TESTING DIFFERENT STRUCTURES OF SPATIAL DYNAMIC PANEL DATA MODELS

There are different ways to model spatial data. One is by means of the so called spatial econometric models, originally introduced by Anselin (2001) for cross-sectional data and generalized in the last decade to different frameworks. In particular, in the recent years, the attention has been devoted to spatio-temporal data, with the introduction of the Spatial Dynamic Panel Data models. In this work we propose a strategy for testing different structures of such models, basing on a new estimation procedure and a multiple testing scheme that allow to choose between a general version of the model and some specific versions derived from the general one by imposing particular constraints on the parameters