Credit without Commitment: The Implications of the Bankruptcy Law and the Determination of Public Debt

这个项目从有限承诺的角度来看是关于信用的。这项工作的主要主题是，虽然诱使债务人履行承诺的手段有限，但发放的信贷数额相当大，违约情况普遍存在。这些主题将这个项目的两个部分结合在一起。第一部分涉及私人信贷和美国破产法的明确考虑，破产法基本上是根据债权人的要求给予保护。该提案包括发展理论，以分析一个法律制度与美国相同的环境，在这个环境中，所有的代理人（家庭、企业）都是理性的。值得注意的两个特征是：不可观察的消费者异质性（“好”和“坏”借款人）和旨在限制不良贷款危害的合同；与现有的有限承诺理论相反，这里的违约者主要是收入受到严重冲击的消费者。该理论用于定量分析政策变化，这反过来又要求模型能够同时计算观察到的信贷和违约金额。该项目的另一部分涉及公共债务，分析政府无法承诺未来政策的环境。这项研究对在宪法框架下的政策选择做出了预测。从经济政策设计的角度来看，这两个项目的答案都非常重要。在第一部分中，分析了破产法的备选设计，包括最近在美国国会讨论的一些改革，而第二部分将有助于理解历届政府之间的相互作用，从而理解宪法对预算赤字和公共债务限制的影响。在美国，《破产法》保障公民（在宪法第一条第8款中得到承认）向债权人申请保护的权利。这一权利在经济学中大多被忽视，但在美国，每年有超过100万人申请破产，因此它经常被行使。拟议的工作从理论上和数量上研究无担保消费信贷行业的环境，具有以下特点：信贷供应商以给定的利率吸收存款，并通过一系列信贷水平和相关利率向家庭提供贷款。贷款行业竞争激烈，可以自由进入，借款人有一个违约选择，在过程和后果上都类似于根据美国破产法第7章申请破产。特别重要的是对家庭信用评级的分析，当家庭采取储蓄和违约行为时，信用评级就会更新。家庭的信用评分或信用等级因评估保险风险等多种原因而越来越重要。债务管理的研究一直围绕着假设政府对未来政策作出承诺的最优概念。在缺乏这种承诺的情况下，由于最优政策不具有时间一致性，财政政策就无法得到很好的理解。特别是，一个经典模型表明，在承诺下，政府在第一阶段总是增加债务，这表明在没有承诺的情况下，债务水平不断增加。然而，目前的项目揭示了一种没有承诺的债务管理的一般机制，这意味着虽然政府债务最初会增加，但债务扩张必须有一个限制。原因是在这些情况下产生的内生和自然的非凸性。更广泛的影响在美国法律体系的框架内理解信贷是解决各种问题的关键因素，例如信贷紧缩在经济衰退中是如何运作的，以及获得信贷的不同途径如何影响人们的终身机会。此外，国会正在考虑修改《破产法》。可以用这个模型来分析现行准则的替代方案的影响，包括对人们信用评分系统如何运作的规定。对公共债务的分析可能会影响对平衡预算修正案等问题的讨论。

This project is about credit from the point of view of limited commitment. The main themes that inform the work are that while there are limited means to induce debtors to act as they promised, the amount of credit issued is quite large, and the instances of default are pervasive. These themes unite the two parts of this project. The first part deals with private credit and the explicit consideration of the U.S. Bankruptcy Code that grants protection from creditors essentially on demand. The proposal includes the development of theory to analyze an environment where the legal system is that of the U.S., and where all agents (households, firms) act rationally. Two features are noteworthy: unobservable consumer heterogeneity ("good" and "bad" borrowers) and contracts designed to limit the harm from taking on bad loans; and those who default here are mainly consumers with bad income shocks, in contrast with existing limited-commitment theory. This theory is applied to analyze policy changes quantitatively which in turn requires that the model is capable of accounting simultaneously for the observed amounts of credit and of default. The other part of the project deals with public debt, analyzing an environment where governments cannot commit to future policies. The work yields predictions over policy choices given a constitutional framework. The answers from both projects are very important from the point of view of the design of economic policy. In the first part alternative designs of the Bankruptcy Code are analyzed, including some reforms that have been recently discussed in the U.S. Congress, while the second part will help understand the interactions between consecutive governments and hence the implications of Constitutional limits on budget deficits and public debt. In the U.S. the Bankruptcy Code guarantees the right of citizens (recognized in Article I, section 8, of the Constitution) to file for protection against creditors. This right is mostly ignored in Economics, yet it is exercised very frequently as more than one million people file for bankruptcy in the U.S. each year. The proposed work studies, theoretically and quantitatively, an environment with an unsecured consumer credit industry with the following features: Credit-suppliers take deposits at a given interest rate and offer loans to households via a menu of credit levels and associated interest rates. The loan industry is competitive, with free entry, and borrowers have a default option that resembles, in process and consequence, a bankruptcy filing under Chapter 7 of the U.S. Bankruptcy Code. Of special importance is the analysis of the credit rating that households have and that is updated when households take savings and default actions. The credit score or credit rating of households is an increasingly important feature as it is used for a variety of reasons such as assessing insurance risks.The study of debt management has been centered around notions of optimality which assume the commitment of governments to future policies. In the absence of such commitment, fiscal policy is not well understood since optimal policy is not time-consistent. In particular, a canonical model shows that under commitment, the government always increases its debt in the first period, suggesting ever-increasing debt levels when there is no commitment. However, the present project is uncovering a general mechanism underlying debt management without commitment which impliesthat although government debt will tend to increase initially, there must be a limit to debtexpansion. The reason is a non-convexity that arises endogenously and naturally in these contexts.The Broader ImpactsUnderstanding credit within the framework of the U.S. legal system is a crucial ingredient for various questions such as how credit crunches operate in recessions and how differential access to credit shape people's life-long opportunities. Moreover, Congress is considering the revision of the Bankruptcy Code. The implications of the alternatives to the current Code can be analyzed with this model, including regulations of how the system of credit scores of people operate. The analysis of public debt may shape the discussions over such issues as balanced-budget amendments.