制度溢出效应的地区固定时间随机模型

Spatial Fixed and Time Period Random Model of Institutional Spillover Effect

为了控制个体效应的异质性,研究模型的固定效应和随机效应,尤其是因变量之间、解释变量之间和误差项之间存在的交互效应。借鉴国内外文献常用的一些制度代理变量、控制变量和扩散变量建立空间计量分析的指标体系。通过Matlab软件和R软件程序代码适用模型的比较,发掘其中的互补之处,针对固定效应可能的不足,着重分析同时或交叉使用地区或时间的固定效应和随机效应,即地区固定时间随机模型、时间固定地区随机模型、地区时间双固定模型和嵌套模型(SARAR)4种情况。1994—2014年中国内地31个省份面板数据的分析结果证实存在显著的制度溢出效应,并且地区固定时间随机模型较好地控制空间结构,得出与以往研究不同的一些结论。

For controlling the individual heterogeneity effect, research on fixed effect model and random effect, especially the interaction effects between the dependent variables, explain variables and the error term. Drawing from the domestic and foreign literature of some commonly used system proxy, control and diffusion varia-bles to establish the index system of spatial econometric analysis. Through the comparison of the Matlab and R soft-ware code for models, exploiting complementary between them, in view of the lack of fixed effects, focuses on the analysis of cross use area or time, namely spatial fixed and time period random models, time period fixed and spa-tial random model, spatial and time period double fixed model, nested model （SARAR） . An empirical panel data analysis of 1994-2014, confirm the existence of institutional spillover effect, spatial fixed and time period random models better control of space structure, some different conclusions appears with the past.