Military mutinies are critical to scholars’ collective understanding of civil–military relations. This article introduces a new dataset that systematically codes mutinies across all regions and conducts an exploratory analysis of these new data. The primary contribution made here is the introduction of a new dataset that provides scholars with a sample of mutinies across region, space, and time. The new Military Mutinies and Defections Database (MMDD) codes events of military indiscipline from 1945 to 2017. This dataset uses geocoding techniques that will enable scholars to explore the spatial patterns and diffusion associated with mutinies. The second contribution is the preliminary exploration of these new data. Of note, I demonstrate that over one-third of all mutinies are violent, 6% of mutinies are associated with civilian deaths, and anocracies are more likely to experience mutinies than democratic or autocratic counterparts. MMDD provides investigators with an exciting new tool to explore dimensions of military disloyalty.

军事叛变对于学者对军民关系的集体理解至关重要。本文介绍了一个新数据集，该数据集系统地对所有地区的兵变进行了编码，并对这些新数据进行了探索性分析。这里的主要贡献是引入了一个新数据集，为学者提供了跨地区、空间和时间的兵变样本。新的军事叛变和叛变数据库 (MMDD) 对 1945 年至 2017 年间的军事违纪事件进行编码。该数据集使用地理编码技术，使学者能够探索与兵变相关的空间模式和扩散。第二个贡献是对这些新数据的初步探索。值得注意的是，我证明超过三分之一的兵变是暴力的，6% 的兵变与平民死亡有关，与民主或专制的同行相比，无政府状态更有可能发生兵变。MMDD 为调查人员提供了一种令人兴奋的新工具来探索军事不忠的维度。