Determinants of greenhouse gas emissions from household consumption in Sweden: Time-series and cross-sectional analyses

Other conference contribution, 2011

This presentation provides findings from two separate empirical analyses of energy use and greenhouse gas (GHG) emissions from Swedish households. The first analysis uses expenditure data from 2000 households with 100 expenditure categories and coupled energy and emissions intensities. The second analysis follows the development of aggregated consumption patterns and GHG emissions in Sweden over time from 1993 to 2006. Emphasis is put on the total GHG emissions of the households (transport, heating, food, entertainment etc), recognizing that reduced spending in one category may cause rebound effects through increasing spending in other categories (Nässén & Holmberg, 2009). The author belongs to a research group which consists of energy system modellers, economists and sociologist. Structural determinants of consumption are emphasized (e.g. income, urban form) while psychological determinants are not included. This research primarily builds on an empirical tradition rather than a theoretical, but an overarching aim is to add to the understanding of how society can be transformed towards long-term climate targets, for example if ecoefficiency will be enough (ecological modernization theory) or if more radical changes will be required (e.g. to the work-life balance; Nässén et al, 2009). The cross-sectional analysis confirms results from other countries showing that income (or total expenditures) is the most important determinant of energy use and GHG emissions. A 1% increase in total expenditures corresponds to increases of 0.78% in energy use and 0.83% in GHG emissions. Spatial determinants also proved to be important. For example, given the same household size and income, an average household living in a detached house in a non-urban area caused 28% more GHG emissions than an average household living in an apartment in a large city. Contrary, the level of education did not show any statistical significance. In the time-series analysis, a decomposition method is used to describe the change in GHG emissions as a series of factors (GHG/energy, energy/energy service, energy service/consumption, consumption/cap, cap). The development of energy service demand is here calculated as energy use under constant technical energy efficiency. This is done in order to separate the technical progress from the structural change of consumption patterns. A 1% increase in consumption corresponds to an increase in energy service demand by 0.80% which is very close to the result from the cross-sectional analysis. This finding gives some support for the use of results from cross-sectional analyses in scenarios of energy use if these are complemented by scenarios of technological change. The aggregated effect of structural changes of consumption, energy efficiency and fuel substitutions was a decrease in GHG emissions by 5% while consumption increased by 39%. While this can be seen as a substantial decoupling, this rate will not be enough to reach long-term climate targets.