Peralta, SusanaWinkler, Marc Sebastian2022-07-212022-07-212022-01-132021-12-17http://hdl.handle.net/10362/142245Making use of EU-SILC longitudinal data, I investigate the persistence and dynamics of energy poverty in Portugal as measured by the ability to keep the home warm. I estimate a dynamic random effects probit model and find state dependencies of 15.2% in energy poverty. A household’s income, employment status, educational degree, and tenure status are found to be important predictors of energy poverty. I apply a multinomial logit model and find poverty duration states to highly depend on a household’s region of residence. Using a discrete time hazard model, I identify trigger events but find no significant effects. Key words: EU-SILC, Energy poverty, Fuel poverty, Dynamic random effects probit, Multinomial logit, Hazard model.engEu-silcEnergy povertyFuel povertyDynamic random effects probitMultinomial logitmaster thesis202975177