طراحی بانک مرکزی با عوامل ناهمگن

We study alternative institutional arrangements for the determination of monetary policy in a general equilibrium model with heterogeneous agents, where monetary policy has redistributive effects. Inflation is determined by a policy board using either simple-majority voting, supermajority voting, or bargaining. We compare the equilibrium inflation rates to the first-best allocation.

This paper studies alternative decision-making models for the determination of monetary policy. We consider a general equilibrium model with heterogeneous consumers. Differences in preferences yield diverse inflation aversions. Monetary policy is decided within a policy board that represents the agents’ preferences. This allows us to compare the outcome of the following central bank designs: first, we analyze simple-majority voting and supermajority voting, i.e. majority voting with veto power for the minority. Then, we analyze a policy board, where the representatives of each group of agents bargain over the money growth rate with and without allowing for lump-sum transfers.1 Our framework builds on the representative agent model of Lagos and Wright (2005). Their model is useful because it allows us to introduce heterogeneous preferences for consumption while still keeping the distribution of money balances tracktable. Agents have either high or a low utility from consumption. The main consequence is that there is a two-point distribution of money holdings and, therefore, monetary policy has redistributive effects as the inflation tax affects agents differently.2 The following results emerge from the model. First, the social planner's desired inflation rate is the Friedman Rule, i.e. an inflation rate at the rate of time preferences. This efficient outcome is attained if, under simple-majority voting, the agents with low-inflation preferences have the majority. Under bargaining, first-best can only be attained when transfers are feasible. Second, under all other central bank designs the equilibrium outcome does not attain the first-best allocation. In particular, under some simple-majority voting, when the agents with high inflation preferences have a majority, the resulting inflation is strictly above the Friedman Rule. The same is true for bargaining and for supermajority voting. An interesting aspect of our findings is that when we study two separate economies, one populated by agents with low preferences for consumptions and the other populated by agents with high preferences for consumptions, each prefers the Friedman Rule. In particular for any central bank design the outcome will be the Friedman Rule in each economy. If these two economies form a monetary union, then a deviation of the Friedman Rule can be the outcome. This is due to the redistributional effect of inflation in an economy with heterogeneous agents.3 Bullard and Waller (2004) have the most closely related analysis. They discuss the advantages of various alternative decision-making models in an overlapping-generations model. First, they find that if the inflation-averse agents have a minority, the resulting inflation is infinite. In contrast, under the same institutional rule, the equilibrium inflation rate in our model is above the Friedman Rule, but it is finite. This major difference is due to their use of the overlapping-generations model where the young generation prefers a hyperinflation. Second, in their model, a constitutional rule (i.e. supermajority voting) implements the first-best allocation. In our model, however, supermajority voting never yields the first-best allocation. As they remark (p. 112), “the problems with majority voting and bargaining are remedied by giving the older, minority generation a veto over proposed policy changes [using supermajority voting]. This acts as a form of commitment, causing the young to choose monetary policy based on lifetime utility, and thus to create a stationary equilibrium at the social optimum”. We propose a different solution to attain the first-best allocation: bargaining with transfers. Appropriate transfers allow the economy to move to the Friedman Rule from any status quo inflation rate since the agents who lose from the change of monetary policy are compensated. If the transfers are lump-sum, they yield a Pareto superior allocation. One interpretation of our result is that monetary policy needs to be linked to fiscal policy since this allows for such transfers. We organize the paper as follows. In Section 2, we present the environment. Section 3 analyzes monetary equilibria and the optimality of the Friedman Rule. Section 4 then examines alternative institutional arrangements for the determination of monetary policy and discusses the results. Extending our model in Section 5 allows us to interpret our results similar to those of Erosa and Ventura (2002). Finally, we provide a brief conclusion.

We consider a micro-founded model of money where monetary policy has redistributive effects. In this framework we analyze the outcome of various central bank designs. We find that the only central bank design that consistently yields the first-best allocation is Nash bargaining with transfers. This result provides an interesting link between monetary policy and fiscal policy. It states that when different groups in society have different inflation goals, the best monetary policy can nevertheless be implemented. It requires, however, that fiscal and monetary policy are jointly determined. That is, the policy board implements the first-best inflation rate and at the same time fiscal policy is used to compensate the party that loses from this new monetary arrangement. However, this finding requires some caution. In our analysis we have completely abstracted from asymmetric information problems and public choice issues. Incorporating such considerations into the analysis may change our results. These are interesting questions for future research.