Analysis And Application Of Panel Data Model

With the development of international economic and the improvement of information technology, the data form change from single time series data and cross-sectional data into panel data gradually. Panel data model becomes one of the most important research area of theory and application of non-classical econometric research.This paper is on the study of content includes three parts, which is the research of the non-stationary panel data model, the research of GARCH model in panel data, the research of dynamic panel data model.Firstly, this paper explains the system of the research methods in panel data model after a systematic classification and carding of the documents of panel data model, and discusses the estimate methods for fixed-effects model and random-effects model in panel data. Also, an applied study of regional disparity of grain production in our country is analyzed.Secondly, have explained theory of non-stationary panel data model, this content includes limit theory, unit roots test and cointegration test. Besides, the cointegration relationship of panel model between economic growth and consumption in China is considered in this paper.Thirdly, this paper discusses GARCH model in panel data, and proposes variable intercept panel GARCH model with random effects, then gives hypothesis test and parameter estimate. Volatility of FDI in China is analyzed using REVI-panel GARCH model, empirical results indicate that the model outperforms variable intercept panel GARCH model with fixed effect.Finally, random-effects dynamic model and fixed-effects AR model in panel data are explained, meanwhile this paper summarizes methods for model specification,hypothesis test and model estimation.This paper summarizes relevance literature, and gives theory extension and applied study, displays that panel data model can explain economic problems more exactly.