Unit root and cointegration tests for cross-sectionally correlated panels. Estimating regional production functions

This paper employs recently developed non stationary panel methodologies that assume some cross-section dependence to estimate the production function for Italian regions in the industrial sector over the period 1970-1998. The analysis consists in three steps. First, unit root tests for crosssectionally dependent panels are used. Second, the existence of a cointegrating relationship between value added, physical and human capital variables is investigated. The Dynamic OLS (DOLS) and Fully modified (FMOLS) estimators developed by Pedroni (1996, 2000, 2001) and the Panel Dynamic OLS (PDOLS) estimator proposed by Mark and Sul (2003) are then used to estimate the long run relationship between the variables considered.