• rebound effect
• energy efficiency
• carbon dioxide emissions
• SVAR models

To reduce the carbon dioxide emissions, one key strategy is to decrease the energy consumption. We can achieve this target by improving the energy efficiency. However, there is a rebound effect, which is the difference between the expected decreased amount of energy consumption and its actual amount. Thus, it is crucial to study deeply on rebound effect and to estimate its size. We study the impact of the increased energy efficiency on carbon dioxide emissions, using structural vector autoregressive (SVAR) model, identified by FASTICA. Under limited assumptions, we use U.S. quarterly data by SVAR model to estimate the rebound effect. We found that the carbon dioxide emissions do not decrease in the long run.