We propose a procedure, which combines hierarchical clustering with a test of overidentifying restrictions for selecting valid instrumental variables (IV) from a large set of IVs. Some of these may be invalid in that they fail the exclusion restriction. We show that if the largest group of IVs is valid, our method achieves oracle properties. Unlike existing techniques, our work deals with multiple endogenous regressors, weak instruments, heterogeneous effects and near validity. In simulations our procedure outperforms the Hard Thresholding and the Confidence Interval method. The method is applied to estimating the effect of immigration on wages and the return to education.
翻译:我们提出一个程序,将等级分组与从一大批四人组中选择有效工具变量(IV)的过度识别限制测试结合起来。其中一些可能无效,因为它们没有达到排除限制。我们表明,如果最大的四人组是有效的,那么我们的方法就具有触礁特性。与现有技术不同的是,我们的工作涉及多种内生递减者、薄弱仪器、多种效应和近乎有效的问题。模拟我们的程序优于困难阻力和互信方法。这种方法用来估计移民对工资和重返教育的影响。