|Topic:||Testing Instrumental Variable Validity with High-Dimensional Data and Heteroskedasticity|
|Time:||2:00 pm - 3:00 pm|
|Venue:||Zoom Meeting (please refer to seminar PDF)|
|Speaker:||Dr. Qingliang FAN|
In this paper, we propose a new test of instrumental variable (IV) validity in data-rich environments in terms of either high-dimensional IV or controls. The test works under a growing number of instruments and (or) controls which could be even larger than the sample size. Moreover, the test is shown to have desired properties of size and power under heteroskedasticity. Simulation results show the new test performs favorably compared to existing alternative tests.