Quantifying the Effect of the Principal-Agent Problem on U.S. Residential Energy Use

Panel: Strategies for Appliances, Lighting, Electronics, and Miscellaneous End–Uses

Authors:
Scott Murtishaw, California Public Utilities Commission
Jayant Sathaye, Lawrence Berkeley National Laboratory

Abstract

Quantifying the amount of energy associated with market failures helps to demonstrate the significance of energy efficiency policies beyond price signals. In this report we investigate the magnitude of the principal-agent (PA) problem affecting refrigeration, water heating, space heating, and lighting in the U.S. residential sector. We develop an approach for classifying households into a PA matrix for each end use that allocates end-use equipment into four different PA classifications based on household characteristics. End use energy values differentiated by housing unit type were used to estimate the total energy use associated with the PA problem. We find that the 2003 associated site energy use from these four end uses affected by the PA problem totaled over 3,400 trillion Btu, equal to 35% of the site energy consumed by the US residential sector. In addition, we also estimated an upper bound on annual hypothetical energy savings of 4.8 trillion Btu of site energy if there had been no PA problems affecting refrigerators and water heaters sold in 2003. This savings potential is considerably lower than it would have been in the absence of minimum energy performance standards. Policies and programs that would be effective differ among the four different classes of the PA problem.