Key findings of the international workshops on “Energy Efficiency Policies and the Rebound Effect”,
organized by the Climate Energy Decision Making Center and the University of Stuttgart under IRGC’s support
Authors: Inês Azevedo, Brinda Thomas, Marco Sonnberger, Ortwin Renn, Granger Morgan
The Climate and Energy Decision Making Center (CEDM) (USA) and the University of Stuttgart (Germany), in collaboration with Tsinghua University (China) and the University of British Columbia (Canada) organized two workshops on the topic of “Energy Efficiency Policies and the Rebound Effect”.
The first workshop was held in Washington, D.C. on 27 and 28 June 2011 and led by CEDM, while the second workshop was held in Stuttgart on 13 and 14 October 2011 and led by Dialogik gGmbH and the University of Stuttgart.
The objective of these workshops was to identify and develop research needs in respect to different approaches to measuring direct and indirect rebound effects that may arise from investments in and policies regarding energy efficiency. Before the workshop, CEDM and the University of Stuttgart conducted literature reviews and found a very mixed and diverse set of findings in terms of the magnitude and importance of rebound effects across end-uses, time periods covered, and regions or countries being studied. The goal from the workshops was neither to provide a consensus on the magnitude of the rebound effect nor to perform new research, but rather to identify the gaps in existing research and identify future research needs.
In preparation to each of the workshops, all participating experts were asked to produce a 3-page “think-piece” that addressed a specific set of questions. These are available upon request from Inês Lima Azevedo, at inesliazcmu@gmail.com
Why do we care about characterizing the issue of energy efficiency and the rebound effect?
Improving energy efficiency of the energy system is normally believed to result in a reduction of energy consumption, and in some cases to a reduction in greenhouse gas emissions (GHG). However, a number of authors who have conducted studies to estimate such savings find that part of the theoretically estimated gain in energy efficiency is eroded by additional consumption or changes in consumption patterns. This effect has been described in the literature as the energy efficiency rebound effect. However, the literature is not very clear in this area, in part because different authors use the same general phrase to refer to different phenomena. There is large uncertainty about the magnitudes of rebound effects and about how they differ by end-use and across different regions, time period considered, income groups, and stages of economic development.
Key findings from the participants’ presentations and research needs
First workshop
Many participants highlighted the need for a more carefully articulated taxonomy to explain the impacts from energy efficiency investments and change in effective prices of energy. There is a lot of confusion on what is included and excluded in the estimates produced in the existing empirical studies, and on whether these are even comparable across studies given the different methodologies and scope of the studies. The impact of efficiency gains on consumer behavior can be classified in terms of four different effects:
Direct energy efficiency rebound effect: Efficiency gains lead to an expanded or intensified use of the energy consuming product. For example, when a consumer switches from an incandescent light bulb to a compact fluorescent one, he may leave it on for more hours than he did previously because its operation costs less.
Indirect Rebound effect: The additional income that is freed up by saving energy costs can be used for other energy or carbon intensive consumption. For example, the income gained by installing an efficient furnace and insulating one’s house could be bundled into additional air travel, leading to an overall increase in GHG emissions.
The direct and indirect rebound effect can be decomposed into two familiar economic effects:
● Substitution effect: Efficiency gains in a particular energy service leads to a shift into more consumption of that service and out of other goods and services.
● Income effect: Efficiency gains in a particular energy service makes available additional income from energy cost savings which can be used for greater consumption overall, in both energy and other goods and services.
After accounting for substitution and income effects, the direct rebound effect focuses on net changes in energy service consumption, while the indirect rebound effect focuses on the net changes in other good or service consumption.
Economy wide rebound effect: Energy efficiency investments will lead to change in prices of goods and services, which will lead to structural changes in the economy and a new equilibrium in the consumption of other energy services and goods.
Changes in consumption patterns: Energy efficiency gains lead to changes in behavior (such as buying more frozen food when energy efficient freezers are available).
Without first sorting and classifying studies in these terms, it is not easy to tell when and whether rebound effects are large or small.
Studies on direct rebound effect generally show small effect and rarely show the potential for backfire (i.e., energy consumption increasing so that it is higher than under the baseline scenario). When indirect and economy wide effects are included, the range of rebound estimates increases.
There are very few comparisons for estimates of rebound across different end-uses and even fewer across countries. There was a lot of discussion among participants about the importance of understanding potential rebound effect in developing countries. However, sometimes this discussion included issues of understanding demand growth in developing countries, as opposed to rebound arising specifically from investments in energy efficiency.
Presentations by workshop participants strongly suggested that the biggest impacts are in production and in the developing world, so these should be better studied. There was a discussion of the sorts of tools needed to perform better analysis in some of the key production sectors, and whether one could do an analysis similar to what Jeff Tsao et al. have done for lighting [1] for other sectors.
Several participants expressed concern that the links to policy are not yet as well formed as they need to be. Namely, there is a link between avoiding rebound effect and in pricing energy “correctly” (including policy objectives) that requires further attention.
Participants noted that there has been relatively little application of ideas from the behavioral literature (e.g. bounded rationality) to the study of how people make decisions when selecting energy end-use devices and examination of the impacts of such decision-making processes in the context of estimating rebound effects.
Two other question that arose were: 1) how can energy modelers disentangle the effects of technological change and energy efficiency policies? and 2) what is the impact of energy efficiency investment and rebound effects on societal welfare?
More than understanding and gathering rebound estimates, several participants argued that it would be more important to understand the dynamics of consumption and understand what are the drivers to increase or decrease consumption.
Second Workshop
Interdisciplinary research is needed in order to integrate different disciplinary perspectives into one research framework. So far, the issue of rebound effects has mainly been studied from mere disciplinary (economic) perspectives. However, there are some projects under way trying to integrate a social, psychological and economic perspective on rebound effects. Participants identified some of the examples of ongoing European efforts addressing this issue, which include “REBOUND – The Social Dimension of the Rebound Effect” and “HECoRE – Household Energy Consumption and Rebound Effect”.
As in the first workshop, several participants in the second workshop raised the issue of understanding the interaction between welfare growth and energy efficiency: what are the impacts of energy consumption on social welfare? To what extent do rebound effects cause welfare effects?
Cultural differences in lifestyles and habits explain, at least in part, the differences in energy consumption in the USA and in the EU. While electricity prices are different in the two regions, residential price elasticity is very similar in the two regions [2]. This raises the unexplored question of how do cultural or lifestyle differences influence rebound effects?
A debate was raised by some of the psychologists in the room over whether there are education and learning processes that have to be taken into account and studied more deeply: Can consumers learn to avoid rebound effects like consumers have learned to do recycling?
Indirect and economy-wide rebound effects are very complex and difficult to estimate. More empirical econometric studies are needed. The group at CEDM and several researchers in the EU are now pursuing these efforts. At the same time, the theoretical basis has to be developed further in order to have a sound starting point for empirical research; a clear taxonomy on the effects from investments in energy efficiency and their consequences is still missing.
New models of individuals’ decision-making regarding energy consumption have to be developed. This should be done with the help of experimental research in order to identify causal relationships between certain variables, e.g. between knowledge about impacts of one’s own behavior and actual behavior.
An approach for studying rebound effects in organisations has to be developed. One would expect rebound effects in organisations to be lower than in households, because organisations are assumed to be more rational in their decision-making processes.
More survey research is needed to identify specific societal groups where rebound effects are highest. This should be done by going beyond the mere consideration of income groups (we already know that rebound effects are higher in low income groups). By identifying specific societal groups, intervention strategies can be tailored to their specific social contexts.
Policy implications from both workshops
While rebound effects could be large in the developing world, among low income groups, and in the production sector of the economy, there has been too little study of these groups.
Among reasonable well-to-do consumers, rebound effects appear to be modest (typically < 15%).
There is very little evidence of rebound effects exceeding 100% (so called “backfire”).
In situations in which empirical analyses suggest that rebound effects are non-negligible, these effects should be considered in the design of policy programs.
It is important to help policymakers understand that their policy strategies may not be as effective as simple direct analysis suggests. At the same time, care should be taken that energy efficiency policies are not called into question in general. Energy efficiency policies could be
improved by explicitly taking into account rebound effects. A strategy could be purposed to identify energy efficiency strategies where usage is not likely to change much, and therefore there will be little evidence of a direct rebound effect.
The UK systematically takes into account direct rebound effects, i.e. people increase the temperature of their homes due to financial savings from installed energy efficiency measures, in building design. Elsewhere, rebound effects are generally neglected in policy making. If more countries begin to pursue serious energy efficiency policies, the consideration of rebound effects will become more important.
So far, rebound effects are neglected in energy scenarios and models. For example neither the Stern nor the IPCC reports considered rebound effects. This omission raises the risk of underestimating future energy demand. However, it is difficult to include rebound effects in energy scenarios and models, not least because the knowledge of rebound effects is still vague. Some scenarios and models take into account rebound effects implicitly, but use a different wording. Here, clear definitions and a common wording are needed.
Footnotes:
[1] J Y Tsao, H D Saunders, J R Creighton, M E Coltrin and J A Simmons, Solid-state lighting: an energy-economics perspective, J. Phys. D: Appl. Phys. 43 (2010)
[2] Azevedo, I. L., Morgan, M. G., Lave, L. “Residential and regional electricity consumption in the US and EU: How much will higher prices reduce CO2 emissions?”. The Electricity Journal Jan./Feb. 2011, Vol. 24, Issue 1 1040-6190