Research

We augment a parsimonious monetary policy Bayesian VAR with heterogeneous expectations from the Michigan Survey to investigate the macroeconomic effects of shocks to the distribution of short term inflation expectations. A first surpris- ing result is that (the Michigan survey) inflation expectations do not seem to be much influenced by macroeconomic developments, while the opposite is not true. Moreover, a comprehensive density impulse response function analysis shows that it matters to take into account the whole expectation distribution. First, it matters because considering only the first and second moment of the distribution leads to an underestimation of the macroeconomic effects of expectations shocks. Second, mean and variance shocks are stagflationary, while dispersion shocks might be re- cessionary. Third, the effects are sharper when the shock mass is condensed on the tails. Specifically, left-tail perturbations account for the largest effect of expectations shocks on macroeconomic fluctuations. It follows that central bank communication should focus on the tails: reducing the noise/dispersion might be more effective than anchoring the mean.

In U.S. data, inflation and output are negatively related in the long run. A Bayesian VAR with stochastic trends generalized to be piecewise linear provides robust reduced-form evidence in favor of a threshold level of trend inflation of around 4%, below which potential output is independent of trend inflation, and above which, instead, potential output is neg- atively affected by trend inflation. Moreover, this negative relationship is quite substantial: above the threshold every percentage point increase in trend inflation is related to about 1% decrease in potential output per year. A New Keynesian model generalized to admit time-varying trend inflation and estimated via particle filtering provides theoretical founda- tions to this reduced-form evidence. The structural long-run Phillips Curve implied by the estimated New Keynesian model is not statistically different from the one implied by the reduced-form piecewise linear BVAR model.

We propose a generalization of the rational expectations framework to allow for temporarily unstable paths. Our approach introduces multiplicative sunspot shocks and it yields drifting parameters and stochastic volatility. Then, we provide an econometric strategy to estimate this generalized model on the data. The methodology allows the data to choose between different possible alternatives: determinacy, indeterminacy and temporary instability. We apply our methodology to US inflation dynamics in the `70s through the lens of a simple New Keynesian model. When temporarily unstable paths are allowed, the data unambiguously select them to explain the stagflation period in the `70s. 

Link to AER

An earlier version of this paper circulated under the title: "Rational Sunspots".

This paper uses the Kalman filter to estimate potential output as a latent process. We estimate two Dynamic Linear Models, comparing the results obtained through a traditional and a New Keynesian model. We verify that the traditional measures of output gap, even if usually applied in the estimation of the New Keynesian Phillips curve, are not consistent with the theory. We propose a New Keynesian measure that overcomes this limit. We suggest it as an alternative to the use of marginal costs.

Winner of the Angelo Costa Award for the best Italian undergraduate theses in Economics.