Content area
Full Text
We consider per-capita carbon dioxide emission trends in 16 early industrialized countries over the period 1870-2028. Using a multiple-break time series method we find more evidence for very early downturns in per-capita trends than for late downturns during the oil price shocks of the 1970s. Only for two countries do downturns in trends imply downward sloping stable trends. We also consider trends in emission composition and find little evidence for in-sample peaks for emissions from liquid and gaseous fuel uses. These results lead us to reject the oil price shocks as events causing permanent breaks in the structure and level of emissions, a conclusion often made in analyses using shorter postwar data.
INTRODUCTION
While consequences of the climate warming remain highly uncertain, most scientists find it likely that emissions of carbon dioxide (CO2) and other greenhouse gases contribute to the warming. The rising atmospheric concentration of CO2 is generally viewed as the most important cause for the greenhouse effect. Economic actions that seek to reverse the "business-as-usual" trend of CO2 emissions may prove to be very costly: according to an overview of estimates produced by thirteen research teams, the implementation of relatively modest reduction targets may generate significant annual costs (Weyant and Hill, 1999; see also McKibbin and Wilcoxen, 2002).
Because of these characteristics, there has been a considerable recent interest in projecting the development of global CO2 emissions to the future. The econometric approach has been to model similarities across countries using reduced-form models estimated with cross-national panel data on CO2 emissions and indicators of economic development (e.g., Shafik 1994, Seiden and Song 1994, Holz-Eakin and Seiden 1995, Grossman and Krueger 1994, and Schmalensee et al. 1998). The global projections are then based on the assumption that development in developing countries brings about the emission pattern estimated for the industrialized countries. Much of the focus has been on a pattern called the "inverted U" relationship between emissions and income levels. The relationship is interesting because it implies that future reduction in emissions might follow as a by-product of economic growth.
This paper is not about the "inverted U. " Rather than testing hypotheses about global similarities in emission-income patterns, we choose a much simpler objective: What are the historical trends in per-capita emissions...