Estimation and Assessment of a Dynamic Spatial Fixed Effect Model -Based on Exogenous Initial Value
This paper proposes and establishes a dynamic spatial fixed effect (DSFE) model. It contributes to the literature on three main fronts. First, we simultaneously introduce both temporally and spatially lagged factors and design a DSFE model with both dynamic and spatial fixed effects. Second, we analyze both observable and unobservable spatial effects by taking the initial value as exogenous, respectively. Third, we derive and prove the asymptotic properties and distributions of the likelihood estimators, and undertake an assessment based on a Monte Carlo simulation. The simulation results show that the likelihood estimators of the model parameters improve as the sample size increases and move closer to their real values. Moreover, the degree to which the estimation results improve depends more on the temporal dimension than on the spatial dimension. Therefore, increasing the temporal dimension can significantly improve the estimation results given a fixed spatial dimension.
dynamic spatial model fixed effect model quasi maximum likelihood estimators monte carlo simulation
GUO Penghui QIAN Zhengming
School of Economics, Xiamen University, P.R.China, 361005
国际会议
威海
英文
614-630
2010-07-24（万方平台首次上网日期，不代表论文的发表时间）