MethodsX (Jan 2020)
How to apply dynamic panel bootstrap-corrected fixed-effects (xtbcfe) and heterogeneous dynamics (panelhetero)
Abstract
The characteristics of panel data namely, inter alia, missing values, cross-sectional dependence, serial correlation, small time period bias, omitted variable bias, country-specific fixed-effects, time effects, heterogeneous effects and convergence often lead to misspecification, and spurious regression, thus, affecting the consistency and robustness of the model. In this regard, a more sophisticated panel estimation technique that accounts for the attributes and challenges is worthwhile. The novel panel bootstrap-corrected fixed-effects estimator (xtbcfe) and heterogeneous dynamics (panelhetero) recommended in this study meets almost all the requirements for robust and consistent panel estimation with an interface for user modifications. We further demonstrate how to use empirical CDF, moments and kernel density estimation to investigate heterogeneous effects. Due to the complexities in the application of xtbcfe and panelhetero algorithm, we provide a step-by-step procedure and guidelines for the estimation approach. We apply the xtbcfe and panelhetero algorithm for global estimation of mortality, disability-adjusted life years and welfare cost from exposure to ambient air pollution. Importantly, the xtbcfe algorithm can be applied to any panel data-based studies in social science, environmental science, environmental economics, health economics, energy economics, and among others. • Procedures useful for data imputation and transforming negative variables for time series, cross-sectional and panel data are presented. • Contrary to traditional models, we show how a novel approach can be modified and used to examine the degree of heterogeneous effects across cross-sectional units of panel data. • We demonstrate how the dynamic panel bootstrap-corrected fixed-effects estimator is useful in estimating higher-order panel data models and accounting for challenges such as omitted-variable bias, convergence, cross-section dependence and heterogeneous effects. • We apply the imputation technique, panelhetero, and xtbcfe algorithms to examine the nexus between ambient air pollution and health outcomes.
