Instrumental variables estimation of a generalized correlated random coefficients model
Instrumental variables estimation of a generalized correlated random coefficients model
复制标题
广义相关随机系数模型的工具变量估计
DOI：
--
复制
发表时间：
2013
期刊：
影响因子：
--
通讯作者：
Alexander Torgovitsky
中科院分区：
文献类型：
--
作者：
Matthew A. Masten;Alexander Torgovitsky
我们学习身份识别？阳离子和估计的平均治疗效果的相关随机系数模型，允许？第一阶段非均质性和二元仪器。该模型还考虑到多个内生变量以及内生变量和协变量之间的相互作用。我们的身份？阳离子方法基于对从以控制函数的不同实现为条件的普通线性回归的集合获得的系数进行平均。这个身份？阳离子策略提出了一个透明的和计算简单的估计修剪平均处理效果构建核加权线性回归的平均值。我们发展了这个估计量，并建立了它的vn-相合性和渐近正态性。蒙特卡罗模拟显示优秀？在精度上与标准两阶段最小二乘估计相当的nite-sample性能。我们应用我们的研究结果来分析空气污染对房价的影响，和？和实质性的异质性？第一阶段的仪器效应以及与家庭分类一致的治疗效应的异质性。
We study identi?cation and estimation of the average treatment effect in a correlated random coefficients model that allows for ?rst stage heterogeneity and binary instruments. The model also allows for multiple endogenous variables and interactions between endogenous variables and covariates. Our identi?cation approach is based on averaging the coefficients obtained from a collection of ordinary linear regressions that condition on different realizations of a control function. This identi?cation strategy suggests a transparent and computationally straightforward estimator of a trimmed average treatment effect constructed as the average of kernel-weighted linear regres-sions. We develop this estimator and establish its vn–consistency and asymptotic normality. Monte Carlo simulations show excellent ?nite-sample performance that is comparable in precision to the standard two-stage least squares estimator. We apply our results to analyze the effect of air pollution on house prices, and ?nd substantial heterogeneity in ?rst stage instrument effects as well as heterogeneity in treatment effects that is consistent with household sorting.
影响因子：
0.8
作者：
E. Mammen;C. Rothe;M. Schienle
通讯作者：
E. Mammen;C. Rothe;M. Schienle