Health/Human Capital Investimates- Long-term impact: Randomized Experiments Kenya

描述（由申请人提供）：拟议的项目将在肯尼亚青年样本中扩展一套关于教育、健康、营养、人口和劳动力市场结果的显著纵向（面板）数据集。现有的数据集是从1998-2009年收集的，部分支持来自NIH/NICHD R01，拟议的项目将把它再延长6年，到2015年。由此产生的17年纵向数据集将有助于严格估计健康和人力资本干预措施对一系列生活结果的长期影响，并有助于探索“成功”过渡到成年的决定因素。特别是，该项目提议找到并重新调查肯尼亚生活小组调查（KLPS）的大约6 800名受访者，这是一个由肯尼亚青年组成的不同寻常的纵向数据集，其中许多人是早期健康和人力资本干预措施的受益者。这些干预措施包括：（一）一项从1998年开始向3万多名小学生提供驱虫药物的保健方案；（二）一项以成绩为基础的奖学金方案，支持2001-2002年招收近12 000名五年级和六年级学生的小学女生的学费；以及（三）2008年启动的一项方案，提供学费券，支持大约1 100名年轻人接受最多两年的职业教育。他们之前都是驱虫或奖学金项目样本的一部分。由于这些早期干预措施的受益人选择是随机的，因此数据收集将使我们能够估计儿童健康或教育（或两者结合）方面的外生收益如何影响广泛的成年生活结果。在过去和正在进行的研究中，这些干预措施显示对受益者有重大的中短期影响。尽管有充分的先验理由预测，更好的儿童健康和教育对劳动力市场和人口有重大的长期好处，但这些影响很少得到实证证明，这在很大程度上是由于几乎完全缺乏我们建议收集的那种扩展的纵向数据。试图在欠发达国家设计有效的卫生和教育项目的公共决策者迫切需要证据。此外，拟议的数据收集将为在贫困率、失业率、生育率和疾病负担高的环境中审查青少年“成功”成年过渡的决定因素提供一个难得的机会。在最近的KLPS数据收集期间（2007-2009年），受访者的平均年龄为22岁，只有38%的人已婚，44%的人有孩子。随着这些人从十几岁到二十岁出头进入成熟的成年期，许多人将开始重要的人生转变，包括结婚和组建家庭、找工作、创业和移民，这是一个难得的机会，可以详细记录他们未来六年的生活轨迹。由此产生的多轮KLPS数据集将跟踪中位数样本受访者从12岁到30岁。

DESCRIPTION (provided by applicant): The proposed project will extend a remarkable longitudinal (panel) data set of educational, health, nutritional, demographic and labor market outcomes among a sample of Kenyan youth. The existing data set was collected from 1998-2009 with partial support from an NIH/NICHD R01, and the proposed project will extend it for an additional six years, through 2015. The resulting 17-year longitudinal data set will allow for rigorous estimation of the long-run impact of health and human capital interventions on a range of life outcomes, and for an exploration of the determinants of "successful" transitions to adulthood. In particular, this project proposes to locate and re-survey approximately 6,800 respondents of the Kenya Life Panel Survey (KLPS), an unusual longitudinal data set comprised of Kenyan youth, many of whom were beneficiaries of earlier health and human capital interventions. These interventions include (i) a health program that provided deworming medication to more than 30,000 primary school children starting in 1998, (ii) a merit-based scholarship award program supporting school fees for female students in primary schools enrolling nearly 12,000 individuals in grades 5 and 6 in 2001-2002, and (iii) a program launched in 2008 to provide tuition vouchers to support up to two years of vocational education for roughly 1,100 young adults, all of whom had earlier been part of the deworming or scholarship program samples. Since the selection of beneficiaries for these earlier interventions was randomized, the data collection will allow us to estimate how exogenous gains in childhood health or education (or a combination of the two) affect a wide range of adult life outcomes. In previous and ongoing research, these interventions were shown to have substantial short- to medium-term impacts on beneficiaries. Despite strong a priori reasons to anticipate significant long-term labor market and demographic benefits from better childhood health and education, these effects have rarely been demonstrated empirically, in large part due to the near total absence of extended longitudinal data of the kind we propose to collect. Evidence is urgently needed by public policymakers attempting to design effective health and education programs in less developed countries. Furthermore, the proposed data collection will provide an exceptional opportunity to examine the determinants of "successful" adulthood transitions for adolescents in a setting characterized by high poverty rates, unemployment, fertility, and disease burdens. During recent KLPS data collection in 2007-2009, respondents were on average 22 years old, merely 38% were married and 44% had children. As these individuals progress from their late teens and early 20s into full-fledged adulthood, many will embark on important life transitions including marriage and family formation, job search, entrepreneurship and migration, presenting a rare opportunity to document in detail their life trajectories over the next six years. The resulting multi-round KLPS data set will follow the median sample respondent from age 12 through to age 30.