طراحی سیاست های پولی و انتقال عدم تقارن در EMU : آیا عدم قطعیت مهم است؟

In this paper, we consider how uncertainty affects the choice between federal monetary policy based on national and union-wide aggregate data under conditions of asymmetry in the transmission of monetary policy. We find that the uncertainty about the transmission process sustains (and, in some cases, even reinforces) the need to take into account information about national economies in the formulation of monetary policy. Also the forecasting process matters when uncertainty is additive: in particular, when union-wide forecasting is more accurate than national-based forecasting, this advantage can compensate for the welfare loss from using union-wide aggregation. There is, however, a strong case for using national information in the optimal design of common monetary policy.

The conduct of monetary policy in the European Union is complicated by within-union asymmetries both with regard to macroeconomic shocks and the transmission of monetary policy. Recent theoretical analysis (see De Grauwe, 2000, Gros and Hefeker, 2002 and Nolan, 2002) has shown that asymmetries in the transmission of monetary policy of the ECB call for design of monetary policies that takes into account national data. So, in order for monetary policies to be set optimally, it is not sufficient to use area-wide (euro) data on inflation and output gaps. Evidence also seems to support this view for the Federal Reserve System in the United States (see Meade and Sheets, 2002).1 This result obtains in models where there is no uncertainty regarding monetary policy. In this paper, we consider whether it holds when there is uncertainty about the transmission process (multiplicative uncertainty) or when the policymaker decides on monetary policy on the basis of forecasts that reflect an imperfect knowledge of shocks (additive uncertainty). Monetary-policy transmission uncertainty is an important issue in the European context. Several studies (see, for example, Dornbusch et al., 1998, Mihov, 2001 and European Central Bank, 2001) have pointed out that the degree of uncertainty in the transmission of monetary policy is important in the euro-zone, partly because of the heterogeneity in national transmission channels. The optimal design of monetary policy with transmission uncertainty has been analysed in detail in the theoretical literature. The main insight is that transmission uncertainty may call for caution on the part of the monetary authorities. Faced with this kind of uncertainty, the authorities should tend to stabilise less than when no uncertainty exists. See Brainard (1967) for the original argument, and Söderström (1999) and Peersman and Smets (1999) for an application to the European context.2 Implications of additive uncertainty for monetary policy have received little attention in the context of a monetary union. A few papers have addressed the issue of whether the national or union-wide based content of forecasts affects the stabilisation properties of monetary policy. Some empirical studies (see, e.g., Massimiliano et al., 2003) have shown that the predictive power of the forecasts made by the central bank of the monetary union is influenced by the decision to base these forecasts on national rather than (direct) union-wide information. To address the uncertainty issue, we use a model for the design of monetary policy in a monetary union with transmission asymmetries across member countries. We allow for uncertainty in the transmission process (Section 2), and introduce forecast-based monetary policies to address the case of additive uncertainty (Section 3). The robustness of our results is briefly discussed in Section 4 where we also conclude our study.

The design of monetary policies in a monetary union is particularly challenging when member countries have maintained idiosyncrasies that create asymmetries in the transmission of common shocks. A general result in the literature is that, when asymmetries in the transmission exist, the common central bank can improve the quality of monetary policy by using national information about inflation and the output gap rather than focusing only on the union-wide aggregates. The contribution of this paper consists in analysing whether this conclusion holds when uncertainty (of the additive or of the multiplicative type) is present in the design of monetary policy. We find that the answer depends on the type of uncertainty, although, in many of the cases that we analyse, the presence of uncertainty sustains the need to use national data on inflation and output gaps. Using only union-wide aggregated information about these variables is therefore likely to be suboptimal. Extensions of this study, some of which we have mentioned, would analyse more dimensions of transmission asymmetry and/or uncertainty. For example, the use of a model that allows for asymmetries in the transmission of shocks on the demand and the supply sides would be interesting30. Another extension is to model explicitly the voting behaviour underlying the national aggregation strategy: the questions that arise here are whether such a strategy can be sustained by majority voting among the national representatives in the Governing Council (see Hefeker, 2003, on this) and how accounting for uncertainty may affect the outcome (see Gerlach-Kristen, in press, for some evidence, and Bottazzi and Manasse, 2002).