Simula@BI seminar: Geert Mesters

Simula@BI invites Associate Professor Geert Mesters, Universitat Pompeu Fabra to give a talk within the field of econometrics.

Abstract

The narrative identification approach is often only able to identify a small number of exogenous events accurately, leading to consistent but high variance estimates of dynamic causal effects. In contrast, the time series model identification approach ---e.g., recursive, long-run, max share--- leads to low variance but possibly inconsistent estimates because of omitted variables. In this work, we introduce a new method ---innovation-powered narrative inference---, which combines the benefits of both identification approaches by using innovations from time series models to boost the power of narrative methods while preserving consistency. The method is illustrated using the narrative monetary series of Romer & Romer (1989, 2023), documenting considerable reductions in confidence intervals for the causal effects of monetary policy on inflation and output. 

This is a joint work with Regis Barnichon, San Francisco Fed.  

• Read more about the speaker
• Visit our website to read more about the research center Simula@BI