Romero-Avila, DiegoOmay, TolgaEconomics2024-07-052024-07-05202221420-20261573-296710.1007/s10666-022-09835-42-s2.0-85130553155https://doi.org/10.1007/s10666-022-09835-4https://hdl.handle.net/20.500.14411/1752This study applies a large battery of state-of-the-art nonlinear unit root tests to examine the stationarity properties of carbon dioxide emission series for 28 industrialized countries, five BRICS and seven transition economies over a very long horizon, in some cases over more than two and a half centuries. The application of time-dependent and state-dependent nonlinear unit root tests separately provides mixed evidence regarding the time-series properties of CO2 emissions and a high degree of variability across the different tests. However, the use of hybrid nonlinear unit root tests, combining the presence of structural breaks with symmetric or asymmetric ESTAR adjustment, leads to the rejection of the unit root hypothesis in each of the countries under study with at least one of the hybrid tests. This has important climate policy implications.eninfo:eu-repo/semantics/closedAccessCO2 emissionsNonlinearitiesUnit rootTime dependenceState dependenceLSTAR processESTAR processAESTAR processStructural breaksArticleQ3Q2274621643WOS:000799068900001