A note on the long-run elasticities from the energy consumption-GDP relationship