Efforts to reduce carbon emissions significantly will require considerable improvements in energy intensity, the ratio of energy consumption to economic activity. Improvements in energy intensity over the past thirty years suggest great possibilities for energy conservation: current annual energy consumption avoided due to declines in energy intensity since 1970 substantially exceed current annual domestic energy supply.
    While historic improvements in energy intensity suggest great scope for energy conservation in the future, I argue that estimates of avoided energy costs due to energy conservation are overly optimistic. Avoided costs are likely to be significantly higher than estimates from recent energy technology studies suggest once behavioral responses are taken into account.
    I then analyze a data set on energy intensity in the United States at the state level between 1970 and 2001 to disentangle the key elements of energy efficiency and economic activity that drive changes in energy intensity. Rising per capita income plays an important role in lower energy intensity. Higher energy prices also are important. Price and income predominantly influence intensity through changes in energy efficiency rather than through changes in economic activity.