A0436
Title: Estimation of dynamic quantile panel data model with interactive effects
Authors: Chaowen Zheng - University of York (United Kingdom) [presenting]
Abstract: The estimation of a dynamic quantile panel data model with unobserved interactive effects is considered. A two-step procedure is proposed. In the first step, we apply the iterative principal component analysis to estimate the unobserved common factors. In the second step, we construct an augmented model using these estimated factors and then run quantile regression to estimate the slope parameters together with the individual effects (factor loadings). To facilitate asymptotic analysis, we smooth the quantile objective function. We show that the proposed two-step estimators are consistent and asymptotically normal-distributed, though subject to asymptotic bias due to the incidental parameter problem. The split-panel jackknife is then employed for correcting the bias. Monte Carlo simulations confirm that our proposed bias-corrected estimator has good finite sample performance.
