Inflation Dynamics and Monetary Policy

宏观经济学的核心问题之一是短期通货膨胀动力学的性质。这个问题也是争论最激烈的问题之一，经过几十年的调查，几乎没有明确的答案。除其他事项外，商业周期的性质以及货币政策的适当行为应该是什么。 为了应对这一挑战，近年来在通货膨胀动态的理论建模方面取得了重要进展，但伴随的计量经济学分析相当有限，迄今为止的工作对现有理论提出了一些令人不安的问题。 本课题就是在这种背景下，以正在进行的通货膨胀动态研究为基础，开发并估计通货膨胀的结构性计量经济学模型，研究其对经济周期和货币政策的影响。它也建立在一些研究者现有的工作，探讨了短期通货膨胀动态的性质对最优货币政策的影响。第一阶段涉及开发和估计结构菲利普斯曲线，生存传统的假设检验。该方法的显著特点包括：（一）使用真实的边际成本，而不是按照理论所建议的产出差距计量;（二）扩展基线理论，以考虑到可能使用经验法则定价的一部分公司;（三）确定和直接估计模型的原始参数。 第二阶段是对真实的边际成本的周期性行为进行建模。它的动机是来自第一阶段的一些初步结果，这些结果表明，以真实的边际成本路径为条件的通货膨胀行为与强调前瞻性行为的现有理论大致一致（即，我们发现经验法则行为在统计上是显著的，但在数量上并不重要）。然而，这一理论似乎在描述总体活动与真实的边际成本之间的联系时出现了问题。具体来说，在一定程度上，单位劳动力成本是它们的合理特征，真实的边际成本的变动往往滞后于整个周期的产出，而不是像现有理论所认为的那样同时变动。在这种情况下，初步研究结果表明，通货紧缩的产出成本可能与真实的单位劳动成本对产出变动的缓慢调整有关，而不是通货膨胀或适应性预期的任意粘性。为此，研究了真实的边际成本的行为。 最后，将经验菲利普斯曲线纳入动态一般均衡框架（在单位劳动成本中加入适当的刚性），从实证和规范两个角度分析其对货币政策的影响。

Among the central issues in macroeconomics is the nature of short run inflation dynamics. This question is also one of the most fiercely debated, with few definitive answers available after decades of investigation. At stake, among other things, is the nature of business cycles and what should be the appropriate conduct of monetary policy. In response to this challenge, in recent years important advances have come about in the theoretical modeling of inflation dynamics, but accompanying econometric analysis has been rather limited and the work to date has raised some troubling questions about the existing theory. It is in this context that this project develops and estimates a structural econometric model of inflation, and then studies implications for business cycles and monetary policy.The proposed research builds on some work in progress on inflation dynamics. It also builds on some of the investigators' existing work that explores the implications of the nature of short run inflation dynamics for optimal monetary policy.The research has three main stages. The first stage involves developing and estimating a structural Phillips curve that survives conventional hypothesis testing. Distinctive features of the approach include: (i) the use of real marginal costs as opposed to an output gap measure, as the theory suggests; (ii) extension of the baseline theory to allow for a subset of firms that may use rule-of-thumb pricing; and (iii) identification and direct estimation of the model's primitive parameters. The second stage involves modeling the cyclical behavior of real marginal costs. It is motivated by some preliminary results from the first stage that suggest that the behavior of inflation conditional on the path of real marginal costs is roughly consistent with existing theories that stress forward looking behavior (i.e., we find that rule of thumb behavior is statistically significant but not quantitatively important). Where the theory appears to break down, however, is in characterizing the link between aggregate activity and real marginal costs. Specifically, to the extent they are reasonably characterized by unit labor costs, movements in real marginal costs tend lag output over the cycle, rather than move contemporaneously as much of the existing theory suggests. In this vein, preliminary findings suggest that the output costs of disinflation may be related to sluggish adjustment of real unit labor costs to movements in output, as opposed to arbitrary stickiness in inflation or adaptive expectations. For this reason, the behavior of real marginal costs is studied. The final stage of the project involves integrating the empirical Phillips curve within a complete dynamic general equilibrium framework (with the appropriate rigidities in unit labor costs) and then analyzing the consequences for monetary policy, both from a positive and normative standpoint.