated until they reach their targets. In other words, in accordance with what Qvigstad recommends, the Fed should be seeking to achieve a period of stagflation.
But the question is: how much stagflation can the Fed tolerate?
Stagflation Is Just What the Economy Needs
By: Narayana Kocherlakota – University of Rochester
The long-run relationship between inflation and economic activity
In this paper, Ascari et al. investigate the nature of the long-run relationship between inflation and economic activity using both reduced-form and structural macroeconomic models. They develop a new empirical framework to investigate this relationship which generalizes the Bayesian VAR with stochastic trends to a piecewise linear case. The main result is the evidence in favor of a threshold level of trend inflation below which potential output is independent of trend inflation, and above which, instead, potential output is negatively affected by trend inflation. The threshold level of inflation is slightly below 4%, above which every percentage point increase in inflation is related to about 1% decrease in potential output per year.
The authors then estimate a Generalized New Keynesian (GNK) model which reproduces reproduces very well the evidence of the reduced-form Bayesian VAR (BVAR) model of a negative long-run relationship between inflation and output. In particular they find that:
The long-run Phillips curve (LRPC) is not vertical but negatively sloped and non-linear. In particular, it is vertical for very low levels of inflation and then it exhibits an increasingly negative slope as the long-run inflation rate increases above 3-4%. In terms of output losses, going from 2% to 4% inflation target causes an output loss of roughly about 0.65% per year. The effect is highly non-linear such that a 5% and a 6% inflation target would imply an output loss (relative to 2% target) of roughly 1.2% and 2% per year, respectively.

The Long-Run Phillips Curve is … a Curve
Authors: Guido Ascari, Paolo Bonomolo, Qazi Haque
From: De Nederlandsche Bank, The University of Adelaide