AN EMPIRICAL ANALYSIS OF MARGINAL EFFICICENCY GAINS WITHIN AND ACROSS END USES: IMPLICATIONS FOR THE REBOUND EFFECT FOR HOUSEHOLDS
Nour-El Imane Bouhou, The University of Texas at Austin, (512) 366-2741,
Dr. Mike Blackhurst, The The University of Texas at Austin, (512) 925-5645,
[Lead Author’s Name, Affiliation, Phone, email]
[2nd Author’s Name, Affiliation, Phone, email]
[Other Author’s Name, Affiliation, Phone, email]
[Format: single space, 10 point font, Times New Roman]
Overview
The effect of energy efficiency to reduce net demands has long been uncertain. Policymakers and researchers have emphasized demand-side strategies to reduce consumption of non-renewable energy sources [NRC 2009; EPA 2008]. As a result, efficiency improvements have become central to energy and environmental policy decisions. DSIRE (2013) reports over 1,400 U.S. programs that provide financial incentives for efficient technology adoption, administered at all levels of government. Spending on demand-side management has also increased. Incentives for electricity demand-side efficiency more than doubled from $1.9B to$4.2B from 2005 to 2010 [EIA 2013], and the Federal government invested $11B in efficiency as part of the American Recovery and Reinvestment Act of 2009 [DOE 2009].Engineering (or technical) assessments used to inform policy interventions assume an increase in technical operating efficiency leads to an equivalent decrease in energy use (ΔEfficiency = -ΔEnergy). [for examples see NRC 2009; EPA 2008; Creyts et al. 2007; Blackhurst et al. 2010]. These assessments disregard behavioral responses to technical change. Increased efficiency decreases the implicit cost of energy services, and consumers respond by increasing quantity demanded. This behavioral response is often called the “rebound effect and is often cast as “eroding” some or all of the technically feasible savings.Though several empirical analyzes have been conducted to evaluate the magnitude of the Rebound effect, the behavioral drivers of residential energy technology choice and use and respective net effects on energy consumption are relatively unknown. Previous quantitative assessments of such behaviors are generally limited to models of efficiency change of a single energy service using relatively simplistic demand functions or limited empirical assessments. This study thus provides a unique opportunity to empirically assess the electricity consumption (and thus rebound) implications of marginal efficiency change within and across end uses.
Methods
We develop empirical mixed regression models of marginal technical change within and across multiple household electricity end-uses, including air conditioning, appliances (refrigeration, clothes washing, and dishwashing), devices (televisions, computers, and tables) as well as photovoltaic panels. We use interaction terms in the designed mixed models to better understand how such marginal technical change affects consumption.We apply this model to data provided by the Pecan Street Project Research Institute. The sample includes 75 homes in Austin, Texas described by nearly 30 variables representing monthly consumption, household demographics, and physical household descriptions.
Results
We show empirically that direct rebound is relative to the baseline net technical efficiency of the affected service.Results demonstrate declining marginal rebound with consistent efficiency improvements in air conditioning, for example. As the efficiency of a given service consistently improves, consumers become satiated with the service and thus “take” marginally less back for additional services and that “enough” technical improvement within a given end-use will eventually overcome behavioral responses, leading to net energy reductions. Further, homeowners are not predicted to leverage efficiency gains across energy services for appliance use, but less robust evidence suggests that homeowners may rebound into increased consumption for devices, even in the short-run. We also find that the order of marginal efficiency change affects fitted values of consumption marginal. For example, consumption is expected to decrease for homes with higher R-values that then adopt efficient appliances but increase for homes with efficient appliances that then increase insulation. In addition, results indicate that homeowners receiving heavily subsidized solar panels do not use less grid energy, i.e., they used their subsidized energy source for more end-uses.
Conclusions
To better understand potential rebound effects within and across end uses, we develop and apply an empirical mixed regression model of marginal technical change for multiple residential electricity end-uses, including air conditioning, appliances (refrigeration, clothes washing, and dishwashing), and devices (televisions, computers, and tables). Results indicate that the rebound effect declines marginally with consistent technical improvements for air conditioning, suggesting that “enough” efficiency improvements can overcome respective behavioral responses within end uses. Across end uses, results indicate that homeowners do not leverage efficiency gains for appliance services, but less robust indications of expansion into consumption for devices is predicted. These results indicate that the net effect of technological change in households is relative to a home’s baseline technical efficiency and the degree to which homeowners seek new energy services (income and substitution effects). These results challenge single-service models of rebound when quantifying the net economic and environmental effect of efficiency. Results further challenge nearly all of the literature characterizing the efficacy of solar PV to meet environmental goals and policy mechanisms (incentives) for doing so. However, they are not entirely surprising. Solar PV’s demonstrate nearly free long-run operating costs, and grid-connected homes have access to an affordable source of energy. Our sample may therefore represent the long-term behavioral response to distributed solar PV panels.
References
DOE (U.S. Department of Energy). “Energy Efficiency and Renewable Energy: American Recovery and Reinvestment Act”. 2009. Home Page:
DSIRE (Database of State Incentives for Renewables and Efficiency). “Financial Incentives for Energy Efficiency,” 2013.
EIA (Energy Information Administration). 2013a. “Annual Electric Power Industry Report: Form EIA-861 Fata File.”
EPA (Environmental Protection Agency). National Action Plan for Energy Efficiency Vision for 2025. National Action Plan for Energy Efficiency Vision for 2025: A Framework for Change. Washington D.C., 2008.
NRC (National Research Council). Realistic Prospects for Energy Efficiency in the United States. Washington, D.C.: National Academies Press, 2009.
Blackhurst, M., I. Lima Azevedo, H. Scott Matthews, and C.T. Hendrickson, (2011a). “Designing Building Energy Efficiency Programs for Greenhouse Gas Reductions.” Energy Policy.
Creyts, J., A. Derkach, S. Nyquist, K. Ostrowski, and J. Stephenson, 2007. “Reducing US Greenhouse Gas Emissions: How Much at What Cost? “Executive Report; US Greenhouse Gas Abatement Mapping Initiative . McKinsey & Company Analysis, Washington, DC.