How Transparent About its Inflation Target Should a Central Bank be? An Agent-Based Model Assessment

This paper revisits the benefits of explicitly announcing an inflation target for the con- duct of
monetary policy in the framework of an agent-based model (ABM). This framework offers a flexible
tool for modeling heterogeneity among individual agents and their bounded rationality, and to
emphasize, on this basis, the role of learning in macroeconomic dynamics. We consider that those
three features (heterogeneity, bounded rationality, and learning) are particularly relevant if one
desires to question the rationale for the monetary authorities to be transparent about the inflation
target, and to achieve credibility. Indeed, the inflation targeting’s potential role in anchoring
inflation expectations and stabilizing the inflation and the economy can be analyzed more
realistically if we do not assume a representative agent framework based on substantial rationality
in behaviors and expectations. Our results show that a dynamic loop between credibility and
success can arise, and stabilize inflation, but only in the case of a learning environment that
corresponds to a moderate degree in heterogeneity regarding the behavior and decisions of
individual agents. In a more general way, we analyze, using this ABM, different assumptions about
the nature of the economic volatility, and the degree of disclosure of the target.