Regional Growth and Convergence in Spain: Is the Decentralization Model Important?

This work is rooted on the analysis of growth and convergence at the regional level in Spain. Our contribution to that field is concentrated on the period 1980-2014, period characterized by a weak narrowing of the income per capita gap within regions. Several factors could explain that result. We focus our attention on the role of the political decentralization process in Spain, which actually began in the early eighties, on regional economic growth, a controversial and yet not enough studied issue. In Spain there are different models of decentralization, and even for each model, the different regions involved could follow different speeds gaining new administrative roles. Our econometric methodology is based on the system Generalized Method of Moments estimator. After using a general Mankiw-Romer-Weil approach, which fits well the Spanish data, our empirical work will implement other augmented growth regressions, which allow including a large set of explanatory variables. For such purpose, we try specifications with different proxies for the decentralization variable, as well as interactions with other variables that we think are linked to it, to capture the whole effect of decentralization. To sum up, our results, reinforced by several robustness exercises, are not conclusive on the relevance and sign of the effect of the decentralization path followed by the Spanish regions on growth and convergence, and points out to the importance of alternative factors. This result can contribute to the current debate in Spain on these topics. Several panel data analyses implemented with Stata. This work is rooted on the analysis of growth and convergence at the regional level in Spain. Our contribution to that field is concentrated on the period 1980-2014, period characterized by a weak narrowing of the income per capita gap within the territory. Several factors could explain that result. We focus our attention on the role of the political decentralization process in Spain, which actually began in the early eighties, on regional economic growth, a controversial and yet not enough studied issue. In Spain there are different models of decentralization, and even for each model, the different regions (Comunidades Autónomas) involved could follow different speeds gaining new administrative roles. Our econometric methodology is based on several panel data techniques implemented with Stata. Using a general Mankiw-Romer-Weil (MRW, 1992) approach, which fits well the Spanish data, our empirical work includes the decentralization variable as an additional regressor in the growth equation. For such purpose, we try specifications with different proxies for the decentralization variable, as well as interactions with other variables that we think are linked to it, to capture the whole effect of descentralization. To sum up, our results, reinforced by several robustness exercises, are not conclusive on the relevance and sign of the effect of the decentralization path followed by the Spanish regions on growth and convergence, and points out to the importance of other factors. This result can contribute to the current strong debate in Spain on these topics.