I recently had the chance to participate in a panel about energy efficiency at the Aspen Ideas Festival in Colorado. I expected the usual discussion of all the opportunities we’re missing to be more energy efficient and save money in the process—what Amory Lovins calls the “free lunch… you’re paid to eat.”

Instead, I found myself vehemently defending the very idea of energy efficiency against an idea with the odd name of Jevons paradox, which is undergoing a resurgence since David Owen’s article on it in The New Yorker.

Jevons paradox is named after William Jevons, who observed in the 19th century that an increase in the efficiency of using coal to produce energy tended to increase consumption, rather than reduce it. Why? Because, Jevons argued, the cheaper price of coal-produced energy encouraged people to find innovative new ways to consume energy.

Jevons paradox, as currently stated by Owen and others, is really an extreme statement about an effect economists commonly observe called “rebound”: some of the gains from energy efficiency are lost because people’s consumption rises in response to lower prices.

For instance, when the federal government requires more fuel efficient cars, aggregate demand by cars for gas is less. So prices tend to decline, and (to a limited effect) that lower price motivates a few people to drive a little more than they might have, perhaps taking advantage of the lower prices to take an extra weekend trip to the beach.

Jevons paradox claims that, over the very long term, the rebound effect can dramatically exceed the original gains from energy efficiency. A classic example is lighting, which has gotten vastly cheaper per unit as the world has moved from lamp oil to tallow candles to incandescent bulbs to fluorescent bulbs. Yet people now use more resources for lighting than we ever have in the past, since we have chosen to put lights almost everywhere.

Notice that this argument doesn’t just hold for energy, but really applies to use of any resource. If humanity is going to feed another 2.3 billion people by 2050, and accommodate increases in meat and dairy consumption from the rising middle class in places like India and China, we will need to roughly double food production. If we are to avoid having to plow under the Earth’s remaining natural forests and grasslands to reach this target, then clearly we will need to get more efficient in how we grow food.

However, Jevons paradox would suggest that in the process of making agriculture more efficient we will increase total food consumption, perhaps by supporting greater meat and dairy consumption than would otherwise be affordable for many people.

To mainstream environmentalists, this whole line of reasoning is blasphemy. Efficiency is seen as an unqualified good, a necessary first step toward a more sustainable society. If energy efficiency is the free lunch one is paid to eat, the sad truth is that environmentalists have only been partially successful at getting people to pick up that lunch: there’s still a lot of food rotting on the table. If environmentalists have had only partial success at promoting energy efficiency, what are the prospects of fighting for an even more fundamental change in our society’s relationship to resources?

I’ve struggled since Aspen to figure out why Jevons paradox seems to me so meaningless from the perspective of actual policy decisions facing society. Here’s what I’ve come up with:

1) Energy use by itself is not a bad thing. Indeed, anyone reading this blog online would view the life of hundreds of millions of the world’s poor, living in villages without electricity, one of extreme deprivation. With food, it’s even more clear: Access to enough food is a basic human right which close to a billion people are denied, although granted some of us in the developed world (myself included) sometimes eat so much it damages our health.

The issue is not consumption of a resource, but the environmental costs of satisfying demand. In other words, focus on limiting greenhouse gas pollution or erosion, not on limiting energy or agricultural production.

2) Jevons paradox suggests a false choice to policymakers: either make energy production and consumption more efficient, or do something more fundamental. It’s not clear why society can’t work on both options in parallel, especially since empirically the rebound effect for most technologies over the timescale of decades is much smaller than the original efficiency gain.

I like to call the next few decades “The Great Crunch.” As humanity strives to meet the resource demands of more than 10 billion people, many of them aspiring to live as resource-intense a life as people reading this blog, we will struggle greatly to protect or restore nature and the benefits it supplies us. Efficiency gains buy us time to make our whole economy more sustainable.

The more I confronted the Jevons paradox argument, the more it seemed to be just an excuse to stand back and do nothing. Why should government promote efficiency, ask proponents, when Jevons paradox would imply it’s a wasted effort?

To me they seem like fisherman on a sinking boat that, when the boat begins to take on water, would rather finish the beer they have on board than start bailing: “Pass the beer boys, nothing to do but enjoy the time we have left.” A very convenient attitude, but a dangerous one to the extent that it distracts people from, say, putting on their lifejackets or trying to build a lifeboat.

What do you think?

(Image: Compact fluorescent light. Source: Flickr user VickyvS via a Creative Commons license.)

If you believe in the work we’re doing, please lend a hand.


  1. Resilience is the new “sustainability”. Resilience includes conservation but comes at a cost of reduced efficiency. Globalized “Just-in-time” manufacturing is efficient but not at all resilient.

    Jevons Paradox occurs in a business-as-usual situation with increasing resource supplies. But as finite resources deplete we will be forced into consuming less in total amount, with increased resilience by switching fuels where possible in order to meet needs and demands. And increased resilience means for decreased efficiency, so Jevons Paradox won’t generally apply in a post-peak world.

  2. Hi Rob,

    There’s nothing particularly new about the Jevons paradox or its update, the Khazzoom-Brookes postulate. And the two are completely compatible with some very simple concepts in economics and plain old logic.

    I take exception to your comment that, to environmentalists, this line of thinking is blasphemy. First, environmentalism (and I consider myself one who supports addressing environmental concerns) is not an exercise in belief so characterizing these views on the effects of efficiency as ‘blasphemy’ rings hollow. And many people with well developed environmental sensibilities are sensitive to exactly this issue: “What’s the point of tryin’ to win this thing on the basis of efficiency? I go out and spend extra money for a Prius? That just leaves more cheap fuel for the guy who drives his 4X4 pickup to the office! I’m fed up!” You haven’t heard a comment like that before?

    Efficiency is great because it allows us to do more with less but that only helps the environment if we use LESS. Efficiency without constraints on access to carbon-based fuel just means more people deriving the same amount of energy benefit or the same number of people getting more benefit or some combination thereof.

    That’s why there has to be policy that limits production of carbon. If users react by demanding more efficient energy appliances (heating, autos, etc) so that they can get by with less fuel and more expensive fuel … great! But policy has to make carbon fuel use less attractive.

    This is the whole discussion about externalities. The market clears value. The market doesn’t recognize the negative value in atmospheric carbon. Policy has to create ways of bringing the negative value of emissions into the market as a COST. There has to be a limit or a penalty.

    Unless we deny China, India, and the global South any opportunity to develop, CFL light bulbs and electric cars aren’t going to mean a damned thing. Energy efficiency means nothing. It’s not the energy efficiency of the application, it’s the emissions intensity of the energy production.

    If many mainstream environmentalists are stuck in ‘change-your-lightbulbs’ mode (I think only the most naive and poorly informed are) that’s a shame.

  3. As an avid environmentalist, I certainly disagree with your take on how environmentalists stand on this issue.

    I believe that our current environmental problems (as well as just about every other problem from crime to poverty) are the result of poor behavior and policy, not because of some lack of technology. We could literally solve our environmental (or otherwise) problems overnight without any further advances in technology by just by willing it. For example, we could all choose tomorrow to stop driving everywhere and use public transit, walk, or cycle; to stop flying all over the world on vacations; to stop using throw-away plastics; to start composting food and recycling everything that can be; to stop developing our cities so that you have to drive; to just turn off all the lights on all the skyscrapers and roads where 99% of the time they aren’t used (we have an obsession with lighting up everything); to start buying food that is local and organic; etc, etc.

    But instead of focusing on the real cause of our problems — bad policy and the ensuing bad behavior — we love to worship at the altar of technology and claim that: “if we only had more efficient stuff, we could continue to act like wasteful fools.” The thing is: it’s the technology that has created our problems! Or rather, it is the gap between our technology and our understanding and implementation of that technology.

    For example, you can argue that crime is caused by a society which, though genetically the same as our hunter-gatherer ancestors, has become so removed from that environment in which we evolved — because of technology — that we are unable to handle the social implications of these detachment from our roots and people slip through the cracks and aren’t taken care of by the “tribe” and resort to crime. That doesn’t mean we need to go back to being hunter-gathers or that technology is bad, but simply that we love to only look at the good side of technology and mostly ignore the bad. This has allowed us to push technology to advance exponentially while almost completely ignoring the social implications of that technology (which are just as much bad as good). This has created a huge gap between the two, and that is where our problems lie. We are racing ahead with technology while almost ignoring any advances in our social institutions.

    Jevons Paradox is nothing more than the manifestation of the idea that you can’t use technology to solve what is, at its root, inherently a political and behavioral problem.

    Further, Jevons Paradox doesn’t say that efficiency is bad, but only that efficiency which is *not* accompanied by effective policy (ie, taxes, working less hours, etc.) will not have the desired effect of reducing consumption of a given resource. Again, technology is neither good nor bad, but just “is”. Once we realize this, we see that we must spend at least as much time understanding the effects of the technology on our society and behavior as we spend on creating the new technology.

    With all this in mind, it becomes very clear that any true progress in our society, from environmental to educational to health, *requires* social and behavioral progress; technological progress (ie, efficiency) is optional. However, if we went technological progress, than it must be accompanied by social and behavioral progress. We have failed miserably at this, and Jevons Paradox predicts this.

    So my problem with many environmentalists is that they over-emphasize technological progress and under-emphasize social/behavioral progress. Jevons Paradox is a useful way of understanding that this type of behavior will not solve our problems. We need to re-orient the environmental problems we face to address policy and behavior first, and any technology progress, ie, efficiency, second.

  4. Your post gives more weight to Jevon’s paradox than it does to refute it – i.e. Assuming that the world could support 10 billion humans without efficiency increases.

    Historically inaccurate and without cheap energy impossible.

    We’ve squandered our endowment (of cheap oil) and technology will not likely produce an equal. Solar, Wind, etc all require more energy to produce the components required to make panels, windmills etc let alone produce a net energy gain.

Add a Comment