Replication data for: Not by the Pen Alone

Recent studies of civil war onset emphasize the need to differentiate between hard and soft power as measures of state capacity. To better discriminate between these concepts, Warren proposes a measure for soft power (the "Media Density Index''), which he shows enhances state ability to produce substantial barriers to domestic militarized challenges. In this paper, we build on Warren's findings. We provide one possible measure of hard power---the "M Score''---to contrast with MDI. We find that this measure is insignificant but argue that better measures for hard power must be found. Second, we test an important part of Warren's theory that he neglected---the idea that there are increasing returns on the negative effect of the Media Density Index on the probability of civil war outbreak. We find that this part of his theory does not hold; there is little to no evidence of this type of relationship between MDI and civil war onset. We also provide a measure of civil wars in neighboring states. We find that this variable proves to be a strong predictor for civil wars, indicating that there is reason to doubt the assumption of between-country independence made by Warren and many other scholars of civil war onset.

现有关于内战爆发的研究均强调，需区分作为国家能力衡量维度的硬实力与软实力。为更清晰地辨析这两个概念，沃伦提出了软实力的衡量指标——“媒体密度指数（Media Density Index）”，并证实该指数可提升国家构建国内军事化挑战的实质性壁垒的能力。本研究基于沃伦的研究结论展开拓展。其一，我们提出一种硬实力的可行衡量指标——“M得分（M Score）”，以与媒体密度指数形成对照。我们发现该指标的统计显著性较弱，但主张学界仍需探索更完善的硬实力衡量方式。其二，我们检验了沃伦的理论中被其忽略的重要环节，即媒体密度指数对内战爆发概率的负向影响存在边际递增效应这一观点。研究结果显示，该理论环节并不成立，几乎无证据表明媒体密度指数与内战爆发之间存在此类关联。此外，我们还构建了邻国境内内战的衡量指标。研究发现，该变量可作为内战爆发的强有力预测因子，这表明我们有理由对沃伦及其他诸多内战爆发研究学者所提出的“国家间独立性”假设提出质疑。