On the rebound
Nearly everyone agrees now that global warming is real, and I’m sure we all agree that the challenge we face in preventing catastrophic climate change is incredibly daunting. Brace yourself, though, because I bear some bad news: it turns out that the challenge might be even greater than is commonly acknowledged by policymakers, and that energy efficiency might not be the holy grail that some claim it could be.
The added worries about climate change result from ‘rebound effects’, behavioural or systematic responses to new technologies or energy-efficiency measures that offset some – or, in extreme cases, all – of the environmental benefits. Rebound effects can either be direct (as when energy-efficient light bulbs make lighting cheaper, so people leave lights on or increase levels of illumination) or indirect (as when money saved through energy-efficiency measures is used to purchase goods and services, such as a holiday abroad). The effects are never intended and, to date, rarely accounted for in official policy, though they can be significant: a rebound effect of 30 per cent, for example, would mean that only 70 per cent of the anticipated/calculated efficiency savings would be achieved.
If you’ve never heard of rebound effects you’re far from alone. I ask Dr Steve Sorrell, Senior Lecturer in Science and Technology Policy Research at the University of Sussex and lead author of several papers on rebound effects, why this is the case, and his response is frank: “It’s been known to specialists for quite a while, though it’s always been a subject of great controversy. One of the reasons for the controversy is it is quite difficult to calculate the size of these effects, so people can be quite selective with the evidence and point to some studies that suggest very small effects and others can point to studies that suggest very large effects, and so on. That situation is changing in that there’s more and more evidence and more and more studies, and it is better understood now. I wouldn’t say that we’ve got to a consensus yet, but it is improving.
“Another reason is that it’s quite an uncomfortable message for policymakers and people concerned about climate change. To suggest that energy efficiency is not going to deliver the energy savings and emission savings we are expecting, and hence to suggest that a lot of the models’ projections and scenarios are wrong or overly optimistic, is not a comfortable message nor one that people are very open to. You do get quite defensive reactions to it, and that’s particularly evident in the United States, where advocates of energy efficiency, like Amory Lovins, for example, have been quite dismissive about the impact of rebound effects.”
Indeed, as far back as 1988, Lovins – energy efficiency guru and proponent of a ‘negawatt’ revolution – said: “The concept of a nontrivial rebound effect… is without basis in either theory or experience. It is, I believe, now widely accepted to be a fallacy whose tedious repetition ill serves rational discourse and sound public policy.” His tune has not changed much in the intervening years, and as recently as 2011, he wrote in a letter to The New Yorker magazine that ‘take-backs’ (another term for rebound effects) ‘are usually between zero and a few per cent, rarely ten to thirty per cent, and never more than a hundred per cent… Rebound effects are small in energy-using devices for three reasons: no matter how efficient your house or washing machine becomes, you won’t heat your house to sauna temperatures, or rewash clean clothes; you can’t find an efficient appliance’s savings in your un-itemized electric bill; and most devices have modest energy costs, so even big savings look unimportant. Respending a saved energy dollar does indirectly use energy, but, from 1986 to 2007, only six to nine cents’ worth on average, and no respent dollar can buy more energy than the hundred-per-cent energy in that original dollar saved.’
Others, though, increasingly calculate that rebound effects aren’t so trivial, and a growing number of studies are agreeing about the extent of certain rebound effects. Sorrell tells me that, at least when it comes to fuel-efficient cars, there’s growing consensus that in OECD countries, “if you make cars more fuel-efficient, then they become cheaper to run, and that does encourage more driving – not so much that it offsets the energy savings, but most studies suggest something to the effect of 15 or 20 per cent or so”. In developing economies, the effect is likely to be much higher, as cheaper, more fuel-efficient cars place car ownership within the reach of more and more people, putting more and more cars on the road, though Sorrell cautions that “there are far fewer studies there and the data is harder to come by”.
Even in developed economies, it gets more complicated when you calculate both direct and indirect rebound effects, especially if you’re looking at measures that result from both efficiency (technological improvement) and sufficiency (behavioural change) measures – all of which take place in complex economic systems. Indeed, studies into rebound effects from efficiencies in transport and utilities in Sweden in 2007 and Japan in 2008 came to staggeringly different conclusions, with the former saying rebounds offset savings by 120-175 per cent (meaning efficiency measures ‘backfired’), while the latter suggested rebound effects of just 12-38 per cent!
That being said, even the International Panel on Climate Change now says that rebound effects ‘cannot be ignored’, with the Working Group III on Mitigation of Climate Change noting in 2014: ‘Rebound effects need to be taken into account in interpreting these findings of environmental effectiveness of energy efficiency regulations… direct rebound effects… are commonly found to be in the range of 10 – 30 per cent in various sectors in developed countries, and higher in developing countries.’
The level of rebound also varies widely depending on the measure it relates to. A 2011 study that Sorrell contributed to, ‘Exploring rebound and backfire effects in UK households’, examined three sufficiency measures: reducing internal temperatures by 1°C, reducing food expenditure by a third through eliminating food waste; and walking or cycling instead of driving on trips shorter than two miles. The study, which calculated impacts of both re-spending the avoided expenditure as well as saving it, found that the rebound effect would likely be seven per cent for lowering the thermostat, 25 per cent for avoiding car travel, and 51 per cent for reducing food waste! (It does depend, of course, what people invest in, and the overall rebound effects of the three sustainability actions in combination could actually be negative, if avoided expenditure is put into ‘green’ investments, but could be as high as 515 per cent, if savings are re-invested in greenhouse gas intensive categories.)
This is bound to be dispiriting for our waste and resources industry, where increases in awareness of food waste and corresponding decreases in food waste itself are one of our great success stories. I ask Sorrell what causes the large rebound in this area, and he explains that the effect is bigger there than with the other two sufficiency measures, because food has lower carbon intensity than the fuel saved through turning the heating down or not driving: “By avoiding throwing away food, you’re saving money, and then spending that money on other things. And some of the things that you’re spending money on are more carbon-intensive than food is. On the basis of our analysis of household spending patterns, we estimate that some of the money will be spent on gas, some of it on electricity, some on vehicle fuel, some on aviation – they’re all very carbon-intensive things to be spending money on.” He cautions, though, that the study has limitations in that it used an average carbon intensity for food, and “different foods are more carbon intense than others”, and points out that the large potential rebound “is not an argument for not driving food waste reduction, which has all sorts of resource implications”.
Indeed, Sorrell and others raising the issue of rebound effects are at pains to insist that they are not arguing against energy-efficiency measures, which will in most cases still deliver savings, and which have benefits such as improving living standards: “It’s not an argument to change energy-efficiency policy, it’s an argument to be more careful with the accounting, and to recognise this problem rather than ignoring it.” Indeed, even studies that suggest that energy-efficiency measures could actually increase emissions in developing economies still advocate energy efficiency for other reasons – as with a recent Breakthrough Institute report, ‘Energy Efficiency Backfire in Emerging Economies’, which concludes: ‘Energy efficiency brings many social benefits, reducing the cost of energy services, contributing to the achievement of human development goals, and opening new frontiers for economic production and consumption.’
Asked what policymakers need to do, given the increasing consensus that rebound effects are real, Sorrell responds: “You need to start taking them seriously rather than ignoring them… In most areas… it’s generally assumed that the rebound is zero, and then you end up with overoptimistic estimates of energy savings, which will create a problem going forward. And it does reinforce the arguments in favour of putting a price on carbon emissions or putting a cap on carbon emissions.”
In instances where people save emissions in their own country through energy efficiency and then contribute to emissions in other countries by using the money saved to buy goods produced abroad, Sorrell adds that ‘border carbon adjustments’ could help protect efficiency savings: “There is growing attention being paid to consumption-based emissions, which is trying to follow the whole emissions supply chain of the goods that we are buying, right back to where they’re made. Whilst UK emissions have been reducing within our borders, until recently, the consumption-based emissions have been going up… It points to the long-term importance of a global agreement on emissions because partial agreements will only cover certain sectors or certain countries, and that allows a form of leakage – you don’t emit here, you emit there instead.
“It could also be dealt with by putting a levy on the goods that are imported in proportion to the carbon emissions that have been created in producing those goods, which levels the playing field if there is a carbon tax on domestic energy.”
Others have suggested that rebound effects reinforce the desperate need to decarbonise the energy sector as quickly as possible (though even without rebound effects, the IPCC suggests we must be producing zero emissions by the end of the century), and also to raise consumer awareness of putting the money they’ve saved through energy efficiency into low-carbon investments.
That latter idea could be a rather tricky one to communicate, though – how much of a positive response do you imagine you’d get from a campaign that ran along the lines of: ‘Avoid food waste, save money, and then make sure you put that money into low-carbon investments or activities’? Because of this potential confusion Sorrell suggests that it’s more important at this stage to engage with policymakers: “I’m concerned that raising awareness amongst consumers on this issue is just going to confuse them and be counterproductive.”
He notes, though, that most of the research to date on rebound effects has been by economists looking at price mechanisms, but “there could be all sorts of other psychological things going on that are less-well studied”, such as moral licensing, by which “if you’re being good environmentally in some area, like you’re doing your recycling, then you might feel that justifies you in doing something that is environmentally damaging in another area, like flying abroad for your holiday”. But that’s a topic for a whole other article…