Why efficiency won't save us
Rebound effects, explained - and updated with the latest science! Conclusion: we need to make choices about the world we want.
Energy efficiency. “The ratio of useful outputs to energy inputs for a system, like a motor, tool, process, firm or even an entire economy” (Brockway et al 2021). It’s been a mainstay of environmental policy since the 1973 oil crisis. It is super important. Greater efficiency can reduce our use of fossil fuels and thus curb emissions directly. It can also reduce the energy demand of our electric devices and systems, which will allow us to go farther on our still highly insufficient sources of renewable energy, which in turn will reduce our reliance on fossil fuelled electricity. Great 👍.
But efficiency has a dirty - albeit open - secret.
“Large rapid absolute reductions of resource use and GHG emissions cannot be achieved (through technological improvements)” (Haberl et al, 2020)
Improvements in energy efficiency do not actually translate into the expected reduction of emissions. Why is that?
The answer lies in good part with a phenomenon called the rebound effect, or the process whereby (some of) the gains are gobbled up by increased consumption afterwards - whereby energy use ‘rebounds’, so to say.
In other words: as our technology or logistics improve, it becomes cheaper or more powerful, and we start doing more things, consuming more, and growing our economy. And because we haven’t been able to undo the laws of thermodynamics, all that growth means more energy demand and, so far, most of that demand has been met by burning fossil fuels.
OK, so let’s review the latest evidence for the what, how, and why of the energy rebound effect. It’s important that we come to grips with it and that we understand its consequences for policy. In particular, it opens up the door to that space where we talk about what an actually sustainable society could look like.
What is the rebound effect?
Most likely, this is not the first time you heard about the rebound effect. At the very least you are likely familiar with the “direct rebound”. That happens when we use something more because it has become cheaper. The prime example is your neighbour's house seemingly on fire with LED lamps during the holidays.
Another form is the indirect rebound. That happens when efficiency frees up money for other things. So, a cheaper cost of heating means being endowed with some extra fundalation for a vacation trip.
These are both examples on individual (or household) level rebounds, but similar effects can be seen at aggregate level too. I won’t go into those, because I want to dwell a bit longer on a more complex type of rebound at the collective level: the “economy-wide rebound”. This is:
the percentage of the expected economy-wide energy savings, as estimated from a counterfactual scenario where none of these adjustments occur. (Brockway et al, Renewable and Sustainable Energy reviews, 2021)
Let’s start with an example of one study.
Koesler et al. found that energy efficiency improvements in German industry improve the competitiveness of German producers, encourage increased exports and thereby reduce production and energy use in other regions. At the same time, energy efficiency improvements increase German GDP and wages, increase domestic demand and imports, and thereby increase production and energy use in other regions. The net result is that (in this case) the global rebound effect is smaller than the rebound effect within Germany alone. (Brockway et al, 2021)
In other words: when you start looking at system-wide ripple effects of certain improvements, it gets very complicated very quickly. If calculating direct and indirect rebound effects is more like a controlled lab experiment, with the economy-wide rebound effect, you're in the deep end, trying to model the real world as best as possible, where everything influences everything.
So, all that to say: it's a bit of a crapshoot. There are a lot of estimates and assumptions involved. As a consequence, calculations can very quite a bit (and the rebound effect has been controversial because of it).
Still, after reviewing dozens of studies from the past 15 years or so, Paul Brockway and his colleagues detect a clear pattern: the economy-wide rebound is real and it is large. At least half of efficiency gains are lost to it, they reckon.
And the thing is: we're not really taking this into account. The authors also combed through the predominant forecasting studies, the famous Integrated Assessment Models, or IAMs, that also undergirds much of the IPCC reports (read more about IAMs here, it you'd like a primer). It turns out they use a much cruder statistical trick to estimate rebound and come up consistently with lower losses. That means we still might be in for a nasty surprise.
A failure to achieve the anticipated structural break in the rate of growth of global energy demand could have important consequences. (Brockway et al)
Is it the economy stupid?
If we can say with some confidence that systemic rebound effects are considerable, then the next step is to understand why this is happening. Sociologist Ray Galvin, writing in Energy Policy (2020), is surprised to find that we don’t actually ask that question very much - or at least not beyond superficial reasons.
Like, on the individual, household or firm level, the traditional explanations are pretty straightforward: a simple matter of economics (it’s only natural to reinvest the gains) or a matter of psychology (people are creatures of habit and conformity, if they’re not consciously willing to keep the gains, it will slip right through their inattentive fingers). But while on the surface this is not wrong, it’s just that: on the surface. These habits and these decisions do not take place in a vacuum.
If we want to benefit from technological improvements to reduce our energy demand and emissions then we need to understand that context and reshape it accordingly.
Case study: the all-American car
To understand how efficiency gains actually disappear, Galvin takes a look at the American car. After the oil crisis, and under government mandate, the industry produced spectacular enhancements in fuel economy. In 1987 cars were 70% more efficient cars than in 1975. But then car makers started increasing the weight and the horsepower of the cars, obviously undoing some of the gains made. How much? Well, based on official test results over the last 20 years or so, Galvin sat down to do some calculating. He wound up with a 28% reduction for highway driving and 58% for urban (start and stop).
Whhyyyy?? 😭, Gavin proceeds to ask.
Well, basically because car manufacturers made it happen and were allowed to make it happen. This wasn’t an automatic process — thermodynamic rebound effects are not thermodynamic law. It took a lot of effort! People in companies have lobbied for laxer rules and standards (limiting the absolute fuel economy goal, making fuel economy requirements dependent on the size of the car); for tradeable efficiency credits (making it possible to run a little less hard to meet the next standard); and for a shift to “voluntary” standards (...). On the consumer side, meanwhile, they have promoted and normalized the desire for more horsepower.
When American car makers are like “When you wanted bigger cars, we built bigger cars. When you wanted smaller cars, we built bigger cars”.
What to do?
In an early explainer about the rebound effect, energy journalist David Roberts’ concluded half-jokingly that the rebound effect doesn’t matter - we still need to keep making our technology and processes less energy demanding anyway.
So… yes and no.
Yes - we do need to keep making efforts to improve efficiency - at the very least as long as energy comes from burning fossil fuels.
No - rebound effects do matter, and our understanding of them should have consequences.
This is why the full quote by Helmut Haberl and his colleagues from the introduction actually reads as follows:
Large rapid absolute reductions of resource use and GHG emissions cannot be achieved (through technological improvements), (so in addition we need) sufficiency-oriented strategies and strict enforcement of absolute reduction targets. (Haberl et al, 2020)
This is where we get in Bigger Swing territory. But: this newsletter has been long (and technical) enough as it is (you made it 🙌!), so TBC next week. I’ll explain what these authors mean by sufficiency, what would be some examples and why even with these strategies we need…
Take care for now!
Marten
My conclusion: pursuing energy efficiency always makes sense, but we need to be well aware of the potential rebound effects and address these where possible through behavioural changes. Do not install that extra LED, just because you can ..., do not buy that SUV because you can afford it ...