摘要翻译：
在一类结构模型中，我们提出了一个描述反事实对关于潜在变量分布的参数假设的敏感性的框架。特别是，我们展示了如何表征反事实的最小值和最大值，因为潜在变量的分布跨越了研究者参数规范的非参数邻域，同时保持了模型的其他“结构”特征。我们的程序用有限维凸规划代替了关于分布的无限维优化，因此计算简单。我们开发了一个新的MPEC实现我们的过程，以进一步简化模型中的计算，这些模型具有由平衡约束定义的内生参数。我们的方法在大邻域上恢复了非参数识别的反事实集的尖锐界限，并与小邻域上的灵敏度分析的局部方法有联系。我们提出了最小反事实和最大反事实的插件估计器和两个推理过程。我们通过对匹配模型和动态离散选择的经验应用来说明我们的过程的广泛适用性。
---
英文标题：
《Counterfactual Sensitivity and Robustness》
---
作者：
Timothy Christensen and Benjamin Connault
---
最新提交年份：
2021
---
分类信息：

一级分类：Economics        经济学
二级分类：Econometrics        计量经济学
分类描述：Econometric Theory, Micro-Econometrics, Macro-Econometrics, Empirical Content of Economic Relations discovered via New Methods, Methodological Aspects of the Application of Statistical Inference to Economic Data.
计量经济学理论，微观计量经济学，宏观计量经济学，通过新方法发现的经济关系的实证内容，统计推论应用于经济数据的方法论方面。
--

---
英文摘要：
  We propose a framework for characterizing the sensitivity of counterfactuals with respect to parametric assumptions about the distribution of latent variables in a class of structural models. In particular, we show how to characterize the smallest and largest values of the counterfactual as the distribution of latent variables spans nonparametric neighborhoods of a researcher's parametric specification while other "structural" features of the model are maintained. Our procedure replaces the infinite-dimensional optimization with respect to the distribution by a finite-dimensional convex program and is therefore computationally simple to implement. We develop a novel MPEC implementation of our procedure to further simplify computation in models featuring endogenous parameters defined by equilibrium constraints. Our procedure recovers sharp bounds on the nonparametrically identified set of counterfactuals over large neighborhoods and has connections with local approaches to sensitivity analysis over small neighborhoods. We propose plug-in estimators of the smallest and largest counterfactuals and two procedures for inference. We illustrate the broad applicability of our procedure with empirical applications to matching models and dynamic discrete choice.
---
PDF链接：
https://arxiv.org/pdf/1904.00989