LightbulbThe rebound effect: a light that never goes out.Cross-posted from the Natural Resources Defense Council.

Every few years, a new report emerges that tries to resurrect an old hypothesis: that energy efficiency policy somehow results in consumers using more energy instead of less. This hypothesis was introduced in the 19th century by economist William Stanley Jevons, who argued that increases in the energy efficiency throughout a nation would lead to increases in coal consumption, rather than decreases.

Recent articles have attempted to revive these claims, also known as the “rebound effect” — restating that energy efficiency tends to encourage more energy use, not less, and that if a consumer’s immediate goal is to tackle climate change, then it seems risky to count on reaching it by improving efficiency. Assuming rebound effects eat up most of the energy savings, such claims then argue that efficiency cannot be a good policy to reduce energy consumption or combat climate change.

However, in a new report [PDF] published yesterday in the online journal,, my colleagues at the Natural Resources Defense Council dispel these claims, finding that:

Grist thanks its sponsors. Become one.

  1. Rebound effects are small, at high-end, and at the low-end, very well might have the opposite effect — efficiency might cause people to save more than was already expected.
  2. Rebound effects do not jeopardize our ability to reduce greenhouse-gas emissions or to lower our energy consumption because they do not change energy efficiency savings significantly. Indeed, to the extent they do, the effects appear to be positive.
  3. Grist thanks its sponsors. Become one.

  4. There are two different types of rebound theory, both of which have been discredited. The first, known as the end-use rebound theory, hypothesizes that people who own efficient appliances would use them more and thereby consume more energy. This is contradicted by comprehensive studies showing end-use rebounds to be small and decreasing over time, as well as the aggregate consumption metrics. For example, when environmentally conscious consumers switch to energy efficient lightbulbs, they don’t necessarily leave the lights on longer. The second, the economy-wide theory, hypothesizes that any energy savings from efficiency would be offset by money savings re-spent on activities that demand additional energy consumption. This is refuted by the fact that only a small portion (maximum of 6-8 percent) of those expenditures ever goes to energy. For instance, consumers who save money on their electric bill may use those savings for any number of things — food, movie tickets, or a child’s college fund. 

Our analysis — “Are There Rebound Effects from Energy Efficiency? — An Analysis of Empirical Data, Internal Consistency, and Solutions” — takes a detailed look at what rebound theory really says, how or whether this theory has been tested in the real world, or even how it could be tested, and what policy recommendations and results would flow from it if it were correct.

The first thing we found in our research is that rebound enthusiasts rarely define what they actually are predicting. This is an important failing, because the science of economics demands that theories be tested in such a way that the evidence either disproves or supports the hypothesis. Most rebound effect hypotheses are so casual that proponents of rebound can use any real world situation they choose to use and claim that it validates their ideas. Our study was able to find only two ways of stating rebound theory that are rigorous and capable of being tested. Both are firmly refuted by the evidence. The first version of a rebound theory that is scientific enough to be tested asserts that energy use grows in fixed proportion to the economy (to GDP). As my colleague Sierra Martinez elaborates, the history of the last 40 years in the United States and virtually all other developed economies shows this to be false. After implementing energy efficiency policies, many economies have indeed broken the lockstep increase in energy consumption with production of wealth.

The second version asserts that the savings to a state through efficiency policies will be much less than the sum of the expected savings from the policies and technologies one by one. However, California showed actual reductions in electricity use per capita over the last 40 years (compared to the rest of the nation) that not only equal the sum of the predicted savings but are actually four times as big. Other states showed similar, if smaller, results. Nowhere were there serious predictions of savings accompanied by disappointing results.

Of course, there are small rebounds in a limited number of end uses. For example, if you weatherize a drafty home, the occupants may be able to afford to keep it heated more comfortably, and there are indirect effects of efficiency on energy price and thus overall usage. But these effects have been incorporated into most energy models since the 1970s, and are found to be very small and decreasing over time.

Rebound proponents often want to rely solely on supply-side solutions. But if rebound suggests that efficiency may not save much energy, similar theories on the supply side would show that new clean energy sources may not reduce the usage of older dirty ones as much as expected either.

As you’ll read in the report, our analysis found that energy efficiency policies are not only the fastest way to reduce energy use but continue to be most the effective solution to combat climate change. The objections raised by rebound enthusiasts about efficiency policy and its effectiveness are inconsistent and so vague that they cannot be proven (or disproven). The data that is available about rebound theory indeed shows that its predictions are refuted. What we know is certain: Energy efficiency continues to offer us a strategy that allows people to enjoy a higher standard of living, with increased energy services, while decreasing energy consumption and combating climate change.