Transparency, Accountability, and Corruption Displacement

尽管国际社会提供了充足的资金，但反腐败干预在很大程度上未能遏制腐败。学者和政策制定者一致认为，提高透明度和改善问责制应能减少腐败。有了腐败行为的信息，公民将惩罚腐败官员。问责机构是这一机制的关键部分，因为它们提供了公民和政治上级可以根据信息采取行动的渠道。然而，支持这些观点的证据是混杂的。腐败在不同的透明度和问责制环境中都很常见，从信息和反腐败行动渠道丰富的国家到信息和反腐败行动渠道缺乏的国家。这个项目将解决理论和结果之间分歧的几个可能的原因。利用行为经济学的研究成果，它将辨别政客们是否通过将腐败转变为不易被发现和惩罚的形式来战略性地应对反腐败干预。此外，它将把自上而下的反腐败努力（如大规模审计账户）与自下而上的针对公民的努力（通常是分开执行的）结合起来。最后，本项目的干预措施将教育公民了解政府的角色和责任，以及他们可以采取的行动方案，并告知他们政府官员的腐败情况。这个项目的干预将发生在马拉维的地方（区）政府。为了解决过去反腐败干预措施的一些不足之处，这些干预措施将结合对地区政府自上而下的定期审计和自下而上的干预，旨在让公民参与反腐败过程，并教育他们了解他们的政府和他们可以采取的行动途径。公民透明度干预将包括在当地电台广播对官员的主要调查结果，向随机选择的一组公民提供有关受调查官员的书面报告，并将选定的地区官员召集到一组酋长、公民社会代表、媒体代表和随机选择的一组公民面前讨论调查结果。会议还将教育公民了解举报腐败和要求减少腐败的途径。每个地区内腐败模式的变化将通过在透明度干预后对地区官员进行调查并将结果与2013年基线调查进行比较来衡量。这项工作为政策制定和科学研究带来了巨大的希望。捐助者和国际组织认为，腐败破坏了非洲繁荣与和平的政府。通过提供具体的行为理论和创建个人腐败的精确衡量标准，该项目将推动有关制度、政治家激励和行为的辩论。通过使用随机实验，该项目采用了社会科学家可以使用的一些最严格的设计和测试。最后，通过在测量结果的调查中抽样实际的政治家，本项目与实际从事政治活动的人一起探讨政治现象。从这个项目中获得的见解不仅可以为腐败研究提供宝贵的信息，而且可以为其实际层面提供信息。该设计采用了由官方政府人员执行的实际审计，使本研究的结果和影响与那些对反腐败感兴趣的人非常相关。项目成果将以讨论、政策文件和学术、政府和政策受众的研究文章的形式呈现给政策制定者和研究人员。该项目还将培训数十名公民——包括私人和政府官员——学习研究方法。

Despite ample funding from the international community, anti-corruption interventions have largely failed to contain corruption. Scholars and policymakers agree that increasing transparency and improving accountability should reduce corruption. Armed with information about corrupt actions, citizens will punish corrupt officials. Institutions of accountability are a key part of this mechanism, as they provide the channel through which citizens and political superiors can act on the information. Yet, the evidence to support these arguments is mixed. Corruption is common across different transparency and accountability environments, from countries rich in information and avenues for taking action against corruption to countries lacking in both. This project will address several possible reasons for this divergence between theory and results. Drawing on work in behavioral economics, it will discern whether politicians respond strategically to anti-corruption interventions by shifting their corruption to forms less vulnerable to detection and punishment. Furthermore, it will combine top-down anti-corruption efforts, such as large-scale audits of accounts, with a bottom-up effort targeting citizens, which typically are executed separately. Finally, the interventions in this project will educate citizens about the roles and responsibilities of government and the action options available to them, in addition to informing them about corruption among their government officials. The interventions of this project will occur in Malawi's local (district) government. To address some of the deficiencies of past anti-corruption interventions, the interventions will combine a regularly scheduled top-down audit of the district government with a bottom-up intervention designed to involve citizens in the anti-corruption process and educate them about their government and avenues of action available to them. The citizen transparency intervention will include broadcasting the main findings about their officials on the local radio, providing written reports about the treated officials to a randomly selected group of citizens, and calling selected district officials before a group of chiefs, civil society representatives, media representatives, and a randomly selected group of citizens to discuss the findings. The meeting will also educate citizens about the avenues of reporting corruption and demanding a reduction in corruption available to them. Shifts in corruption patterns within each district will be measured by conducting a survey of the district officials after the transparency interventions and comparing the results to the 2013 baseline survey. This work holds great promise for both policymaking and science. Donors and international organizations believe that corruption undermines prosperous and peaceful governments in Africa. By providing a specific theory of behavior and creating precise measures of individual corruption, this project will advance the debate regarding institutions, politician incentives, and behavior. By using randomized experiments, this project employs some of the most rigorous designs and tests available to social scientists. Finally, by sampling actual politicians in the surveys measuring outcomes, this project explores political phenomena with those who actually practice politics. Insights from this project can prove invaluable to inform not just the study of corruption, but also its practical dimensions. The design uses actual audits executed by official government personnel, making this study's findings and implications very relevant to those interested in combating corruption. The project findings will be presented to policymakers and researchers in the form of discussions, policy papers, and research articles to academic, government, and policy audiences. This project will also train dozens of citizens -private and government officials- in research methods.