Youth Violence and Housing Programs to Deconcentrate Poverty

描述（由申请人提供）：问题陈述：不同社区的青少年暴力犯罪和受害率差异很大，这导致了美国不同种族和社会阶层的健康差异。拟建项目目的：本项目旨在评估芝加哥市随机住房政策干预对青少年暴力犯罪和受害的影响，该政策旨在消除贫困和暴力地区的集中。使用随机分配的干预等待名单以及管理数据，使我们能够进行令人难以置信的成本效益评估，与任何随机实验一样严格，但成本只是其中的一小部分。方法和研究对象：1997年，芝加哥12年来第一次重新开放住房券计划的等待名单。共有82,607个符合条件的家庭申请，远远超过可用的代金券数量，因此申请人被随机抽签分配到等待名单上。我们关注的是10%住在公共住房的家庭。下面我们提供的初步结果表明，对这些家庭来说，代金券在社区环境中产生了相当大的变化。另一方面，候补名单的位置与基本的社会人口特征无关——也就是说，随机券的抽奖确实是随机的。我们通过与伊利诺伊州警察局（ISP）的行政逮捕记录相匹配来衡量青少年参与暴力犯罪，并使用国家死亡指数（NDI）的死亡率数据来衡量青少年暴力受害情况。住房券对青少年暴力行为结果的因果影响主要是通过比较家庭被随机分配到“好”和“坏”候补名单位置的青少年的结果来确定的。根据基线特征对青少年的影响如何变化的证据可以提供社区层面的风险和保护因素如何与青少年层面的风险和保护因素（如年龄、性别、家庭结构）以及发展因素（如先前的犯罪轨迹）相互作用的信息。有关机制的信息将来自分析代金券对个人和社区一级候选中介因素的影响。最后，我们将比较那些家庭使用住房券进行不同类型迁移的青少年的结果，以检验社区对青少年暴力的影响在特定社区属性方面是线性的还是非线性的。对预防的影响：我们项目的发现应该与大量的住房政策决定有关，这些政策决定有可能影响美国青少年暴力的集中程度和总体数量。我们的研究也可以直接帮助设计新的面向社区的健康干预措施，针对社会环境中影响青少年暴力的具体因素。本项目旨在评估芝加哥市随机住房政策干预对青少年暴力犯罪和受害的影响，该干预旨在分散贫困和暴力地区。使用随机分配的干预等待名单以及行政数据，使我们能够对一项帮助青少年转移到不那么弱势、危险的社区的干预措施进行令人难以置信的成本效益评估，这与任何随机实验一样严格，但成本只是其中的一小部分。这项研究的结果应该与大量的住房政策决定有关，这些决定有可能影响美国青少年暴力的集中程度和总体数量，也可能直接有助于设计新的面向社区的健康干预措施，针对影响青少年暴力的社会环境中的具体因素。

Description (Provided by Applicant): Statement of the Problem: Rates of youth violence perpetration and victimization vary dramatically across neighborhoods, contributing to health disparities within the U.S. across race and social class lines. Purpose of the Proposed Project: This project seeks to evaluate the effects on youth violence perpetration and victimization of a randomized housing policy intervention in the city of Chicago designed to de- concentrate areas of poverty and violence. The use of random assignment to the intervention wait list together with administrative data enables us to carry out an incredibly cost-effective evaluation that is as rigorous as any randomized experiment - but at just a small fraction of the cost. Methods and Study Population: In 1997, Chicago re-opened the housing voucher program's wait list for the first time in 12 years. A total of 82,607 eligible families applied, far in excess of the number of available vouchers, and so applicants were assigned by random lottery to a wait list. We focus here on the 10 percent of families living in public housing at baseline. Below we provide preliminary results showing that for these families, vouchers generate sizable changes in community environments. On the other hand wait-list position is unrelated to baseline socio-demographic characteristics - that is, the random voucher lottery was indeed random. We measure youth involvement in violence perpetration by matching to administrative arrest records from the Illinois State Police (ISP), and measure youth violence victimization using mortality data from the National Death Index (NDI). The causal effects of housing vouchers on youth violence outcomes are identified by essentially comparing outcomes for youth whose families are randomly assigned to "good" versus "bad" wait-list positions. Evidence on how effects vary across youth by baseline characteristics can provide information about how community-level risk and protective factors interact with youth-level risk and protective factors such as age, gender, and family structure, as well as developmental factors such as prior offending trajectory. Information about mechanisms will come from analyzing voucher impacts on individual- and community-level candidate mediating factors. Finally, outcomes will be compared for youth whose families make different types of moves with their housing vouchers to examine whether neighborhood effects on youth violence are linear or non-linear with respect to specific neighborhood attributes. Implications for Prevention: The findings from our project should be relevant to a vast set of housing policy decisions that have the potential to shape the concentration and overall volume of youth violence in America. Our study may also directly aid in the design of new community-oriented health interventions targeted at the specific factors within the social environment that influence youth violence. This project seeks to evaluate the effects on youth violence perpetration and victimization of a randomized housing policy intervention in the city of Chicago designed to de-concentrate areas of poverty and violence. The use of random assignment to the intervention wait list together with administrative data enables us to carry out an incredibly cost-effective evaluation of an intervention that helps youth move to less disadvantaged, dangerous neighborhoods that is as rigorous as any randomized experiment - but at just a small fraction of the cost. Findings from this study should be relevant to a vast set of housing policy decisions that have the potential to shape the concentration and overall volume of youth violence in America, and may also directly aid in the design of new community-oriented health interventions targeted at the specific factors within the social environment that influence youth violence.