Time Variation, Asymmetry and Threshold Effects in Malta’s Phillips Curve

Authors: William Gatt

Corresponding: William Gatt (gattw@centralbankmalta.org)

Keywords: Inflation, NAIRU, time-varying parameters, Bayesian methods, Metropolis-Hastings, Gibbs sampling

Doi: http://dx.medra.org/10.7423/XJENZA.2018.2.06

Issue: Xjenza Online Vol. 6 Iss. 2 - December 2018

Abstract:
This paper estimates a Phillips curve for Malta using data since the 1960s, using Bayesian methods to estimate a time-varying parameter model with stochastic volatility. It presents evidence that the curve has flattened over time. This implies that the link between inflation and economic activity has weakened, consistent with findings for other countries. This phenomenon is driven by downward price stickiness and threshold effects, where inflation is generally unresponsive to domestic economic conditions unless the economy is going through a strong boom. Meanwhile, this study finds an increasingly important role for import price shocks in driving inflation in the Maltese economy, owing to its increased openness and trade integration. The estimated variance of shocks to inflation was high in the 1980s, but has fallen greatly since then, rising somewhat in the run-up to the Great Recession.

Download Article: