Given the growing availability of big datasets which contain information on multiple dimensions and following the recent research trend on multidimensional modelling, we develop three-dimensional panel data models with threeway error components that allow for strong cross-sectional dependence (CSD) through unobserved heterogeneous global factors, and propose appropriate consistent estimation procedures. We also discuss the extent of CSD in 3D models and provide a diagnostic test for cross-sectional dependence. We provide the extensions to unbalanced panels and 4D models. The validity of the proposed approach is confirmed by the Monte Carlo simulation results. We also demonstrate the empirical usefulness through the application to the 3D panel gravity model of the intra-EU trade flows.
Modelling in the presence of cross-sectional error dependence
Mastromarco C.;
2017-01-01
Abstract
Given the growing availability of big datasets which contain information on multiple dimensions and following the recent research trend on multidimensional modelling, we develop three-dimensional panel data models with threeway error components that allow for strong cross-sectional dependence (CSD) through unobserved heterogeneous global factors, and propose appropriate consistent estimation procedures. We also discuss the extent of CSD in 3D models and provide a diagnostic test for cross-sectional dependence. We provide the extensions to unbalanced panels and 4D models. The validity of the proposed approach is confirmed by the Monte Carlo simulation results. We also demonstrate the empirical usefulness through the application to the 3D panel gravity model of the intra-EU trade flows.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
