The Fragility of Sensitivity Analysis: An Encompassing Perspective*
指出，在Leamer定义的敏感性分析中，将数据生成过程纳入模型类既非稳健性的必要条件也非充分条件，即使正确设定的模型有显著系数。涵盖性原则解释了这一现象，并连接了更直观的稳健性概念，澄清了模型平均和预测合并的讨论。
Abstract Robustness and fragility in Leamer's sense are defined with respect to a particular coefficient over a class of models. This paper shows that inclusion of the data generation process in that class of models is neither necessary nor sufficient for robustness. This result holds even if the properly specified model has well‐determined, statistically significant coefficients. The encompassing principle explains how this result can occur. Encompassing also provides a link to a more common‐sense notion of robustness, which is still a desirable property empirically; and encompassing clarifies recent discussion on model averaging and the pooling of forecasts.