Title: Power Enhancement Test of Slope Heterogeneity in Panel Data Models with Interactive Fixed Effects
Speaker: Jie Wei, HuaZhong University of Science and Technology
Host: Zaichao Du, Professor , RIEM
Time: 14:00-15:30, April 7, Friday
Venue: H501, Liulin Campus
Abstract: In this paper, we construct a test of the slope heterogeneity in a panel data framework with interactive fixed effects. The model is general to accommodate dynamic effects and correlations in the errors both in cross sections and time series. We start with a quadratic form test and prove its desired size properties and non-trivial power against certain types of local alternatives. However, like other classical Wald type tests, the quadratic form test suffers from low power problems when the alternative is sparse relative to N. So we combine the quadratic test with a power enhancement component to boost the power while avoid size distortion. We show that the power enhanced test strengthens the power under sparse alternatives. The Monte Carlo simulations highlight the power enhancement performance of our test as well as its non-size-distortion property.