Dynamics and Uncertainty in Macroeconomics

在过去的几年里，宏观经济学在纳入现实的和经验相关的动力来源方面取得了进展，并在处理不确定情况下的决策方面取得了进展。拟议的研究由两个不同但相关的项目组成，这两个项目探索经济动态的各个方面和不确定性的影响。第一个项目开发了新的经验方法来具体说明和衡量与经济模型相关的不确定性，并研究不确定性对决策的影响。第二个项目开发并应用各种方法来研究自适应学习作为经济模型中放大或传播的来源的数量重要性。不确定性在经济学中无处不在，代理人和政策制定者必须不断面对这种不确定性。第一个项目展示了如何构建和测量与估计模型相关的不确定性；开发明确说明模型不确定性的估计方法；提供将这种不确定性的测量最小化的估计；并使用这些结果来设计对我们估计的不确定性具有健壮性的决策规则。该项目为处理不确定性提供了一个新的、连贯的框架，允许对备选决策规则在不确定环境中的表现进行认真的经验评估。这些方法被应用于货币政策规则的设计和金融风险的对冲。尽管商业周期波动是宏观经济学的一个中心问题，但许多领先的商业周期模型很难与观测数据的持续性和波动性相匹配。尽管现代宏观经济学在很大程度上建立在理性预期理论的基础上，但越来越多的文献研究了不完全理性的代理人随着时间的推移学习的经济。第二个项目展示了学习如何在经济冲击的传播中提供额外的持久性，以及在某些情况下学习动态如何提供经济波动的新来源。该项目量化了这些学习动力在解释和产生商业周期波动方面的重要性。该提案中的许多结果都有很强的政策含义。例如，如果学习是经济波动的一个组成部分，那么能够改善信息传递的政策可能是减少波动的关键。更直接的是，第一个项目的一个特别重点是衡量和处理货币政策的不确定性。通过分析不确定性的经验来源，本研究为政策规则的设计提供了帮助，这些政策规则将导致良好的经济绩效。

In the past few years, macroeconomics has advanced in incorporating realistic and empirically relevant sources of dynamics, and in the treatment of decision making under uncertainty. The proposed research consists of two distinct but related projects that explore aspects of economic dynamics and the effects of uncertainty. The first project develops new empirical methods to specify and measure uncertainty associated with economic models, and studies the effects of uncertainty on decisions. The second project develops and applies methods to study the quantitative importance of adaptive learning as a source of amplification or propagation in economic models. Uncertainty is pervasive in economics, and this uncertainty must be faced continually by agents and policy makers. The first project shows how to structure and measure the uncertainty associated with an estimated model; develops estimation methods that explicitly account for model uncertainty; provides estimates which minimize a measure of this uncertainty; and uses these results to design decision rules which are robust to the uncertainty that we estimate. This project provides a new, coherent framework for dealing with uncertainty, allowing for a serious empirical evaluation of the performance of alternative decision rules in an uncertain environment. These methods are applied to the design of monetary policy rules, and the hedging of financial risks. Although business cycle fluctuations are a central issue in macroeconomics, many leading business cycle models have difficulty matching the persistence and volatility of observed data. While much of modern macroeconomics is based on the theory of rational expectations, a growing literature has studied economies with less than fully rational agents who learn over time. The second project shows how learning can provide additional persistence in the propagation of economic shocks, and how in some cases learning dynamics can provide a new source of economic fluctuations. The project quantifies the importance of these learning dynamics in explaining and generating business cycle fluctuations.Broader Impacts. Many of the results in the proposal have strong policy implications.For example, if learning is an integral feature of economic fluctuations, then policies which can improve information transmission may be crucial in reducing fluctuations. More directly, a particular focus of the first project is in measuring and dealing with uncertainty in monetary policy. By analyzing the empirically relevant sources of uncertainty, this research provides assistance in designing policy rules which will lead to good economic performance.