Replication data for: Domestic Strife and the Initiation of Violence at Home and Abroad

The idea of diversionary war, i.e., that leaders may seek to alleviate strife at home by initiating conflict abroad, is an intuitively appealing explanation in many cases. But the extensive empirical literature on the topic has generated mostly negative or mixed support for the diversionary war argument. We argue that the results have been unsupportive primarily because the literature has neglected to distinguish among different forms of domestic strife. Not all conflict at home generates conflict abroad. Further, we argue, conflict at home can motivate domestic repression as well as foreign aggression: diversionary war must be seen as a special case of a larger theory of violence. We accordingly specify the particular form of domestic strife, the contestation of political institutions, most likely to spur leaders to violence. We then statistically examine the effects of contested institutions on the use of violence at home and abroad, using panel data spanning 110 countries and 157 years. Our tests, which are cast at the level of individual states, are based on a fixed-effects duration dependent logit model. Our analysis controls for regime type and regime change, among other factors. The evidence strongly supports our argument. On the basis of this evidence, we conclude that much of the effects of regime type and regime change on the use of force is actually due to contested institutions.

转移战争（diversionary war）的核心逻辑，即国家领导人可通过发动对外冲突缓解国内矛盾，这一理论在诸多情境中具有直观的吸引力。但针对该议题的大量实证研究文献，大多仅为该理论提供了微弱或混杂的支持证据。我们认为，现有研究未能为该理论提供有效支撑的核心原因在于，学界忽略了对不同类型国内矛盾的区分。并非所有国内冲突都会引发对外冲突。进一步而言，国内冲突既可能催生对外侵略行为，也可能引发国内镇压：转移战争实则应被视作更广义暴力理论框架下的一个特殊案例。据此，我们明确了最有可能促使领导人动用暴力的国内矛盾具体类型——政治制度争端。随后，我们采用涵盖110个国家、时间跨度达157年的面板数据，实证检验了制度争端对国内外暴力行为的影响。本研究以单个国家为分析单元，基于固定效应依存时长Logit模型展开实证检验，并控制了政权类型、政权更迭等诸多变量。研究结果有力印证了我们的核心论点。基于上述证据，我们认为，过往研究中观测到的政权类型与政权更迭对武力使用的影响，实则大多源于政治制度争端这一核心因素。