Doctoral Dissertation Research in Political Science: How Do Voters Decide? Ethnicity and Performance in African Elections

主要研究者：Clark C. Gibson和James D. LongTitle：博士论文研究：选民如何决定？非洲公民在民主选举中选择领导人的依据是什么？非洲的选举通常被认为是高度种族化的，候选人通过煽动种族恐惧或使用赞助等选择性激励措施来吸引同种族的选民的支持。这项研究通过确定种族和表现在何时、何地、以何种方式（无论是单独的还是共同的）起作用，增进了我们对非洲人投票行为的理解。因此，这将表明政治家是追求针对自己种族群体的政策、商品和服务，还是更广泛地针对地区或国家。如果种族假设是正确的，那么同种族主义者将一次又一次地回到办公室，并将政策狭隘地针对他们的群体。如果绩效假设是正确的，表现不佳的人将被赶下台，政治家将被鼓励采取针对同种族和非种族的政策。为了了解种族和表现是否以及如何影响投票，本研究建议对肯尼亚即将于2007年12月举行的选举进行选后调查。它将由从具有全国代表性的选区随机选出的选民组成。该调查将询问有关种族、表现评价以及地方和全国选举中候选人的投票选择等问题。本研究利用新颖的个体层面调查数据，对现有研究进行了实证和方法上的改进。这项调查没有假设种族或表现在群体层面上的影响，而是允许对这些变量进行有力和直接的测试，以确定个人的投票。为地方和全国选举进行研究，可以使用在若干特征上有所不同的不同选举单位进行单独测试。它提供了比现有研究更好的绩效衡量标准，现有研究被理解为由税收产生的地方公共服务的交付。通过随机化，该调查还避免了困扰其他研究的选择偏差问题。这些数据将允许我测试关于种族和表现的独立影响的单独假设，以及两者相互作用的方式。这个项目的广泛影响对多民族国家和贫穷国家的发展具有重要影响。这项研究的结果可以为政治家、政策制定者和捐助者提供关于公民与领导人在种族、绩效、公共服务和税收等问题上的关系的关键信息。如果绩效理论是正确的，那么选举就是促使政治家承担责任的主要机制。这些政策可能反过来成为持续经济增长和发展的引擎。如果种族理论是正确的，我们可能会预料到，从长远来看，扭曲的政策偏见对进步是有害的。在地方层面尤其如此，学者和政策制定者越来越多地将地方政府视为发展的潜在催化剂。

Primary Investigators: Clark C. Gibson and James D. LongTitle: Doctoral Dissertation Research: How do Voters Decide? Ethnicity and Performance in African Elections0718782AbstractOn what bases do African citizens choose leaders in democratic elections? African elections are typically understood as highly ethnic, where candidates appeal to voters by inciting ethnic fear or using selective incentives such as patronage to attract support from co-ethnics. The study advances our understanding of African voting behavior by identifying when, where and in what ways ethnicity and performance matter, either separately or in conjunction. Consequently, this will demonstrate whether politicians pursue policies, goods, and services directed at their own ethnic group or more broadly to the district or nation. If ethnic hypotheses are correct, co-ethnics will be returned to office time and again and narrowly target policies towards their group. If performance hypotheses are correct, poor performers will be thrown out of office and politicians will be encouraged to pursue policies targeted towards co-ethnics and non-ethnics alike.In order to understand whether and how ethnicity and performance matter to voting, this study proposes a post-election survey of Kenya's upcoming December 2007 election. It will consist of randomly chosen voters from precincts in a nationally representative sample. The survey will ask questions regarding ethnicity, evaluations of performance, and vote choice for candidates in local and national elections. The study makes empirical and methodological improvements upon current studies by using novel individual level survey data. Rather than postulate the effects of ethnicity or performance at the group level, the survey allows for robust and direct tests of these variables in determining an individual's vote. Performing the study for local and national elections allows for separate tests using different electoral units that vary over a number of characteristics. It provides a better measure of performance than extant studies, understood as the delivery of local public services that result from taxation. Through randomization, the survey also avoids problems of selection bias that plague other studies. Such data will allow me to test separate hypotheses concerning the independent effects of ethnicity and performance, as well as the ways in which the two interact. The broader impact of this project has important implications for development in multi-ethnic and poor countries. The results of this study can provide politicians, policymakers and donors with critical information regarding the relationship between citizens and leaders over issues of ethnicity, performance, public services, and taxation. If performance theories are correct, elections are the prime mechanism that induces accountability on the part of politicians. These policies may in turn become the engine to sustained economic growth and development. If ethnic theories are correct, we might expect skewed policy biases that are harmful for progress over the long-run. This is especially true at the local level, where scholars and policymakers increasingly view local government as the potential catalyst for development.