Essays On Spatial Panel Econometrics

Identificador:  TDX:2813

Handle: https://hdl.handle.net/20.500.11797/TDX2813

Autors:  Miranda Gualdron, Karen Alejandra

Resum:
This thesis contributes to the spatial econometrics literature by presenting a suitable framework for modelling the spatial spillovers of the individual effects in panel data models. The key to this new approach is the use of a correlated random effects specification. Distinguishing the individual effects from their spatial spillovers may provide interesting insights into how the unobserved characteristics of the neighbouring territories affect the output of a certain territory and, conversely, how the unobserved characteristics of a territory affect the output of the neighbouring territories. In growth models, for example, a measure of the unobserved productivity of the regions under study can be obtained from the estimated individual effects Islam (1995). But how does the spatial contagion of the unobserved productivity of these regions affects their economic growth? This thesis considers several model specifications to identify and estimate this kind of effects. Moreover, each chapter of this thesis considers provides illustrative evidence. In particular, this thesis proposes using a correlated-random effects specification (Mundlak, 1978; Chamberlain, 1982) to analyse the spatial externalities of the individual effects. In this vein, the model specifications analysed in this thesis have a spatially weighted error component structure that is closely related to that proposed by Kapoor et al. (2007). In particular, the second chapter considers a correlated random effects specification in a static spatial panel data model and the third does so in a spatial dynamic panel data model. The fourth chapter makes use of the findings of the previous chapters in a growth model with spatial externalities.