Doctoral Dissertation Research: Shocks, Trauma, Behavioral Parameters and Subjective Expectations: The Effects of Abduction and Violence on Economic Behavior in Northern Uganda

正如《2011年世界发展报告》所指出的那样，每四个人中就有一人生活在脆弱和受冲突影响的国家，或者生活在暴力犯罪严重的国家。此外，低收入、脆弱或受冲突影响的国家没有一个实现千年发展目标。尽管冲突影响广泛，对国家和个人层面的发展结果也很重要，但经济学对其长期后果的理解有限。经济学几乎只关注有形的后果，如资产损失或营养和教育成就。在这样做的过程中，经济学忽略了这样一种可能性，即冲击的严重性不仅反映在其破坏性上，还反映在其创伤性上。与这种观点相反，大量的心理健康文献发现，创伤经历会导致持续数十年的行为改变。这些变化可以解释，例如，为什么损失相同的家庭在受到冲击后会有不同的反应，或者为什么冲击可能导致行为改变，尽管约束或机会没有明显的变化。该项目调查了创伤经历对乌干达北部冲突后行为和结果的持久影响。这项研究重新采访了以前有代表性的年轻男性样本，他们中的许多人被反叛武装绑架并经历过暴力。我们特别关注两个主要途径：经济行为参数，如风险规避、时间偏好和信任，以及未来事件的主观概率。例如，对受战争影响儿童的心理健康研究报告称，行为变化与经济行为参数（风险规避、时间偏好和信任）的变化一致，这些参数被认为可以解释经济结果的大部分异质性。因此，该项目将利用最近在实地试验中的创新来得出这些行为参数的估计，并检查它们对相应经济结果的影响。在这样做的过程中，我们加入了最近的少数经济学研究，这些研究挑战了传统的经济学观点，即行为参数是固定的，因此不受冲击的影响。这些新研究的发现是模糊的，发现了行为的变化，但不清楚经济结果的影响。这个项目将确认创伤对行为参数的影响，并检查之前关于经济结果影响的矛盾发现是由于研究的设置还是结果的选择。尽管心理健康和经济学文献对为什么行为会在冲击后发生变化存在分歧，但最终的变化应该反映在未来不确定事件的主观概率中。只有一项经济学研究研究了创伤性事件对行为和主观概率的影响。鉴于之前的研究只关注与创伤性事件相关的冲击，通过引出大范围冲击的主观概率，我们将能够看到创伤性经历是否会影响创伤性事件和非创伤性事件的主观概率。这种可能性是先前关于暴露于暴力后行为选择性改变的研究结果的另一种解释。通过超越传统的经济学冲击模型，开辟了几个新的研究途径。特别是，冲突后以及其他休克后干预措施可能需要重新考虑，因为个人可能有不同的能力来利用它们。例如，当受影响最大的人改变了时间偏好或风险规避参数时，现金转移可能就不那么有用了。同样，收入水平和资产可能不再充分反映冲突后地区的脆弱性。总之，拟议的研究解决了我们对冲击对行为影响的理解中的一个重要空白：创伤经历对行为参数、经济结果和主观概率的影响是什么？该项目通过调查绑架儿童和在乌干达北部青年男子中遭受暴力的持续影响来回答这个问题。除了帮助将心理健康和经济学文献联系起来外，该项目还为经济学、政策和持续遭受创伤的个人作出贡献。

As the 2011 World Development Report notes one in four individuals live either in fragile and conflict-affected countries or in ones with very high levels of criminal violence. Moreover, not a single Millennium Development Goal (MDG) has been achieved by a low income fragile or conflict-affected country. Despite the wide ranging influence of conflict and its importance to country and individual-level development outcomes, economics has a limited understand of its long run consequences. Economics almost exclusively focuses on tangible consequences, such as assets loss or nutritional and educational attainment. In doing so, economics ignores the possibility that the severity of a shock is not only reflected in its destruction but also in its traumatic nature. In contrast to this view, an extensive mental health literature finds that traumatic experiences can lead to behavioral changes that persist across decades. These changes may explain, for instance, why households with identical losses can have heterogeneous responses after shocks, or why shocks may lead to changed behavior despite no visible changes to constraints or opportunities.This project investigates the lasting effects of traumatic experiences on post-conflict behavior and outcomes in Northern Uganda. The research re-interviews an ex ante representative sample of young men, many of whom were abducted by rebels forces and experienced violence. In particular, we focus on two main pathways: economic behavioral parameters, such as risk aversion, time preferences and trust, and the subjective probabilities of future events. For instance, mental health studies of war-affected children report behavioral changes that are consistent with shifts in the economic behavioral parameters (risk aversion, time preferences, and trust) that are believed to explain much of the heterogeneity of economic outcomes. Consequently, the project will use recent innovations in field experiments to elicit estimates of these behavioral parameters and examine their impact on corresponding economic outcomes. In doing so, we add to the handful of recent economics studies that challenge the traditional view in economics that behavioral parameters are fixed and therefore are not affected by shocks. The findings from these new studies are ambiguous, finding changes in behavior but unclear impacts of economic outcomes. This project will both confirm the impact of trauma on behavioral parameters and examine whether the prior contradictory findings regarding the effects on economic outcomes were due the setting of the studies or the choice of outcomes.Although the mental health and economics literature differ on why behavior may change after shocks, the ultimate changes should be reflected in the subjective probabilities of future uncertain events. Only one study in economics has studied the impact of traumatic events on both behavior and subjective probabilities. Whereas the previous study only looked at shocks related to the traumatic event, by eliciting subjective probabilities on a wide range of shocks, we will be able to see whether traumatic experiences affect the subjective probabilities of both traumatic and non-traumatic events. This possibility is an alternative explanation for such findings in prior research on selective changes in behavior following exposure to violence.Several new avenues of research are opened by moving beyond traditional models of shocks in economics. In particular, post-conflict as well as other post-shock interventions may need to be reconsidered as individuals may have varying abilities to take advantage of them. For instance, cash transfers may be less useful when the most affected have altered time preferences or risk aversion parameters. Similarly, income levels and assets may no longer fully capture vulnerability in post-conflict areas. To conclude, the proposed research addresses an important gap in our understanding of the effects of shocks on behavior: what are the effects of traumatic experiences on behavioral parameters, economic outcomes and subjective probabilities? The project answers this by examining the ongoing effects of child abductions and exposure to violence among young men in Northern Uganda. In addition to helping to connect the mental health and economics literatures, the project contributes to economics, policy and the individuals for whom trauma persists.