The Jevons Paradox: Time to Send it The Way of the Dodo?

William Stanley JevonsThe Jevons Paradox has been the elephant in energy efficiency's room since energy efficiency was in diapers. It casts a gloomy shadow over the industry, raises doubts about the sanctity of our mission, and the fact that it exists at all is, frankly, kind of a drag.

But here's the thing: energy efficiency is all grown up now, and Jevons is dead. It's Hans Castorp's grandfather-on-the-wall in The Magic Mountain—archaic, mysterious, and useless. Time to move on.

In case you're unfamiliar, the Jevons Paradox was first suggested by William Stanley Jevons in 1865 in a book called The Coal Question, and essentially claims that as technological improvements increase the efficiency with which a resource is used, use of that resource increases rather than decreases. The classic case in point was that when James Watt invented his coal-fired steam engine, which was drastically more efficient than Thomas Newcomen's earlier design, coal became a more cost effective power source, its range of applications increased, and ultimately, coal consumption boomed. Good bye, blue skies.

The same principle has held true for nearly every other major fuel source since. Likewise, in more modern times, the Jevons phenomenon has manifested itself in the growing efficiencies of typical houses (as measured by mmbtu's/sq ft), that have been offset by—guess what—larger houses. (This may be changing; see below.)

But while it continues to be broadly accepted as an economic principle, we think there's growing reason to question the validity of Jevons in a new era, and certainly good cause for those of us in the energy efficiency business to stop talking about it as though it were an immutable law of nature. Here's why:

1) It's no longer the height of the Industrial Revolution. James Watt was born in 1736. Times have changed (needless to say, but we'll say it).

2) Times have changed more recently, too. The housing boom is over. Americans are no longer looking to get rich quick (or feel rich quick) by investing in bigger, badder, further out of town houses. As Warren Buffett wrote of housing in his 2009 shareholder letter, "prices will remain far below 'bubble' levels..." I question whether those days will ever return. 

3) Today's economic environment is uniquely bad, and, despite our highest hopes, may stay so for some time; so the trend of increased consumption that's carried on for the past 50 years may be grinding to a halt. As Nobel-prize winner Paul Krugman predicts, things "could be unpleasant for a very long time." 

4) Demographics are changing. Today's generation, the "Echo Boomers," is big, and entering the home buying years, but not as wealthy as the baby boomers. They like to spend money, but maybe not quite as much as their parents. And perhaps more importantly, research suggests that today's up-and-coming, soon-to-be-home-buying generation is the most socially conscious, environmentally driven generation in recent history. I'm hopeful that this is the renovation generation, less concerned with real estate flipping, more committed to quality of place and the impact of their housing on the planet. 

5) Houses are no longer growing. While increased efficiency in homes over the past 30 years was offset by bigger homes, what we're seeing in the post-bubble housing market is a veering away from McMansions in favor of smaller, more affordable, more efficient homes

6) Energy prices are growing faster than incomes. Despite the energy price hiccup of the last couple years, the broad trend is that energy prices are on a steady rise, taking into account the recession drop, and even in the absence of shock factors like peak oil or major supply disruptions. The growing disparity, and increase in energy costs as a percentage of family income, means that people will need efficiency just to survive, unless we want to live in a state of permanent recession.

Of course, this is, in the end, an opinion piece, and we'd be tremendously curious what you might have to say about the matter. Let us know in the comments section.

ENDNOTE: Two posts on the topic of energy efficiency versus conservation lit the match for this one. Thanks to Allison Bailes' post on Energy Vanguard and to Sean Lintow's post on the SLS Construction Homeowner's Resource Center Blog. Both highly recommended reads. 


When I saw the title of this article, I was ready to pounce. After reading it, though, I have to say that I agree with your take on this topic, Peter. You've painted a pretty good picture of the evidence supporting your thesis that Jevons's Paradox doesn't rule the day anymore.

You even mentioned what I consider the driving force behind this shift - peak oil. As long as we have untapped energy sources large enough that allow supply to stay ahead of demand, Jevons's Paradox will be with us because more efficiency creates more demand.

As supplies get tight, however, conservation becomes a requirement. If the economy stops growing, people get poorer and need to cut back.

There are a whole lot of peak oil books and websites out there now, so I'm not going to explain the issue here, but I recommend and to anyone who wants to learn more.

Thanks for another great article in this series, Peter, as well as your recommendation of my article.

I think it's important to make a distinction between The Jevons Paradox and the Rebound Effect.  The former postulates that efficiency improvement increase the use of the fuel; the latter says that the effect of efficiency improvements are offset to some degree by the resulting increases in demand for and price reduction of the fuel.

It is hard to argue that the Rebound Effect will not come into play -- in fact, we're probably seeing it in oil prices now.  After the oil price spike that culminated in 2008, there have been some dramatic shifts toward use of more efficient alternatives -- smaller cars, Cash for Clunkers, more efficient vehicle fleets, the death of the Hummer, etc.  Of course all of this is muddled with the economic crash, which had more, eh hem, traditional causes for reduction in prices -- other causes resulting in lack of demand because people had much less money to spend.

That said, I think the current situation is quite different than the past.  

As noted, Peak Oil seems likely to gradually squeeze the availability of the oil resource, at least in the US.  Recent technological improvements (arguable that they are "better") in natural gas extraction has resulted in a shift to this fuel source from both oil and coal because the cost has fallen, relatively.  To the degree that coal is a substitute for oil, we may also find a shift towards coal (all other things being the same) as oil prices rise because it is so plentiful in the US -- and if all-electric cars like the Volt and Leaf are successful, this will likely drive some of this shift.

But perhaps we also have to talk about other changes.  For one, some renewable energy sources (e.g. Solar PV) are near "grid parity" costs today, and as technical improvements in efficiency arise the same kind of shift I mentioned that has happened with gas would be expected to happen with renewables: the cheaper alternative is likely to be chosen.

So yes, a rebound effect may occur as these efficiency improvements occur.  But what's notable, both in the case of coal to gas, and coal or gas to renewables is that even if we do end up getting cheaper power and using more of it (as rebound effect predicts), in this case, it's a good thing!  The Jevons Paradox was about a new technology for turning coal into power, and more coal got used as the more efficient steam engine was deployed.  But natural gas releases 1/2 the CO2 as coal when burnt, and of course renewables have a far, far lower CO2 footprint.  Yes, reduced demand for other fuels will tend to lower their prices, so it's an uphill battle, but it is attenuated by other factors like peak oil.

Whether this year or next, or in 2020, eventually the world will realize that spewing CO2 into the sky is a costly thing, and we'll come to our senses and start adding that cost to the product.  And this will change everything.  Even in the absence of legislation to force this, many other forces are making coal (in particular) less attractive.  Environmental issues with mountaintop removal have recently resulted in the EPA taking action to prevent this practice to some degree.  Fewer and fewer coal plants are "just getting built" as they were in the past, and this raises the cost of delivering power.  Resistance is not futile!

So if the total amount of energy consumed eventually increases in accordance with the Rebound Effect or Jevons Paradox, and if the cost of oil, coal and even gas fall because of competition from renewables, it's OK.  The problem we have to resolve is mainly climate change.  While coal (and then other fossil fuels) used to be the only real energy source, that is less true today than ever before, and many forces are directing us to move towards renewables.

And since buildings use about 40% of our energy, perhaps readers here can do whatever we can to support renewables, support carbon pricing legislation and help educate people on peak oil, climate change, and yes: continue to be disruptive of efforts to increase production of fossil fuels.

I agree that the paradox may be breaking down for now. I reserve judgement for the longer term.

Perhaps too fine a point, but most of the reasons you posit are economic which are having an affect, but are short-term actors. The larger question is whether we believe the the meme of "growth" (expansion, manifest destiny, GDP must rise ) has been replaced by a different meme (sustainability, stableness, flat GDP.) I don't think it has yet. (I haven't given up hope that it will, but I'm beginning to think that it is just hope and not inevitable.) The resources saved by increased efficiency may no longer fuel larger homes (McMansions have become so passé) but they will be consumed in some other way—new iPad, larger backyard BBQ, extra vacation requiring one more airplane flight. This doesn't change the validity of striving for hyper efficient homes, but it may change the underlying reason for doing so. IMO, it's not going to change the relentless increase in energy consumption, but it may buy us enough time to find a none carbon alternative before the system breaks down entirely.

Great discussion.

If we’re talking about the relevance of Jevons in light of our ability to move the needle on climate change, we ought to consider a global context. You might be able to make the case that “Jevons is dead” (or at least dying) IN THE AMERICAN RESIDENTIAL MARKET. You guys (and folks like us) are trying to make sure of it. In terms of electricity consumption, that’s about 30% of 23% of global usage… or 7% of the global energy total. I suspect that Jevons is hard at work on a big chunk of the remainder of our global energy use though.

For folks really into this, you might like to pick up The Jevons Paradox and the Myth of Resource Efficiency Improvements. We did a post on it a while back:

It's not dead, the paradigm has changed. It's the end product cost that drives increased production.

If a hamburger takes $2 worth of energy you might only sell ten a day.

If the efficiency improvement means the hamburger costs $1 worth of energy and you sell 30 a day, you are selling 3x the hamburgers and using 1.5 time the energy.

The cost of energy must RISE with efficiency improvements if the goal is reduced consumption.

Search PIGOU CLUB for a great Harvard essay on this.

The Jevons Paradox is a long term marcoeconomic phenomenon that can't be explained away by circumstantial observations.

Tom's point that we can substitute in renewal sources is hopeful, but doesn't really challenge the underlying axiom that efficiency drives increases in consumption.

As Bertoli writes, it's important to remember that Jevons paradox applies exclusively to the production sphere: the world of the so-called ‘consumer’ behaves very differently.

In the consumer sphere, energy consumption is not affected by efficiency improvements, unless there are price controls in place to prevent the Piggy Principle.

Alex -- 

To be clear, the rebound effect is the more general case which says only that efficiency gains driving a decrease in the use of a resource are offset by increase in use due to lower price.  Jevons' finding was that the offset was significant enough to cause a net increase in use.  Jevons is an edge case.

I am not sure about the distinction between production and "consumer".  Production is presumably driven by consumption (or at least it was when I was in college).  The rebound effect certainly occurs in simple consumer cases -- people with high-mileage vehicles tend to drive more, etc.  I have a Prius, whose actual mileage is 2x that of the average mileage of the US auto fleet (at least as of 2009).  

The application of this family of economic phenomenae would predict that a doubling of mileage would potentially result in both increased driving, as well as an overall shift from low-mileage cars to higher-mileage ones.  It's possible that both occurred ... or would have if all other things were the same ("ceteris paribus" in econ speak).

But all other things never are the same.

Priuses are not SUVs or BMWs.  They cost a little more than an otherwise equivalent car, for example.

I don't think Peter's assertion is that the rebound effect is gone.  Indeed, if there is an overall increase in energy use then this would even be consistent with Jevons.  My point, and I think Peter's, is that enough other systematic things now underway in our economy and society are not the same.  

If so, it's arguable that this particular case (medium-term, like say the next decade) is one where Jevons is not a useful economic principle as a predictor.  Jevons and rebound have been used to make assertions including that attempts at efficiency or conservation are actually counter-productive in the broader goal of reducing greenhouse gas production.  I would argue that this kind of argument is counter-productive, at least in the medium-term.

So (not speaking for the author of this post), I would agree with the simple assertion that Jevons and rebound may be alive and well.  But for this particular phase of our battle with climate change and energy security, there are many confounding issues that would tend to interact with whatever "pure" demand response curve an economist might draw from their predictions.  Wars and stuff, to name just one.

We economists (which I might be able to claim as a result of getting a degree in the early 1980s) tend to oversimplify and extract or normalize-out confounding variables.  But if economists would start acting more like the sociologists we should be (as opposed to the scientists we claim to be), we would realize that as a predictive tool, Jevons and rebound are not of much use in the discussion of how to address climate change.

And finally, as with all such discussions, I assert that anything I say is just an hypothesis.  I have not done the rigorous empirical study that would be needed to validate the hypothesis.  That said, neither has anyone else, because we can only know the truth after the fact.  I prefer to take the course of action that seems more prudent, which is to proceed with efficiency measures.  But that's just me.

An interesting discussion here, but I think the framing is somewhat misdirected. Troast correctly identify some factors that will reduce demand, but none of them really owe to efficiency, or apply to Jevons' observation at all.

The fundamental assertion of the Jevons Paradox seems inarguable. Nearly the entire history of man's energy consumption shows that higher efficiency (or better, more utility) has a nasty way of encouraging demand.

The pertinent question is: What causes demand growth, and what can be done about it?

Our transitions from wood to coal to whale oil to petroleum also led to growth, because the new fuels were denser, more convenient, more versatile, and/or more abundant. Similarly, the creation of fiat currencies and "financial innovation" led to economic growth and consequent fuel demand, by providing greater liquidity and ways to claim future income in the present.

Most importantly, growth has been driven by an ever-increasing population.

Teasing the effects of the Jevons Paradox out of the overall growth picture would be difficult. And if one did, the result wouldn't be particularly useful.

Indeed a lot has changed since the Industrial Revolution. For one thing, we no longer have an abundance of cheap fuel to draw upon. As depletion gradually overwhelms new production and supply constricts, demand will be forcibly reduced as the result of higher prices and outright unavailability. Efficiency can dampen the effects of this and help minimize the pain, but I suspect the paradoxical effects of higher efficiency will be overwhelmed by the sheer magnitude of declining primary energy supply. In any case, Jevons' concept would be merely rendered irrelevant, not nullified.

No tree grows to the sky, and America has apparently reached some key limits to its growth--namely, being badly overextended on everything from energy consumption to debt. Declines in homebuilding, lower general economic activity, and a gradual trend toward smaller and more efficient things are the natural result of these macro factors, not greater efficiency.

Permanent recession seems unavoidable as primary energy supply continues to tighten, since efficiency gains are far too slow and incremental to outweigh it.

The Jevons Paradox will not only be irrelevant in the future, it will be inverted: Declining fuel supply will lead to greater efficiency, and lower demand.

Integrating this concept will undoubtedly be a slow and reactionary process, since the belief that trees can grow to the sky seems embedded in our current economic theory. As Gregor Macdonald observed in Jevons and the Six Day Car Crash, we simply find it difficult to correctly perceive long duration problems and slow changes.

I have not seen one model that convinced me that greater efficiency will permit demand growth to continue in the face of declining supply. I have seen, however, a great deal of fundamental misunderstanding about the scale and time-to-market issues that bedevil our attempts at addressing the developing energy crisis. Other thoughts of mine on that subject:
195 Californias or 74 Texases to Replace Offshore Oil
’Peak Demand,’ Yes, But Not the Nice Kind

No-Cost Marketing Benchmark Analysis

Admit it...keeping up with digital marketing can be challenging. Let us help.