The paradox of efficiency: Can uncertainty be governed?

Zora Kovacic, Louisa Jane Di Felice and Tessa Dunlop

In a world of limited resources and increasing human impact on the environment, using resources more efficiently seems sensible. Many policies see efficiency as an important instrument to achieve their goals. In the case of energy policy, the EU has published in 2012 a directive on energy efficiency and in June EU energy ministers agreed to support a 30% energy efficiency target for 2030 as part of proposed legislation to improve the EU's electricity market. In water management, efficiency is seen as a means to deal with water scarcity in arid regions. In waste management, resource efficiency is pursued as a means to reduce waste production. But does efficiency guarantee that less resources will be used? Does it guarantee that resources will be used better? The Jevons paradox suggests that the answer is not so straightforward and that efficiency policies may not achieve the desired results.

In 1865, William Stanley Jevons observed that increased efficiency in coal engines led to an increase in consumption of coal in a wide range of industries. The improvements in coal engines made it possible to use engines not only in coal mines, but also on rail and sea transport. Jevons concluded that, contrary to common intuition, increases in efficiency do not necessarily reduce resource consumption because they also open up for new applications and uses and ultimately new demands. This is called “the Jevons paradox”. This paradox is one of the many ways that complexity displays itself. In a complex system, if a part is changed or taken out and substituted with a different part, interactions within the system may change and lead to surprising and paradoxical changes throughout the entire system. The Jevons paradox suggests that efficiency policies may not lead to the desired outcomes, because the economic system will adapt to increased efficiency and technological improvements.

A similar concept has emerged also in economics, called the rebound effect. The rebound effect is the reduction in expected gains from increases in efficiency, because of systemic responses to the increase in efficiency. While the rebound effect recognises that systemic responses may offset the benefits of technological improvements, it does not presuppose changes in the essential workings of the system. The rebound effect can be calculated through mathematical formulas, which assume that the interactions between the parts of the system remain stable. There are sometimes varying definitions, but scholars generally differentiate between 1) direct, 2) indirect 3) economy-wide and 4) transformational rebound effects, with the latter most comparable to the Jevons paradox. From the point of view of complexity, however, the rebound effect is different from the Jevons paradox in as far as changes in complex systems cannot be precisely calculated.

What this means is that the rebound effect essentially leads us to do more of the same thing, while Jevons paradox leads us to do something different. To make this distinction clearer, we can draw a parallel with diets. If I am trying to cut my calories to lose weight and decide to buy fat free yogurts, I may end up eating two fat free yogurts instead of a regular one – leading overall to a higher caloric consumption. This would be the rebound effect. On the other hand, I could also eat a fat free yogurt and then, feeling that I have saved on calories, I could take the bus instead of walking, or go out and eat a slice of pizza. This would be the Jevons paradox. This doesn’t necessarily mean that one should stop buying fat free yogurts, or stop improving our efficiency, but it does have implications for governance.

The existence of direct rebound effects is uncontroversial, with quantitative evidence in a large number of studies. The possible effects of the Jevons paradox and how to measure it, however, are in dispute. But rather than focusing on technicalities, the Jevons paradox reveals an important philosophical dilemma regarding complex systems. Because it focuses on unforeseen changes in the interactions between the parts and the identity of the whole, the paradox cannot be modelled nor predicted with precision. Therefore The Jevons paradox and the rebound effect have different implications for policy, and cannot be treated as equivalent. The rebound effect suggests that gains in efficiency can be estimated and that efficiency policies are a means to govern complex systems (although these are not as effective as one may hope). The Jevons paradox instead suggests that complex systems cannot be controlled, and that increases in efficiency may not produce the expected results. Given this uncertainty, which theory should policy rely on for advice? If one takes the Jevons paradox seriously, governance is as much a matter of relying on evidence as it is about taking into account uncertainty.



Sorrell, S. Jevons’ Paradox revisited: The evidence for backfire from improved energy efficiency. Energy Policy. 37 (2009) 1456–1469. (footnote for paradox being in dispute)

Greening, L. A., D. L. Greene, and C. Difiglio. 2000. Energy efficiency and consumption—The rebound effect—A survey. Energy Policy 28(6–7): 389–401.