An econometric time-series analysis of global CO2 concentrations and emissions

This paper extends previous work on the econometric modelling of CO;! concentrations and emissions. The importance of such work rests in the fact that models of the Cohen-Labys variety represent the only alternative to scientific or physical models of CO2 accumulations whose parameters are inferred rather than estimated. The stimulation for this study derives from the recent discovery of oscillations and cycles in the net biospheric flux of CO2. A variety of time series tests is thus used to search for the presence of normality, stationarity, cyclicality and stochastic processes in global CO;? emissions and concentrations series. Given the evidence for cyclicality of a short-run nature in the spectra of these series, both structural time series and error correction models are applied to confirm the frequency and amplitude of these cycles. Our results suggest new possibilities for determining equilibrium levels of CO2 concentrations and subsequently revising stabilization policies.