Quantile regression for panel data models with fixed effects under random censoring

Abstract

The locally weighted censored quantile regression approach is proposed for panel data models with fixed effects, which allows for random censoring. The resulting estimators are obtained by employing the fixed effects quantile regression method. The weights are selected either parametrically, semi-parametrically or non-parametrically. The large panel data asymptotics are used in an attempt to cope with the incidental parameter problem. The consistency and limiting distribution of the proposed estimator are also derived. The finite sample performance of the proposed estimators are examined via Monte Carlo simulations.

Keywords:

• Quantile regression
• panel data
• fixed effects
• random censoring
• Kaplan-Meier estimator

Additional information

Funding

The work was partially supported by the National Natural Science Foundation of China (No. 11861042), the Fundamental Research Funds for the Central Universities, and the Research Funds of Renmin University of China (No. 18XNL012), China Statistical Research Project (No. 2016LD03), the Fund of the Key Research Center of Humanities and Social Sciences in the general Colleges and Universities of Xinjiang Uygur Autonomous Region, and Shanghai Natural Science Foundation (No. 18ZR1427200). The work of Man-Lai TANG was supported by a grant from the Research Grant Council of the Hong Kong Special Administrative Region (Project No. UGC/FDS14/P01/16), a grant from National Natural Science Foundation of China (Grant No. 11871124), and the Big Data and Artificial Intelligence Group of the School of Decision Sciences at the Hang Seng University of Hong Kong.