About a quarter of US energy demand is for domestic use. Yet an understanding of the processes, determinants, and consequences of household energy demand remains elusive. Conventional energy policy has overwhelmingly focused on improving energy efficiency of the buildings. This research applies a non-linear methodology and an interdisciplinary approach to household energy demand. Using data from the US residential sector (2009 Residential Energy Consumption Survey), this research performs Covariance Structure Analysis to isolate direct and indirect effects of household and housing characteristics on total annual domestic energy use. Outcomes uncover some of households' indirect effects on energy demand, which in this research mainly happen through household effects on building characteristics, highlighting the indirect role of household choices in shaping residential energy demand patterns. To maximize its efficiency in reducing energy demand and GHG emissions, this paper suggests that in addition to investing in energy efficient technologies, energy policy should incorporate indirect effects of household choices on the configuration of their future homes. (C) 2015 Elsevier Ltd. All rights reserved.