Time-varying inflation targeting after the nineties

This paper provides empirical evidence on monetary policy making for the US, the UK and the EMU using the decomposition in persistent and transitory monetary shocks proposed in Andolfatto, Hendry, and Moran [Journal of Monetary Economics 55 (2008) 406-422]. We use the particle filter to overcome the non-optimality of the Kalman filter that arises as a consequence of the nonlinear dynamics for the time evolution of monetary shocks. This estimating procedure allows us to estimate all the parameters involved in the monetary policy, providing an alternative way to calibration. We present empirical evidence for the US, the UK and the EMU to show the potential applicability of our estimation method. Our findings show that the evidence of a regime change in US monetary policy making from 1997 to 2001 is weak. However, September eleven and the recession that started in the second quarter of 2001 emerge as events that probably led to regime shifts in US monetary policy making. Also the mortgage subprime crisis is another key event affecting the central bank's decisions worldwide. We show that the use of a Taylor rule with time-varying responses in accordance with a Markov-switching setting leads to empirical findings consistent with those obtained with the particle filter.