Multi-level Model And Static Panel Data Model, A Comparative Study

This paper mainly studies the following two aspects:(1)Using the multi-level model to analyze the regularity of the interactivity between listed companies and the regional economic development. Then compare the results with the static panel data model. By using multi-level model with Chinese province data of 1997-2007,We found that Multi-level model can better identify the inter-provincial heterogeneity and heterogeneity of model parameters ,and model fitting has a good effect . The results show that: there was this"Mathew Effect"in the development of listed companies and the regional economy growth. That is, the more economically developed the region is, the more profit listed companies will gain, the faster the economic grows, and the more favorable conditions provided by regional economy for the listed companies will be. However, in areas where economy develops moderately, listed companies gain less profit, and the conditions provided by regional economy for the listed companies is limited. Combined analysis results to put some related policy recommendations about listed companies development. (2) Summarizes the similarities and differences between the multi-level model and the static panel data model when they dealing with panel data. Thus come up with this multi-level panel data model. Multi-level model is mainly used to analyze statistical data of multi-level structure, while panel data model is a widely used econometric model proposed for the panel data. Under certain condition these two models have some similarities. However, the two-level model and the static panel data model is proposed under different research backgrounds, they have essentially difference. This study has important value on the correct understanding and using two models.Multi-level model is a statistical model which used to deal with hierarchical data developed in recent years, Multi-level model has a wide range of applications in the field of economic analysis. Though analyze the relationship between listed companies and the development of province economics in China, and give a reasonable explanation for the heterogeneity of model parameters, it provides a new analytical tool for this study. To compare with multi-level model and static panel data model can provide a good reference in the correct understanding and using two models; while proposed multi-level panel data model can integrated multi-level model and panel data model, it give a new theoretical framework for panel data analysis, it has a important practical significance in the field of theoretical study and practical application.