On Robustness of Average Inflation Targeting

This paper considers the performance of average inflation targeting (AIT) policy in a New Keynesian model with adaptive learning agents. Our analysis raises concerns regarding robustness AIT when agents have imperfect knowledge. In particular, target steady state can be locally unstable under if details about are not publicly available. Near low interest rates at zero lower bound, does necessarily outperform standard policy. Policymakers improve outcomes by (i) discounted inflation, or (ii) communicating data window for target.