Extant research has used the COVID-19 pandemic as a context to test the “women leadership advantage during crisis” hypothesis. An influential paper reported that women U.S. governors were associated with fewer COVID-19 deaths. Building on this work, we demonstrate that methodological assumptions play a critical role in our interpretation of findings. First, we conduct a literal replication (Study 1) of the original study to validate our dataset. Second, a series of constructive replications (Studies 2A-D) shows the results rely on methodological assumptions that are not fully supported. Without these assumptions, we find no evidence for the “women leadership advantage during crisis” hypothesis. Third, in two constructive replications focusing on U.S. counties and Brazilian municipalities, we causally test the relationship between strategic leader gender and COVID-19 deaths using a geographic matching design (Study 3A) and a regression discontinuity design (Study 3B). Again, we find no evidence for the “women leadership advantage during crisis” hypothesis. Collectively, we demonstrate that when following the methodological precedent of extant research, we were able to replicate previously identified relationships between gender and leadership outcomes, but after accounting for endogeneity and basic assumptions of linear models, we were no longer able to replicate these effects. In all our constructive replications, we found no significant difference in the effectiveness of women and men strategic leaders in crises.

中文翻译：

---

在危机期间，女性战略领导者是否比男性战略领导者更有效？COVID-19 危机期间战略领导者性别与结果之间关系的因果分析


现存的研究以 COVID-19 大流行为背景来检验“危机期间女性领导优势”假说。一篇有影响力的论文报道称，美国女性州长与 COVID-19 死亡人数较少有关。在这项工作的基础上，我们证明了方法论假设在我们解释结果中起着关键作用。首先，我们对原始研究进行文字复制（研究 1）以验证我们的数据集。其次，一系列建设性的重复（研究 2A-D）表明，结果依赖于不完全支持的方法论假设。如果没有这些假设，我们就找不到“危机期间女性领导优势”假说的证据。第三，在以美国县和巴西市为重点的两次建设性重复中，我们使用地理匹配设计（研究 3A）和回归不连续性设计（研究 3B）因果测试战略领导者性别与 COVID-19 死亡之间的关系。同样，我们没有找到“危机期间女性领导力优势”假说的证据。总的来说，我们证明，当遵循现有研究的方法论先例时，我们能够复制先前确定的性别和领导结果之间的关系，但在考虑了线性模型的内生性和基本假设后，我们不再能够复制这些影响。在我们所有的建设性重复中，我们发现女性和男性战略领导者在危机中的有效性没有显著差异。