New evidence on the convergence of per capita carbon dioxide emissions from panel seemingly unrelated regressions augmented Dickey–Fuller tests

Using the data for per capita carbon dioxide (CO2) emissions relative to the average per capita emissions for 21 countries in the organisation for economic co-operation and development (OECD) covering the period 1960–2000, this paper seeks to determine whether the stochastic convergence and β-convergence of CO2 emissions are supported in countries with the same level of development. In other words, are shocks to relative per capita CO2 emissions temporary in industrialized countries? We respond to this question by utilizing Breuer et al.'s [Breuer JB, McNown R, Wallace MS. Misleading inferences from panel unit-root tests with an illustration from purchasing power parity. Review of International Economics 2001;9(3):482–93; Breuer JB, McNown R, Wallace MS. Series-specific unit-root tests with panel data. Oxford Bulletin of Economics and Statistics 2002 64(5):527–46] panel seemingly unrelated regressions augmented Dickey–Fuller (SURADF) unit-root tests, which allow us to account for possible cross-sectional effects and to identify how many and which members of the panel contain a unit root. Our empirical findings provide evidence that relative per capita CO2 emissions in OECD countries are a mixture of I(0) and I(1) processes, in which 14 out of 21 OECD countries exhibit divergence. The results reveal that conventional panel unit-root tests can lead to misleading inferences biased towards stationarity even if only one series in the panel is strongly stationary.