A Free-Market Energy Blog

Energy ‘Rebounds’ and ‘Backfires’: An Introduction and Literature Overview

By -- July 17, 2012

Much of today’s energy policy assumes that regulations mandating greater energy efficiency will reduce energy use. But that isn’t always the case, and energy efficiency improvements are seldom as large as promised by engineering calculations because of “rebounds.” Such is the most general conclusion from hundreds of studies pertaining to the effects of energy efficiency, whether market or nonmarket.

Such is the message from my literature review published by the Institute for Energy Research (pictured here):

For example, people who install lighting that is 50 percent more efficient frequently leave the lights on longer, negating some of the energy savings from greater efficiency. This is called an energy efficiency rebound. Sometimes these mechanisms even bring about net increases in energy use known as backfires.

Rebounds have a direct implication for energy efficiency mandates and incentives. If rebounds are substantial, efficiency policies will be less effective at reducing air pollutants, for example, because the “saved” energy gets consumed elsewhere. Energy consumption may even increase on net to cause backfires.

Rebounds, and certainly backfires, fall into the ‘unintended consequences’ category of government intervention into the complex market. It is reason to ‘let the market decide’ with energy usage, as in energy production.

Four ‘Rebound’ Types

There are four basic types of rebound that might result from improved energy efficiency, defined by the markets in  which their effects occur: direct rebound, indirect rebound, economy-wide rebound, and embedded energy.

Direct rebounds are adjustments in the production or consumption of a good whose energy efficiency has increased. For example, improved vehicle fuel economy that lowers the per-mile cost of driving may motivate drivers to travel more miles than otherwise. Increased gasoline used is a direct rebound.

Indirect rebounds are changes in the production or use of goods related in use to the activity being improved in efficiency. For example, increased fuel economy that leads to more driving also indirectly increases the demand for tires. The resulting increase in the tire industry’s energy use is an indirect rebound.

Economy-wide rebounds are the impacts of an efficiency improvement summed over all affected economic activities. If drivers are traveling more often, hotels will require additional energy to meet increased demand for rooms and services; hotel furniture manufacturers likewise increase their energy consumption to boost outputs.

Embedded energy is where energy is expended in the process of creating more energy-efficient goods. Though high-efficiency building insulation lowers annual energy use, for example, the manufacture and installation of more efficient building materials also requires energy inputs that must be accounted for.

Direct Rebounds

More than 200 studies have defined and estimated direct rebounds. One of the first was by W. S. Jevons, who explained what became known as the “Jevons Paradox” in his 1865 classic, The Coal Question. One of the most recent was published by the Left-of-center Breakthrough Institute in 2011, Energy Emergence: Rebound and Backfire as Emergent Phenomena.

Though most research on direct rebounds has generally verified their existence, their magnitudes vary greatly: their values depend on the energy-using activity under study (e.g., cooking), characteristics of the subject population, and possible biases in sampling.

For example, studies have found:

1) Household behaviors before and after installation of energy-efficient appliances produce wide ranges of rebounds, for example between 10 and 60 percent for electric heating in the short run.

* As a practical example of a 10-60 percent rebound range, assume a 2000 square feet house has a 20 kw electric furnace that uses 2,434 kwh per month, and operates 4 hours a day during the heating season only (about 6 months). At a cost of $0.0698 per kwh, it costs $170 per month to heat the house with this equipment.

* Assume that the furnace is replaced with an electric heat pump, which consumes 1,642 kwh a month, and a heat pump fan that consumes 90 kwh per month. Energy use is decreased to 1,732 kwh and results in a bill of $120.96, assuming that there is no change in power consumption.

* Now we can experiment with rebounds. In the range of Sorrell’s figures (from Chapter 3) the gross saving of 2434 – 1732 = 702 kwh can be netted against rebound. If the rebound is 10 percent the household consumes 1,732 + .10*702 = 1,802 kwh, with a new bill of $125.79. At a high-end 60 percent rebound consumption is 1,732 + .6*702 = 2,153 kwh, with a new bill of $150.29.

2) In particular, wealthy households that already own all major appliances do not reduce energy consumption after buying more efficient ones; instead, further increases in their incomes are often spent on energy-intensive services like travel.

3) A high percentage of utility-sponsored conservation and efficiency programs have found that actual savings fall short of projected ones, a possible manifestation of rebounds.

4) Improvements in energy efficiency can raise the productiveness of other inputs, e.g. works in a better-lit plant are more productive. These increases in non-energy productivity increase the profitable scale of production and bring higher energy use.

Economy-wide estimates of rebound use “computable general equilibrium” (CGE) models to track input-output relationships between sectors and have the ability to simulate the economy over longer durations during which capital investments in energy and energy-using equipment are taking place.

Though the limited number of CGE studies makes it difficult to draw firm conclusions, a general principle appears to be emerging: rebound effects are greater for models with more comprehensive structures and with simulations of longer duration. Over half of the available studies using CGE show rebounds that approach or exceed 100 percent. In other words, their net result is that more energy was consumed than saved.


Groups such as the Breakthrough Institute contend that if greenhouse gases are to be reduced, rebound must be overcome by direct government intervention in markets, rather than with efficiency mandates; however, in the absence of uniform worldwide policies, the ubiquitous nature of greenhouse gases would render such policies ineffective. Implementation would also require governments to identify “winner” technologies, regardless of their capability to do so.

The pervasiveness of energy efficiency rebounds illustrates that attempts to plan or direct energy policy toward desired goals will likely fall far short of expectations. Instead of imposing energy efficiency mandates, energy policy should embrace market prices and disruptive innovations to guide energy to its most valuable uses.


Robert J. Michaels is Professor of Economics at Cal State Fullerton and a senior fellow of the Institute for Energy Research. At his website is a  full bio.


  1. Robert Michaels: Energy ‘Rebounds’ and ‘Backfires’: An Introduction and Literature Overview | JunkScience.com  

    […] MasterResource Share this:PrintEmailMoreStumbleUponTwitterFacebookDiggRedditLike this:LikeBe the first to like this. This entry was posted in baseload energy, Energy efficiency. Bookmark the permalink. ← Tim Blair: It’s a hit! […]


  2. Ed Reid  

    I suspect that the Jevons Paradox is most obvious in situations in which those purchasing or using the more efficient device have been practicing conservation with regard to the utility provided by the more efficient device. For example, if I have been keeping the temperature in my home at 72F, because that is the temperature at which I feel most comfortable, there is little likelihood that I would change that temperature setting because I purchased a new, more efficient furnace or heat pump/air conditioner. However, if I were still keeping my thermostat at 66F during the winter (and wearing my Jimmy Carter cardigan) and at 78F during the summer, either to conserve energy or reduce operating costs, I might be inclined to reset the thermostat to more comfortable levels after purchasing a higher efficiency furnace or heat pump/ air conditioner. Similarly, if I have been cutting down on driving to conserve energy or reduce operating costs, I might choose to travel more frequently after purchasing a more fuel efficient vehicle.

    Paradoxically, in these cases, efficiency offsets conservation.


  3. Andrew  

    I suspect what is at work here is something that applies more generally than to energy. An0ther example where instituting a particular policy simply leads to changes in behavior that wipe out any of the intended “benefit” would be laws requiring people to wear bicycle helmets: I believe it has been shown people tend to ride much less carefully when they think that the helmet will protect them from serious injury.

    Oh, I am also partial to that episode of King of The Hill where Arlen bans trans-fats. Bill thinks he is now safe to eat like crazy, because the government is protecting him from unhealthy food: he winds up fatter than he already was.

    Fundamentally the problem the government faces in making policy is predicting human behavior in reaction to government policy. To accurately predict such changes and design your policy around them would require knowledge the government does not-cannot-possess.

    But I suspect I am preaching to preachers on this one.


  4. Energy Conservationism: Taxpayers, Consumers BewareInstitute for Energy Research | Institute for Energy Research  

    […] And remember: increased energy efficiency does not translate into less energy usage overall, as explained by Robert Michaels in his new study, Energy Efficiency and Climate Policy: The Rebound Dilemma, summarized here. […]


Leave a Reply