Abstract
We propose a methodology for determining the inflation-unemployment trade-off faced by a central bank, i.e., the ability of a central bank to transform unemployment into inflation (and vice versa) via its interest rate policy. We introduce the Phillips multiplier as a statistic to non-parametrically characterize the trade-off and its dynamic nature. Inference on the Phillips multiplier is based on a simple instrumental variable regression of cumulative inflation on cumulative unemployment using monetary shocks as instruments and weak instrument robust methods. We compute the Phillips multiplier for the US and the UK and document that the trade-off went from being very large in the pre-1990 sample period to being small (but significant) post-1990. In contrast to earlier evidence of a substantial flattening of the slope of Phillips curve, we find that the decline in the trade-off is mostly due to the anchoring of inflation expectations.
