سیاست پولی معتبر و اطلاعات خصوصی دولت
|کد مقاله||سال انتشار||مقاله انگلیسی||ترجمه فارسی||تعداد کلمات|
|24654||2001||39 صفحه PDF||سفارش دهید||15416 کلمه|
Publisher : Elsevier - Science Direct (الزویر - ساینس دایرکت)
Journal : Journal of Economic Theory, Volume 99, Issues 1–2, July 2001, Pages 338–376
Credible and optimal monetary policies are considered in environments in which the government observes a signal that is correlated with the state of the economy. When the signal is public information it is optimal for monetary policy to be conditioned upon it. The extent to which such conditioning should occur when the signal is the private information of the government depends upon the government's incentives to misrepresent information. It is shown that in some cases the Ramsey policy is incentive compatible, in others it is not. In the latter cases, policy must be constrained to be incentive compatible. This may result in “penalty phases” along the equilibrium path following apparent “mistakes” by the policy maker; it may result in the optimal monetary policy making no use of the signal. Journal of Economic Literature Classification Numbers: C73, E52, E58.
A credible monetary policy is one such that after every possible history the government has no incentive to deviate from it and implement some publicly observable alternative policy. Credible government policies have received much attention in the literature. In particular, recent papers have reworked earlier reduced form game theoretic models of credible policy to include richer specifications of the non-government private sector.1 Methodologically, progress has been made by making explicit use of recursive equilibrium concepts. This has led to new insights and new com- putational strategies. In contrast, the incentive for a government to exploit any private information that it might have about the economy in its implementation of monetary policy has received much less attention.2
نتیجه گیری انگلیسی
This paper has considered environments in which the government has some private information about the appropriateness of different monetary policies. In general the government may have incentives to misrepresent its information. In the model developed here the strength of these incentives depends on the extent of a distortion in the product market (which can be mitigated through an unexpected monetary expansion). When the incentives to misrepresent are large optimal CIC monetary policy may involve little or no conditioning of policy on the government's private information. Histori- cally, macroeconomists have been interested in the issue of ``macro- economic rules'' versus ``macroeconomic discretion.'' Recasting this language in terms of the current model, when the government's incentives to misrepresent information are high it is given little discretion in the optimal CIC equilibrium. Here ``discretion'' refers to the discretion of the government to condition its policy on its private information without automatic consign- ment to a low continuation payoff equilibrium. When the government's incentives to misrepresent information are smaller the optimal CIC equilibrium involves some amount of discretion in the sense defined above. However, in this case, as long as there is some temptation to misrepresent, penalties must be imposed contingent on those combinations of outcomes that are more likely if misrepresentation has occurred. One way that these (optimal) penalties may be introduced is through rare, but rather drastic adverse penalty phases. This was illustrated via a simulation of an optimal Markov policy equilibrium which had these features. There are, of course, many credible equilibria. Chari et al.  used this degree of freedom to select equilibria that appeared to correspond historical episodes. In par- ticular, they focused on equilibria that were consistent with the high-inflation period of the 1970's. In the analysis described here equilibria are selected to satisfy certain optimality properties rather than to correspond to historical episodes. Nevertheless, these equilibria suggest an alternative explanation for high-inflation episodes. They might be optimal penalty phases in aeconomy with private government information. APS , in a different context, provide conditions under which optimal recursive equilibria in private information games necessarily jump between the extreme points of an equilibrium payoff set. APS refer to such equilibria as ``bangbang.'' It remains to be determined whether there are conditions under which optimal credible equilibria in private information macroeconomic policy games are necessarily bang-bang as well. As developed, the model investigates the relationship between the parameters that define the preferences and technologies of agents and the set of credible, incentive compatible monetary policies. The focus has been on the relationship between the preference parameter , and this set. These relationships are suggestive of positive policy choices that might lead to welfare improvements. For example, any policy that increased the competitiveness of the goods market would tend to reduce the extent of the incentive compatibility problem thus permitting equilibria that allow the government to make better use of its private information. Any policy that reduced the rate at which the government discounted the future would also tend to have beneficial implications. It is well known that in full information environments patient governments can obtain a larger set of credible out- comes. The same result holds in private information environments of the sort analyzed here. This has implications for the design of policy making institutions; e.g., it might rationalize the lengthy terms enjoyed by Federal Reserve governors.20 Further investigation of the relationship between the design of policy making institutions and the nature of credible, incentive compatible policy is left for future work.