Doctoral Dissertation Research in Political Science: Democracy and War: An Experiment

经验研究发现，民主政体比任何其他政体类型更有可能赢得它们所参与的战争。支持这一发现的因果关系仍不清楚。有人提出了几个论点来解释这种模式，包括民主国家比独裁国家更能组织和分享信息，这导致民主国家拥有战场优势。也有人认为，士兵们为民主而战。一种制度性的推理路线认为，民主领导人不会发动最艰难的战争，因为他们对权力的掌握比其他类型政权的领导人更依赖于民众的认可;输掉战争损害了民主领导人的连任机会，对独裁政权领导人的生存影响不大。这个项目研究了为什么民主国家在他们参与的战争中获胜的比例高于独裁国家。为了权衡文献中的争论，该项目提出了一个理论模型，并在实验工作中对其进行了测试。首先，研究者提出了一个国内决策和国家间冲突的模型。该模型将国家间战争的风险、国内资源的分配以及领导人积累战利品的欲望联系在一起。在该模型中，民主被定义为领导者需要获得大多数选票才能赢得资源奖金的群体，而独裁被定义为领导者可以在领导者之外只有一个参与者的支持下赢得奖金的群体。如果没有沟通，预计独裁政权赢得战争的比例将高于民主国家。然而，一个能够沟通的民主国家也有可能有效地组织起来，赢得比非民主国家或没有沟通的民主国家更高比例的冲突。这是可能的，因为如果领导人承诺投资于公共产品，民主国家的非领导人就有动机披露他们的私人信息。独裁者不能承诺投资于公共产品，因此没有从独裁者提供的私人产品中受益的成员没有动机透露他们的私人信息。第二，为了测试民主团体在面对不确定性时是否能有效地组织起来，研究者建议进行小组实验，这是第一批基于团体的冲突相关实验室实验之一。实验将在纽约大学实验社会科学中心（CESS）进行。该项目对公共政策界和军事决策者有几个更广泛的影响。尽管本项目中的模型是以州际冲突为背景的，但如何激励自下而上的信息流动这一潜在主题是广泛适用的。

Empirical research has found that democracies are more likely than any other regime type to win the wars that they enter. The causal relationships underpinning this finding remain unclear. Several arguments have been put forward to explain this pattern, including the claim that democracies can organize and share information better than autocracies can, which leads democracies to have battlefield advantages. It is contended as well that soldiers fight harder for democracies. An institutional line of reasoning posits that democratic leaders do not wage the most difficult wars because their hold on power depends more heavily on their population's approval than is the case for leaders of other types of regimes; losing wars hurts the re-election chances of democratic leaders, and matters little for the survival in power of leaders in autocracies. This project examines why democracies win a higher percentage of the wars they enter than do autocracies. To weigh contending arguments in the literature, the project proposes a theoretical model and tests it in experimental work. First, the researcher presents a model of domestic decision making and interstate conflict. The model ties together the risks involved with interstate war, the distribution of resources domestically, and the desire of leaders to accumulate spoils of office. In the model, democracies are defined as groups where the leader needs a majority of votes to win the bonus of resources, and autocracies are defined as groups where the leader can win the bonus with the support of only one player in addition to the leader. Without communication, autocracies are predicted to win a higher percentage of wars than democracies. It is also possible, however, for a democracy that can communicate to organize itself efficiently and win a higher percentage of conflicts than non-democracies or democracies without communication. This is possible because non-leaders in a democracy have the incentive to reveal their private information if the leader commits to investing in public goods. Autocrats cannot commit to investing in public goods, so members that do not benefit from private goods provided by the autocrat have no incentive to reveal their private information. Second, to test if democratic groups can organize efficiently in the face of uncertainty; the researcher proposes to carry out small-group experiments, one of the first sets of group-based conflict-related laboratory experiments. The experiment will be performed at New York University's Center for Experimental Social Science (CESS). This project has several broader implications of relevance to the public policy community and military decision makers. Whereas the model in this project is framed with interstate conflict in mind, the underlying theme of how to incentivize the flow of information from the bottom up is widely applicable.