(Georgia Institute of Technology, 2011-04-04)
Baek, Youngsun
This study analyzes consumers' behavioral responsiveness to changes in price and policy regarding residential electricity consumption, using a hybrid method of econometric analyses and energy market simulations with the National Energy Modeling System (NEMS). First, this study estimates price elasticities of residential electricity demand with the most recent Residential Energy Consumption Survey (RECS) data, collected in 2005, employing a conventional econometric model and a discrete/continuous choice model. Prior to the NEMS experiments with price shocks and consumers' behavioral features, this study uses NEMS to examine how energy policies would affect changes in retail electricity price in the future.
When climate policies are implemented nationally, electricity prices are estimated to increase by 17% in 2030 with a carbon cap and trade initiatives and by 4% with Renewable Electricity Standards (RES). The short-run elasticity of demand estimated from the 2005 RECS is found to be in a range of -0.81 ~ -0.66, which is more elastic than the current NEMS assumption of -0.15. The 2005 RECS dataset details information about American households' energy consumption. This rich source of micro-level data complements the existing econometric analysis based on time series data.
Electricity price (either census-division average price or household average price), annual income and number of rooms are found to be three major determinants of the level of electricity consumption. The difference in short-run price elasticity leads to a difference in social welfare estimates of energy policies and energy market forecasts. This study suggests that the estimate of social welfare loss caused by electricity price increase is overestimated if the elasticity is assumed to be smaller than the actual responsiveness. Supposing that 1) the short-run elasticity of -0.66 reflects the actual consumers' responsiveness to price changes in the present and future and 2) retail
electricity prices permanently increase by 10%, the welfare loss caused by the price increases would be estimated 0.9 billion dollars less than the current estimates with the elasticity of -0.15. This result suggests that if people are assumed to be more elastic to price signals, the time it takes for a policy to accomplish its goal could be shorter.
In addition to assessing potential savings expected from consumers' behavioral changes with the concept of price elasticity of demand in neoclassical economic theory, this study reviews economic and non-economic theories about behavioral features of energy consumers and discusses how existing information programs could be improved.