Date of Award

Spring 1-1-2015

Document Type


Degree Name

Doctor of Philosophy (PhD)



First Advisor

Richard G. Rogers

Second Advisor

Jason Boardman

Third Advisor

Lori Hunter

Fourth Advisor

Fernando Risomena

Fifth Advisor

Landy Pena-Sanchez


This dissertation examines household energy consumption in the United States over the period of 1987 to 2009, specifically focusing on the role of socioeconomic status, demographic composition, and energy services profiles. The dissertation makes use of four cross-sections from the Residential Energy Consumption Survey data series to examine how household characteristics influence annual energy consumption overall, and by fuel type.

Chapter 4 shows that household income is positively related to energy consumption, but more so for combustible fuel consumption than for electricity consumption. Additionally, results for educational attainment suggest a less cross-sectional association and more longitudinal importance as related to income. Demographic composition matters, as predicted by the literature; household size and householder age show predicted effects, but when considered together, income explains any interaction between age and household size. Combustible fuels showed a far greater relationship to housing unit size and income, whereas electricity consumption was more strongly related to educational attainment, showing important differences in the associations by fuel type. Taken together, these results suggest a life course-based model for understanding energy consumption that may be strongly linked to lifestyles.

Chapter 5 extends the findings in Chapter 4 by examining the patterning of physical characteristics and behaviors within households. The chapter uses Latent Class Analysis to examine a broad set of energy significant behaviors and characteristics to discover five unique energy services profiles. These profiles are uniquely patterned across demographic and socioeconomic compositions of households and have important effects on energy consumption. These profiles are likely byproducts of the lifestyles in which the household takes part, due to factors such as their socioeconomic status and household demographic composition.

Overall, the dissertation finds strong evidence for taking a more lifecycle-based approach to understanding how energy is consumed based on the combined importance of householder age, household size, and socioeconomic status. Additionally, these factors produce very different energy services profiles that I argue are indicative of the lifestyle in which a household engages. Finally, the dissertation argues that these connections are essential to understanding energy consumption and provide a fertile ground for future research.