A SPATIAL DYNAMIC PANEL DATA MODEL WITH BOTH TIME AND INDIVIDUAL FIXED EFFECTS
This paper establishes asymptotic properties of quasi-maximum likelihood estimators for spatial dynamic panel data with both time and individual fixed effects when the number of individuals <italic>n</italic> and the number of time periods <italic>T</italic> can be large. We propose a data transformation approach to eliminate the time effects. When <italic>n / T</italic> → 0, the estimators are <inline-graphic>null</inline-graphic> consistent and asymptotically centered normal; when <italic>n</italic> is asymptotically proportional to <italic>T</italic>, they are <inline-graphic>null</inline-graphic> consistent and asymptotically normal, but the limit distribution is not centered around 0; when <italic>n / T</italic> → ∞, the estimators are consistent with rate <italic>T</italic> and have a degenerate limit distribution. We also propose a bias correction for our estimators. When n<sup>1/3</sup> / <italic>T</italic> → 0, the correction will asymptotically eliminate the bias and yield a centered confidence interval. The estimates from the transformation approach can be consistent when <italic>n</italic> is a fixed finite number.
