Explaining the Election Results in Portugal: A Spatial Econometrics Point of View

The great majority of the theoretical analysis about political business cycles (PBCs) has considered the national space as the territory of interest for the study of the economic consequences of an electoralist behaviour by the central government. This fact, in conjunction to the nature of the data most commonly available, has lead many authors to empirical studies which, by the use of more or less sophisticated econometric techniques, intend to verify the empirical evidence of PBCs. Given that the election results for the main parties, at least for Portugal, clearly reflect some spatial localization we find rather intriguing to verify that so very few of those empirical studies use spatial econometrics techniques. A causal observation on the data concerning the (Portuguese) election results over space shows that the results obtained by the incumbent, at a regional level, should not be considered completely independent of the party ruling the distinct municipalities distributed over the national territory. This being said, the paper’s main objective is the analysis of the results corresponding to the last legislative election that took place in Portugal, from the PBCs viewpoint, by the use of well-known techniques of spatial econometrics. The confrontation of the results with the ones obtained ignoring the spatial localization of the data will lead us to the nature and extent of the improvement on the results obtained by spatial econometrics techniques in what concerns the detection of empirical evidence supporting the existence of PBCs in Portugal.