Search

This paper models externalities of production across regional economies. Under the assumption that knowledge diffuses without political or administrative barriers, we derive externalities that affect the steady state and the process of growth of each economy. The empirical counterpart of the reduced form equation summarizing the process of growth allows us to test for the presence of regional spillovers and to measure their magnitude. Our results for a sample of European regions show that spillovers are far from negligible, are robust to the consideration of variables within each region, and may cause nondecreasing returns at the spatial aggregate level. The paper also relates previous empirical evidence on spatial dependence in growth studies to the externalities modeled here. [ABSTRACT FROM AUTHOR]

Recent research has highlighted that in the last few years, the evolution of regional disparities in many European states has become pro-cyclical. This represents a change with respect to the predominantly anti-cyclical pattern of the 1960s and 1970s. This paper addresses the question of whether and when this change has taken place in the southern periphery of Europe, before analysing the factors that may have played a role in such a change. The analysis relies on a regional database that includes the evolution of the GDP per capita of NUTS II regions in five European countries (France, Greece, Italy, Portugal, and Spain) between 1980 and 2000. The results of the analysis support the hypothesis of a change towards a pro-cyclical evolution of regional disparities in the cases of Italy, Portugal, and Spain, but not in those of Greece and France. A relationship between these pro-cyclical patterns and the emergence of less dynamic sheltered economies is also detected in peripheral regions. This lack of dynamism is related to the fact that numerous peripheral areas in southern Europe have become increasingly dependent on factors such as transfers or public investment and employment, and therefore are less exposed to changes in market conditions. [ABSTRACT FROM AUTHOR]

In this article a semiparametric spatial Durbin model is employed to analyse the growth behaviour of 155 European regions in the period 1988–2000. This specification combines the semiparametric approach with the usual parametric spatial econometric technique to accommodate both spatial dependence and nonlinearities as suggested by recent neoclassical growth models with spatial technological interdependence. The results provide evidence of nonlinearities in the effect of initial per capita incomes and human capital investments. Moreover, the specification used allows identifying the effect of the interaction between the characteristics (initial conditions and structural variables) of each region and those of its neighbours. Finally, it shows some indication of global spillovers across-country and local spatial spillovers from domestic neighbours. Resumen En este artículo se emplea un modelo semiparamétrico espacial de Durbin para analizar el comportamiento del crecimiento de 155 regiones europeas durante el periodo 1988–2000. Esta especificación combina el enfoque semiparamétrico con técnicas econométricas espaciales paramétricas habituales para reconciliar la dependencia espacial y las no linealidades, tal y como sugieren los modelos de crecimiento neoclásicos recientes con interdependencia tecnológica espacial. Los resultados ofrecen pruebas de no linealidades en el efecto de ingresos per capita iniciales e inversiones en capital humano. Además, la especificación utilizada permite utilizar el efecto de la interacción entre las características (condiciones iniciales y variables estructurales) de cada región y las de sus vecinos. Finalmente, da una idea de spillovers (efectos de derrame) globales entre países y spillovers espaciales locales de regiones vecinas dentro del país. [ABSTRACT FROM AUTHOR]