This paper extends the Common Correlated Effects Pooled (CCEP) estimator to homogeneous dynamic panels. In this setting CCEP suffers from a large bias when the time series dimension (T) is fixed. We develop a bias-corrected estimator that is valid for a multi-factor error structure provided that a sufficient number of cross-sectional averages, and lags thereof, are added to the model. We show that the resulting CCEPbc estimator is consistent as the number of cross-sections (N) tends to infinity, both for T fixed or growing large. Monte Carlo experiments show that our correction offers strong improvements in terms of bias and variance.
|Research areas and keywords
- Probability Theory and Statistics
- Factor augmented regression, multi-factor error structure, dynamic panel bias
|Place of Publication||Ghent|
|State||Published - 2016|
|Publisher||Ghent University, Faculty of Economics and Business Administration|