What inflation measure should a currency union target?

What is the appropriate inflation target for the monetary policymaker in a currency union, core inflation or headline inflation? We answer the question in a two-country New Keynesian model with an energy sector. We derive the welfare loss function and find that optimal monetary policy should target output gaps, the terms of trade gap, the Producer Price Index inflation rates, and the real marginal cost gaps. We use the welfare loss function to evaluate two alternative Taylor-type monetary policy rules. We find that the choice of preferred policy rule depends on the shocks. Specifically, when productivity shocks hit the economy, the monetary policymaker should follow the headline inflation Taylor rule, while the core inflation Taylor rule should be followed when a negative energy endowment shock hits the economy. In our extended model in which a representative energy producer in the home country inputs domestic non-energy consumption goods to produce energy to meet the union-wide demands, we also find that the core inflation Taylor rule should be followed when a negative energy productivity shock occurs. In addition, whichever monetary policy rule the monetary policymaker chooses, the welfare level of home households is always lower than that of foreign households following a negative energy productivity shock.