Cointegration and causality: considering Iberian economic activity sectors to test the environmental Kuznets curve hypothesis

Few studies have attempted to study the environmental Kuznets curve (EKC) hypothesis at the individual sector level using more than one sector at once. This paper investigates the existence of the EKC hypothesis in the Iberian countries (Portugal and Spain) using thirteen economic activity sectors for each, analyzing each individual sector's cointegration and causality relationships considering carbon dioxide (CO2) emissions, sector gross value added and energy consumption. The findings of this paper using the autoregressive distributed lag (ARDL) approach only validate cointegration in six Portuguese sectors and in five of the Spanish sectors. Results confirm both short- and long-run bi-directional and unidirectional causality between economic growth, energy consumption and CO2 emissions, using the error correction model (ECM) and Toda and Yamamoto's causality approaches. Moreover, results for Portuguese and Spanish sectors indicate an inverted U-shaped relationship only for one sector each. In some sectors there was evidence of a U-shaped relationship and in others the EKC hypothesis could be verified but no statistical significance was obtained. The study has significant contributions for sector policy, including implications to curtail energy pollutants by implementing environmental friendly regulations to sustain economic development at the sector level in the Iberian market. It also allows inferences to be made about the existence of different behaviors in comparative terms for the same economic activity sectors of the individual countries.