Residential natural gas demand: Assessing the evidence from Greece using pseudo-panels, 2012–2019

This paper investigates the behaviour of residential natural gas demand, employing a pseudo-panel methodology. The empirical analysis is based on annual household panel data for the construction of 185 cohorts in the case of Greece (2012–19). The specification of cohorts is based on household's head year of birth, education level and geographical location. A thorough econometric examination is carried out using a number of static and dynamic models. The empirical results show that residential natural gas demand is mostly price inelastic and almost income elastic. Idiosyncratic factors such as urbanisation and weather conditions seem to have a significant impact on residential natural gas consumption. On the other hand, socio-demographic parameters such as education level and the age structure of the household do not appear to play a significant role. Finally, electricity does not appear to be a substitute to natural gas for residences. Our empirical findings allow for energy policy proposals for the acceleration of transition to green energy sources. •Microdata are used for assessing residential natural gas demand in Greece, 2012–19.•Pseudo-panel static FGLS and dynamic GMM models are employed.•Residential natural gas demand is mostly price inelastic and almost income elastic.•Urbanisation and weather conditions affect residential natural gas demand.•Electricity does not appear to be a substitute to natural gas for residences.