The complexity resulting from intertwined uncertainties regarding model misspecification
and mismeasurement of the state of the economy defines the monetary policy landscape.
Using the euro area as laboratory this paper explores the design of robust policy guides
aiming to maintain stability in the economy while recognizing this complexity. We document
substantial output gap mismeasurement and make use of a new model data base to capture
the evolution of model specification. A simple interest rate rule is employed to interpret
ECB policy since 1999. An evaluation of alternative policy rules across 11 models of the
euro area confirms the fragility of policy analysis optimized for any specific model and shows
the merits of model averaging in policy design. Interestingly, a simple difference rule with
the same coefficients on inflation and output growth as the one used to interpret ECB policy
is quite robust as long as it responds to current outcomes of these variables.