Abstract
Endogeneity problem mainly arises in quantile regression for dynamic panel data models due to the inclusion of lagged value(s) of response variable. The most utilized approach considers the involvement of reliable instrumental variable(s) to reduce estimators' biasedness. We employ the second approach to estimate dynamic panel data utilizing quantile regression according to a two-stage process. A new structural estimator is proposed for estimating dynamic panel data with two-stage quantile regression. To acquire the privilege of robustness, both stages are estimated at the same order of quantile. In addition to that, we provide the properties of the new estimator in terms of equivariance, invariance to monotonic transformations and robustness.