بانک مرکزی مستقل در مواجهه با دولت های منتخب

The literature argues that the benefits of an independent central bank accrue at no cost to the real side. In this paper, we argue that the lack of correlation between monetary autonomy and output variability is due to the proactive role of fiscal policy when faced with rigid monetary objectives. Few of the attempts to measure these correlations actually allow for a changing fiscal role. Yet, when an independent authority handles monetary policy, fiscal and wage/social protection policies remain instruments in the hands of elected governments. We find that, so long as the two authorities pursue their goals independently of each other, a conflict arises that becomes stronger as preferences diverge. We also find that the establishment of a conservative central bank encourages more divergent preferences among the public (as reflected in the government that is elected). The election of more interventionist governments then makes it harder for either authority to reach its own preferred objectives, unless cooperation is possible.

The original research on the importance of coordinating monetary policies was conducted by Hamada, 1976 and Hamada, 1977 and Canzoneri and Gray (1985). Their technique was to examine the costs of noncooperative monetary policies in a multicountry setting, compared to some form of cooperation. Since then, a large literature on the advantages of central bank independence for inflation control has emerged, with clear conclusions for the design of monetary policy and monetary institutions. That literature, however, has proceeded as if we were concerned with a single-issue economy and as if monetary policy were the only policy instrument available.1 This paper returns to the coordination issue but focuses on the issue of coordination between institutions rather than between countries as in the earlier literature. We start with an analysis of noncooperative policy making when there are two (or more) policy instruments controlled by different policy authorities, when there are private forward looking expectations and the possibility of time-inconsistent policies. In our case, these instruments are fiscal and monetary policies. However, they could also represent wage policies, supply side reforms or policy influences from abroad. We then extend that analysis to examine how electoral outcomes will be affected by the conflicts that emerge between the independent policy authorities in this context and what those electoral outcomes may mean for the effectiveness of policy. We find that many of the literature's classic conclusions can be turned on their head. That happens because unrestrained competition between independent policy makers creates conflicts in which one policy has to be used (in part) to neutralise the other. The result is a loss in efficiency as everyone has less than they might. Our first contribution therefore is to show the determinants and extent of the competition between an independent monetary policy and other policies in an economy where private sector behaviour is subject to rational expectations of future developments. We find that potential policy conflicts increase with the differences in preferences between policy makers, and that policy effectiveness falls accordingly. That has consequences for the cost of an independent monetary policy, and hence for the design of the policies which remain in the government's hands. In other words, we are analysing the consequences of delegating monetary policy, not whether one would want to delegate it. Nonetheless, the motivation for delegating monetary policy is the usual one; inflation is higher without delegation because elected governments are less inflation averse than the central bank. More interesting perhaps is our second innovation that government policies will be subject to electoral outcomes and electoral preferences. We show how policy conflicts and policy ineffectiveness affect electoral outcomes, and how those electoral outcomes affect the stance of policy. In general, we find that the more “conservative” the central bank's policies, the more “liberal” or “populist” those of the government become. That of course increases the degree of conflict. Consequently, one-sided interventions designed to limit either fiscal or monetary actions, such as in the Stability Pact or in the Keynesian policies of the 1970s, are no more than a partial remedy. Independently negotiated performance agreements might be a better solution. Throughout the paper, the terms conservative and populist are used according to the definition in Rogoff (1985). Thus, a policy maker is conservative if he/she puts more weight on the inflation objective relative to the median voter, and is liberal/populist if the output objective takes precedence. For the purposes of our paper, we take independence to mean target and instrument independence as defined by Fischer (1995). Alternative schemes could include instrument independence alone, or no independence at all for one of the parties. However, it could also mean a negotiated solution in which each party is an independent actor with full instrument independence and retains a share of target independence. We take the latter case to be the counterfactual.

We have shown that an independent central bank will increase the pressure for an activist fiscal policy. If the central bank is designed to achieve price stability, governments will be left to deliver employment and growth. Our model predicts that, in a world where conservatives worry about inflation and populists about growth, the latter will have a better chance of being elected—the more so, the more conservative is the central bank. That would normally lead to conflicting policies and poor outcomes for both parties, committed solutions included. Two conclusions follow. First, lower inflation cannot be achieved at no cost to output when fiscal policy is included in the model. And, second, the same arguments that sustain time-consistent policies in a world with forward-looking expectations, will also sustain preference consistency. Dissembling governments are not a problem for an experienced, well-informed electorate. Nonetheless, an autonomous central bank does dichotomise policy in a system with elected governments. The more differentiated the preferences between governments or electorates and the central bank, the greater is the scope for conflicts between them and the more difficult it is for them to cooperate. Yet the more important it becomes that they should cooperate. Thus, independence cannot in itself guarantee policy effectiveness in a world of many policy makers because it is too easy for others to use their policies to block those of the central bank in pursuit of their own objectives. The question to be resolved is whether the system can be used to contain those pressures and still produce mutually advantageous outcomes.