Disability Risk, Disability Insurance and Life-Cycle Behavior

本提案的目标是提供一个框架，以理解美国残疾风险和社会保障残疾保险（DI）计划的特征如何在明确的生命周期设置中影响消费和劳动力供给行为。对残障保险计划的分析是一个日益增长的研究主题，但相对而言，它们仍然缺乏资金和研究。例如，失业保险（UI）是一个比个人保险研究得更多的项目，特别是宏观经济学家和劳动经济学家。2007年，残障保险支付的现金福利总额为991亿美元，是失业保险的三倍。残障保险项目也在快速发展。例如，1985年至2007年期间，享受残障保险的人口比例翻了一番，从占覆盖人口的2.4%增至4.7%。大多数研究人员现在都认为，残保计划正在以不可持续的速度增长。整个美国社会保险体系的财务可持续性可能会受到影响。最近残障补助计划规模的增长再次呼吁对该计划进行重新设计，旨在减少福利的慷慨程度，增加对合格残疾的定义的严格程度，并从总体上减少残障补助的道德风险问题的程度，即非残障人士申请福利并被纳入该计划。现有的实证文献确实发现了道德风险效应存在的令人信服的证据，这些证据大多基于简化形式分析。例如，有大量文献记录了残障保险福利对残障保险申请的积极影响，以及残障保险参与率下降和残障保险慷慨度上升之间的强烈相关性。另一方面，由于真实残疾状况的不可观察性，对该计划的任何限制都必然会使一些有资格获得保险的人得不到保险。一些现有的结构性研究试图衡量实施残保计划的福利价值。我们从这些早期的论文中学到了很多东西。这一建议将发展一个丰富、现实的生命周期理论框架，并放宽了早期文献中大多数简化的假设。它将估计模型的参数，然后使用这个理论框架来评估替代政策，权衡该计划的低效率和它提供的保险。特别是，它将评估该计划对工作和储蓄产生的抑制作用的重要性。它还计划使用这一框架来分析替代经济环境中的行为，具体来说，个人行为如何对残保计划特征的变化（如使计划变得不那么慷慨）或新计划的引入（如旨在使一些残保受益人重返工作岗位的政策）做出反应。所提出的模型试图在现实性和可行性之间取得平衡。它考虑到生命周期节省和复杂的工资动态，理由是，鉴于申请残保福利的机会成本以及健康也影响个人生产力的事实，在工资冲击是永久性的情况下，道德风险的考虑应比工资冲击是暂时性的情况更相关。拟议的偏好规范允许消费、休闲和健康状况之间的不可分割性（这意味着残疾消费的下降并不一定表明福利损失，而是一种使各州消费平稳的方法）。最后，该模型考虑了残保计划和其他社会保险计划之间的相互作用，以及劳动力市场的摩擦。模型的结构参数将使用间接推断和来自SIPP和PSID的纵向数据来确定。该模型用于评估各种计划变化对福利的影响：(a)提高残疾测试的严格程度；(b)改变DI重新评估的概率；(c)改变残障残保支付的累进性，(d)降低工作成本或增加残疾人的就业机会等。更广泛的影响：在一个连贯统一的框架内评估上述变化的福利影响的能力具有重要的政策意义。其中一些变化已被明确认为是解决过去二十年来残保人数扩大背后的道德风险问题的可能办法。这种扩张代表了该计划的直接成本，因为为其提供资金的个人减少了。鉴于残保计划是社会保障体系的一部分，其预期扩张对整个社会保障体系的长期可持续性具有重要影响。

The goal of this proposal is to provide a framework for understanding how disability risk and the features of the Social Security Disability Insurance (DI) program in the US affect consumption and labor supply behavior in an explicit life-cycle setting. Analyses of the DI program are a growing topic of research, but they are still relatively underfunded and understudied. For example, Unemployment Insurance (UI) is a much more studied program than DI, especially by macro and labor economists. This is despite the fact that in 2007 DI paid cash benefits totaling $99.1 billions, three times more than UI. The DI program is also growing very fast. For example, between 1985 and 2007 the proportion of people on DI doubled, from 2.4 of the covered population to 4.7 percent. Most researchers now agree that the DI program is growing at unsustainable rates. The financial sustainability of the whole U.S. social insurance system may be affected. The recent growth in the size of the DI program has renewed the calls for a redesign of the program aimed at reducing benefits generosity, increasing strictness in the definition of a qualifying disability, and in general reducing the extent of what is known as the moral hazard problem of DI, i.e., non-disabled individuals applying for benefits and being admitted into the program. The existing empirical literature has indeed found convincing evidence for the existence of moral hazard effects, mostly based on reduced form analyses. For example, there is a large literature documenting a positive effect of DI benefits on DI application and a strong correlation between fall in participation and rise in generosity of disability insurance. On the other hand, due to the unobservability of the true disability status, any restriction in the access to the program would necessarily deny insurance to some deserving individuals. Some of the existing structural studies have tried to measure the welfare value of having the DI program in place. A lot has been learned from these earlier papers.This proposal would develop a life-cycle theoretical framework that is rich, realistic, and relaxes most of the simplifying assumptions made in the earlier literature. It would estimate the parameters of the model and then use this theoretical framework to evaluate alternative policies weighting the inefficiency of the program against the insurance it provides. In particular, it would evaluate the importance of the disincentives to work and save generated by the program. It would also plans use this framework to analyze behavior in alternative economic environments, specifically, how individual behavior responds to changes to the features of the DI program (such as making the program less generous) or to the introduction of new ones (such as policies aimed at moving some DI beneficiaries back to work).The proposed model tries to strike a balance between realism and feasibility. It allows for life cycle saving and complex wage dynamics, on the ground that, given the opportunity cost of applying for DI benefits and the fact that health also affects individual productivity, moral hazard considerations should be more relevant when wage shocks are permanent than if they are temporary. The proposed preference specification allows for non-separability between consumption, leisure and health status (this means that a fall of consumption on disability does not necessarily indicate a welfare loss, but a way to smooth consumption across states). Finally, the model allows for interactions between the DI program and other social insurance programs, as well as labor market frictions. Structural parameters of the model will be identified using Indirect Inference and longitudinal data from SIPP and PSID. The model is used to evaluate the welfare effect of various program changes: (a) Increasing the strictness of the disability test; (b) Changing the probability of re-assessment for DI; (c) Changing the progressivity of DI payments, and (d) Reducing the costs of work or increasing job opportunities for people with disabilities, etc.Broader impact: The ability to evaluate the welfare effect of the changes above in a coherent unified framework has important policy implications. Some of these changes have been explicitly considered as possible solutions to the moral hazard problem underlying the expansion of DI rolls of the last two decades. This expansion represents a direct cost for the program because fewer individuals contribute to its financing. Given that the DI program is part of the Social Security system, its forecast expansion has important implications for long-term sustainability of the Social Security system as a whole.