Mechanisms for Achieving Monetary Stability: Inflation Targeting Versus the ERM

Canzoneri (1985) and Rogoff (1985) showed that imposing symmetric penalties on a central bank for deviations from an announced inflation target would reduce the Barro-Gordon inflation bias, but only at the expense of distorting the central bank’s stabilization effort. More recently, Walsh (1995) and Persson and Tabellini (1993) showed that a simple linear penalty on all observed inflation would eliminate the inflation bias without distorting the stabilization effort. We view this result as a methodological criticism of the Barro-Gordon model, and not a statement about the ease with which the bias can be eliminated. Thus, we begin by modifying the Barro-Gordon model to reintroduce the trade-off between inflation credibility and stabilization, even when linear inflation penalties are imposed. We then discuss how either type of penalty might actually be imposed, and ask whether inflation targeting (supported by one or the other type of penalty) is likely to bring a better resolution to the credibility-stabilization trade-off than the ERM.