Simulation Comparision Of Dynamic Panel Data Model Estimation Method With The Selection Criteria

In recent years, considering the dynamic panel data model with selection criteria are very important in practical applications.It will get a non-random missing data elements added to the model, you can ensure that more data can be used in the model, increasing the precision and accuracy of the model.This article introduced a more natural and rela-tively simple method which wooldrige (2013)proposed about dynamic panel models with sample selection that is two-step NLS estimates and two-step GMM estimates.These two methods are not applied this classic ideological difference, but to use The dependent variable of the initial conditions, which avoids the weak instrumental variables problem that the previous method appeared in the application.Finally, for the model with the explanatory variables and the model without explanatory variables in both cases, this paper use monte carlo method to compare models, with the same set of data were gen-erated using stochastic simulation,Respectively not consider the sample selection factors and consider sample selection factors both cases, Application of six methods to estimate and compare against the six methods for analysis...