Much of international behavior is linked spatially and temporally. Yet, analyses of interstate interactions generally either assume independence among units or resort to technical solutions to dependence that “throw away” relevant information. We detail a more informative and satisfying approach to modeling spatial dependence from extra-dyadic linkages in alliance ties and geographical proximity as specific pathways of conflict contagion. Beyond deterrence, the purpose of alliances is to draw other parties into dyadic contests, but most existing research on conflict onset generally only considers alliance ties within an individual dyad or external intervention in the same dispute. We develop new measures on third- and fourth-party alliance ties, demonstrating direct and indirect spatial effects of alliances on conflict onset. Similarly, ongoing contests can spread geographically, but dyads in some locations are much more at risk for conflict onset than others. We provide a new theory of geographic “inbetweenness” in conflict and show that dyads involving specific locations and ties to the ongoing conflict are much more likely to see dispute onset, even accounting for other purely dyadic factors. Beyond the intrinsic interest in the impacts of extra-dyadic position and alliances on conflict, our spatial approach can be applied more broadly to other extra-dyadic ties.

中文翻译：

---

国际冲突中存在偏见的关系：从联盟关系和中间性分析二元依赖的空间方法

许多国际行为在空间和时间上都相互关联。然而，对州际互动的分析通常要么假设单位之间的独立性，要么求助于“丢弃”相关信息的依赖技术解决方案。我们详细介绍了一种更具信息性和令人满意的方法来模拟来自联盟关系和地理邻近性中的二元联系的空间依赖性作为冲突传染的特定途径。除了威慑之外，联盟的目的是吸引其他各方参与二元竞争，但大多数现有的关于冲突发生的研究通常只考虑单个二元内的联盟关系或对同一争端的外部干预。我们制定了关于第三方和第四方联盟关系的新措施，展示了联盟对冲突爆发的直接和间接空间影响。相似地，正在进行的比赛可以在地理上传播，但某些地方的二元组比其他地方更容易发生冲突。我们提供了一种关于冲突中地理“中间性”的新理论，并表明涉及特定地点和与正在进行的冲突的联系的二元组更有可能发生争端，即使考虑到其他纯粹的二元组因素。除了对二元位置和联盟对冲突的影响的内在兴趣之外，我们的空间方法可以更广泛地应用于其他二元关系。甚至考虑其他纯粹的二元因素。除了对二元位置和联盟对冲突的影响的内在兴趣之外，我们的空间方法可以更广泛地应用于其他二元关系。甚至考虑其他纯粹的二元因素。除了对二元位置和联盟对冲突的影响的内在兴趣之外，我们的空间方法可以更广泛地应用于其他二元关系。