انتخاب بهینه استقلال بانک مرکزی و محافظه کاری تحت عدم قطعیت

This paper examines the optimal monetary delegation arrangement of central bank independence and conservatism in the presence of uncertainty about the central bank’s output objective. We show that full independence is no longer optimal with uncertain central bank preferences, and that optimal delegation involves the choice of both independence and conservatism. Further, we find that the usual trade-off between independence and conservatism no longer necessarily holds. Consequently, high central bank independence may also require a high level of conservatism.

In the debate about the optimal institutional design of central banks, independence and conservatism are usually considered to be the most important ingredients for a stable and successful monetary policy. Building on Rogoff (1985), it is argued that delegating monetary policy to an independent and conservative central bank will improve its credibility and deliver, on average, a lower and less variable rate of inflation, albeit at the price of higher output variability.1 In the literature about monetary policy delegation, however, most papers treat central bank independence and conservatism as a joint variable. Both institutional features are usually modeled by a unique parameter representing the relative weight attributed to inflation stabilisation in the central bank’s objective function. Yet, independence and conservatism (henceforth, CBI and CBC) are two different concepts: independence refers to the central bank’s ability to determine monetary policy without facing pressures from the government, while conservatism refers to the fact that the central bank assigns a higher relative weight to inflation control than society and the government do. In this paper, we explicitly distinguish between CBI and CBC in order to examine the socially optimal choice of both variables. This is done by means of a model of endogenous monetary policy delegation where the government (the principal) chooses the central bank (agent)’s degree of independence and conservatism to minimize the society’s loss function. Up to now only few papers have allowed for a formal distinction between CBI and CBC.2 Important studies are the papers by Eijffinger and Hoeberichts, 1998, Eijffinger and Hoeberichts, 2008, Hughes Hallett and Weymark, 2005 and Weymark, 2007. These authors introduce an explicit parameter for independence in a monetary policy model with a conservative central banker. The idea of government influence on monetary policy is captured by the fact that the objective function that effectively governs monetary policy is a weighted average of the central bank and the government’s objectives. CBI is then defined as the strength of the central bank in the negotiations with the government about monetary policy. Within this framework, they show that there exists a continuum of combinations of CBI and CBC that may be socially optimal. This result implies that an optimal degree of conservatism can be found for each given level of independence (and vice versa). It hence generalizes more specific studies focusing on the optimal degree of conservatism assuming full central bank independence. As pointed out by Hughes Hallett and Weymark (2005), this result also suggests that full independence is not a necessary condition for optimality in monetary policy. There exist alternative optimal combinations of independence and conservatism that lead to an equally effective monetary policy. Moreover, the optimal combinations of CBI and CBC identify both variables as potential substitutes, suggesting the existence of a trade-off between them. Less than full independence can be compensated by appointing a more conservative central banker. In this paper, we add another dimension to the analysis of the optimal delegation arrangements for monetary policy by introducing uncertainty about the central bank’s preference parameters.3 The assumption is that the government and the private sector do not perfectly know the central bank’s preferences.4 As Faust and Svensson, 2001, Faust and Svensson, 2002, Jensen, 2002 and Demertzis and Hughes Hallett, 2007, and more recently Westelius (2009), we assume that there is some uncertainty about the central bank’s output gap target.5 This uncertainty could be explained by a lack of central bank transparency, i.e. the fact that central banks are somewhat reluctant to disclose information about their policy objectives (see, for instance, Mishkin, 2004, Cukierman, 2002 and Cukierman, 2009). An alternative explanation would be the one proposed by Westelius (2009), suggesting that the central bank’s uncertain output gap target reflects its measurement errors of the potential output level. A series of papers has investigated the reaction of the public (the private sector and the government) to uncertainty in monetary policy.6 Yet, the focus here is more normative. Closest in spirit to our analysis are papers by Beetsma and Jensen, 1998 and Muscatelli, 1999 that examine the implications of uncertain central bank preferences for the optimal design of monetary institutions. They demonstrate in particular that high conservatism may be desirable in the presence of monetary uncertainty as it helps to reduce the volatility of central bank decisions. These studies only consider the extreme case of full central bank independence however. We offer new insights into the optimal delegation arrangements for monetary policy. Our first and most important contribution is to show that, in the presence of uncertainty about the central bank’s output gap target, full central bank independence is no longer necessarily optimal.7 Indeed, when the central banker’s output objective is not clearly defined, giving him full autonomy exacerbates the uncertainty surrounding his decisions and may lead to higher macroeconomic volatility. As a consequence, just focusing on the determination of the optimal degree of conservatism while assuming given extreme values for the independence parameter can lead to suboptimal arrangements. The design of monetary delegation arrangements actually involves an optimal choice of both conservatism and independence.8 The second innovation of our paper is that it qualifies the idea of a trade-off between independence and conservatism. More precisely, we show that if the uncertainty surrounding the central bank’s output gap target is high, a positive correlation may emerge between the optimal degrees of independence and conservatism, presenting them as complements rather than substitutes. In this case, rendering the central bank more independent increases the need for conservatism.

This paper has explored the specification of the optimal degrees of independence and conservatism by allowing for the possibility that there exists uncertainty about the CB’s output objective. The main message that emerges from our analysis is that exogenously setting full central bank independence while concentrating on the optimal level of conservatism does no longer lead to an optimal arrangement. This happens because when the CB’s output objective is not clearly defined, it is in the interest of society to restrain the CB’s autonomy in order to avoid excessive volatility of its decisions. Our results also indicate the existence of a positive correlation between the optimal CBC and CBI when there is high ambiguity about central bank preferences. This means that an increase in the degree of independence does no longer allow a reduction of the degree of conservatism. Instead, higher CBI amplifies the need for CBC. An interesting possible application of our study could be the context of the Economic and Monetary Union (EMU). Although the ECB’s priority for price stability is clearly articulated for all to understand, there is greater ambiguity concerning its output objective when the inflation rate is low. Our analysis suggests that under such circumstances, the ECB should not necessarily be given full independence. Moreover, the ECB is at the same time very independent and very conservative, a combination that seems to rationalize our result that independence and conservatism may be complements. From our analysis, one might argue that the ECB’s high conservatism is required due to its large degree of independence. Indeed, the delegation of the monetary instrument to a supra-national authority whose output objective is not clearly defined creates additional uncertainty in the member states. Since the member states’ national authorities are not allowed to interfere in the ECB’s decisions, it is in their interest to make sure that priority will be given to the price stability objective – which is clearly defined – and therefore to select a highly conservative central banker to avoid excessive monetary uncertainty. Finally, for the trade-off between CBC and CBI to be restored at the EMU level – so that the highly independent ECB can afford to be less conservative – our analysis would recommend more transparency and clarification concerning the ECB’s output objective.