Abstract:

In this talk, we will consider estimating problems for dynamic models with short panel data. The problem with estimating the dynamic parameter of interest is that the model contains a large number of nuisance parameters, one for each individual. Heckman(1978) proposed to use maximum likelihood estimation of the dynamic parameter when the prior distribution of individual parameters is normally distributed, which is named as Heckman's estimator. However, the estimator is sensitive to the assumption of prior distributions for short panel data, which estimating results have big difference with choice of prior distributions. We propose new estimators for the dynamic parameter, theoretical properties of our estimators are derived and a simulation study shows the proposed estimators have some advantages compared to Heckman's estimator and the modified profile likelihood estimator(MPL) for fixed effects.