May 28, 2013 — Over the past decade, energy efficiency has come to be seen as a fast, cheap and even profitable way to reduce greenhouse gas emissions. Increasing the efficiency of buildings, vehicles, appliances and industry plays “a key role” in climate mitigation scenarios created by the U.N. Intergovernmental Panel on Climate Change. As governments face political opposition to costly climate policy measures, energy efficiency offers a tantalizing promise of a win-win for both the environment and the economy.
Detailed reports from energy analysts and consulting groups — including the International Energy Agency, McKinsey and Company, and the Rocky Mountain Institute — lend legitimacy to bullish efficiency prospects. IEA’s latest World Energy Outlook touts energy efficiency’s potential to “realize huge gains for energy security, economic growth, and the environment.” It claims an $11.8 trillion global investment in efficiency through 2035 would yield $18 trillion in higher economic output while allowing global carbon dioxide emissions to peak by 2020. An influential 2009 report from McKinsey analysts argued that the U.S. could cut its annual energy use by 23 percent through 2020, abating one-sixth of U.S. carbon emissions and yielding a net savings of roughly $700 billion. And Reinventing Fire, an efficiency-centered roadmap for the U.S. authored by Amory Lovins and published by RMI in 2012, promises to cut projected energy consumption 40 percent while delivering $5 trillion in net energy savings by mid-century.
But behind the scenes a growing number of economists and energy analysts are challenging the assumptions and methods behind these studies. None of them argues against seizing truly cost-effective energy efficiency opportunities. Rather, they caution against overestimating their energy and carbon savings potential. As such, it is time to rethink the privileged place efficiency has taken in the climate and energy strategies of national governments and international agencies.
Historical Perspective
Between 1860 and 1990, the global economy “decarbonized” — reduced how much carbon is required to produce a given unit of economic output — at a rate of 1.3 percent per year. Over the last decade, as coal and heavy industry fueled rapid growth in China and other nations, decarbonization slowed to 0.4 percent per year.
If the world hopes to stabilize atmospheric carbon emissions at 450 parts per million, the target for preventing global temperatures from rising by more than 2 degrees Celsius above preindustrial levels, the decarbonization rate must accelerate to at least 4 percent annually — a 10-fold increase above the last decade’s rate, or three times the 1860–1990 rate, which is an enormous challenge.
The rapid global efficiency gains posited by RMI, McKinsey or IEA are unprecedented.
There are two ways to accelerate decarbonization. The first is to reduce the energy intensity of the economy — how much energy is required to produce a unit of economic activity. Over the past century, energy intensity has improved worldwide by about 1 percent per year. The second is to decrease the carbon intensity of the energy supply — carbon dioxide per unit of energy produced. This has improved more modestly, at about 0.3 percent per year in the 20th century, slowing to about zero over the past decade.
The rapid global efficiency gains posited by RMI, McKinsey or IEA are unprecedented. IEA places more than half of the burden of decarbonization on improving energy intensity, about 2.4 percentage points of an overall 4.2 percent annual decarbonization rate. To achieve that target, global energy intensity would need to decline 2.4 times faster than it has over the past century.
Only four of the 26 developed (OECD) nations we studied — Ireland, the U.K., the U.S. and Poland — achieved energy intensity declines of 2 percent per year or greater between 1971 and 2006, a period that began with energy price shocks and strong measures by nations to increase energy efficiency. Moreover, much of the energy intensity decline in these countries was the result of broad economic changes unrelated to energy efficiency policies.
In the U.S., California is often held up as the real-world proof that efficiency can be radically increased at a profit because per capita electricity use in California has fallen flat since the early 1970s, while energy demand in the rest of the U.S. continued to rise. In fact, a variety of factors unrelated to the state’s efficiency policies explain more than three-quarters of the divergence, according to a study by Anant Sudarshan and James Sweeney of Stanford University’s Precourt Energy Efficiency Center.
A 2007 study of energy intensity by Soham Baksi and Christopher Green reinforces that finding. After accounting for ongoing changes in the share of global economic activity derived from manufacturing, agriculture and services, the two economists found that simply increasing the global rate of energy intensity decline from the historic rate of 1 percent to 1.25 percent would require increasing efficiency within the residential and commercial sectors sevenfold and efficiency in the transportation and industrial sectors fivefold between 1990 and 2100.
So not only are such efficiency gains unprecedented, given that a substantial share of historical energy intensity declines are due to factors outside the domain of energy efficiency policy, they also appear highly unrealistic.
Double Counting
In order to claim that energy efficiency can dramatically reduce emissions, both IEA and IPCC assume that energy intensity will decline faster in the future than it has in the past. In the 2012 World Energy Outlook, IEA assumes in its core reference scenario that energy intensity gains will spontaneously — that is, without any additional policy intervention — increase to 1.9 percent annually through 2035, meaning that some of the fastest national rates observed over the last 40 years will all of a sudden happen globally. This allows IEA to assume away about two-thirds of its projected increase in energy intensity relative to the long-term global trend and focus its detailed plans only on that final 0.5 percentage point increase.
IPCC has made similarly heroic assumptions about energy efficiency, provoking a reaction from leading energy and climate scientists. In a 2008 article for Nature, Roger Pielke Jr., Tom Wigley and Christopher Green faulted IPCC for assuming in its business-as-usual scenarios that two-thirds or more of the overall decarbonization rate required to stabilize greenhouse gases this century will occur automatically, with energy intensity improvements constituting the largest portion of these projected gains.
Making matters worse, even after IPCC and IEA assume higher-than-normal energy intensity declines in their scenarios, their modelers went on to assume there are substantial additional efficiency opportunities left waiting to be unlocked — at a profit — by the right policy actions.
According to RMI, McKinsey, IEA and others, trillions of dollars in untapped efficiency gains await the savvy businessman or policy maker. In practice, private actors have been perfectly happy to leave these trillion-dollar-bills sitting on the sidewalk.
Unfortunately, few models rigorously account for what part of the finite pool of cost-effective efficiency opportunities accounts for the big gains baked into baseline scenarios — such as swapping old incandescent bulbs for efficient new fluorescents or LEDs, an efficiency opportunity likely to be taken up under business-as-usual scenarios — and which are left over to be captured by policy, raising the risk of substantial double counting.
Hidden Costs
According to RMI, McKinsey, IEA and others, trillions of dollars in untapped efficiency gains await the savvy businessman or policy maker. In practice, private actors have been perfectly happy to leave these trillion-dollar-bills sitting on the sidewalk.
Obstacles to greater energy efficiency include split incentives (the people paying for efficiency are often not the same people who receive the benefits) and imperfect information (people often do not know that greater efficiency can pay for itself). And economists warn that there are often hidden challenges associated with overcoming these obstacles.
First are transaction and implementation costs, such as rebates, policy implementation and hired consultants, which can raise the cost of efficiency upgrades 40 percent or more, and which are ignored by the efficiency strategies published by IEA, McKinsey and RMI. In a recent paper, economists Hunt Allcott and Michael Greenstone found that the literature “on the magnitude of profitable unexploited energy efficiency investments … frequently does not meet modern standards for credibly estimating the net present value of energy cost savings and often leaves other benefits and costs unmeasured.”
Another challenge is the assumed rate at which consumers and firms discount future earnings relative to current wealth — what economists call the “discount rate.” McKinsey assumes an across-the-board discount rate of 7 percent per year when determining which investments are profitable. RMI applies a range of rate-of-return hurdles necessary to overcome discount rates in various sectors, ranging from 5 to 33 percent.
But due to opportunity costs, corporate executives are often looking for returns on investments in the range of 20 percent, and individual consumers even higher still. Likewise, efficiency investment today means less money available to invest in other opportunities tomorrow, which can increase the hurdle rate by 10 percentage points.
When more realistic numbers are applied, the amount of cost-effective efficiency measures dramatically shrinks — by 21 percent with the industry-appropriate discount rate of 20 percent, and by 43 percent with a discount rate consistent with the two- to three-year minimum payback period typically needed to entice household efficiency investments.
Beware Rebound
A third reason to question potential efficiency gains is the rebound effect, which posits that efficiency improvements that lower the cost of energy services trigger an increase in energy demand that can erode much of the expected energy savings and climate benefits.
This is because to the extent that energy efficiency measures lower the effective price of energy services, consumers and firms are likely to demand more of them. And when consumers save money through energy efficiency measures, they will likely spend some of those savings on goods or services that require still more energy. Finally, getting more economic activity out of each unit of energy drives economic growth, which further expands energy demand.
Collectively, these economic mechanisms can erode much — and in some situations all — of the reduction in energy consumption predicted by engineering-level analyses. And indeed, economists have documented myriad examples of the rebound effect. A 2007 review commissioned by the UK Energy Research Centre, drawing from more than 500 studies on the topic, found typical rebound levels ranging from 10 percent to as high as 80 percent, depending on the sector in question.
Surprisingly, most efficiency scenarios disregard the possibility of a significant rebound effect. McKinsey, for instance, completely disregards rebound effects and, in Reinventing Fire, RMI dismisses the possibility of significant economywide rebound, while the IEA’s 2012 World Energy Outlook assumes that rebound effects erode only 9 percent of energy savings.
Perhaps most importantly, in emerging economies where energy demand is growing most rapidly and consumers are just starting to acquire many modern conveniences such as air conditioning, personal transportation or even reliable lighting, rebound effects are much larger and can even lead to “backfire,” or a net increase in energy consumption following energy efficiency improvements.
Fortunately, there is still another big lever left to drive global decarbonization: accelerating the transition to low- and zero-carbon energy sources, including renewables such as wind and solar as well as nuclear energy.
To be clear, rebound effects, particularly in emerging economies, mean consumers and firms are using energy efficiency to enhance their economic welfare, getting more energy services out of the same or less overall energy use. That’s a fundamentally good thing. That said, the developing world is projected to account for virtually all energy demand growth in the coming decades. Accurately estimating how much rebound effects erode lasting energy savings is essential to depict the contribution of efficiency to long-term climate and energy strategies.
A Way Forward
Despite the bullish efficiency strategies promoted by IEA and respected consultants such as McKinsey and RMI, economists and energy analysts provide plenty of reason to be cautious about overestimating the potential contribution of profitable efficiency opportunities to global climate mitigation. After avoiding double counting and taking full account of the hidden costs required to unlock efficiency opportunities, the quantity of “low-hanging efficiency fruit” shrinks dramatically. Meanwhile, what truly profitable efficiency opportunities remain will trigger rebound effects, which further erode the delivered long-term energy savings.
If policy makers cannot count on energy efficiency to deliver the lion’s share of the roughly 4 percent per year global decarbonization rate needed to avoid dangerous climate change, are climate mitigation efforts sunk? Fortunately, there is still another big lever left to drive global decarbonization: accelerating the transition to low- and zero-carbon energy sources, including renewables such as wind and solar as well as nuclear energy.
In fact, history suggests reason for hope: Sweden and France each sustained greater than 4 percent annual improvements to the carbon intensity of their energy supplies for more than a decade (from 1974 to 1991 in Sweden and 1976 to 1988 in France) by deploying large amounts of nuclear power, which at the time, was the only zero-carbon energy source ready for global prime time. Iceland achieved similar rates from 1971 to 1985 by tapping the island’s localized geothermal resources.
Today, the world has an expanded suite of low-carbon power sources at its disposal, from increasingly competitive wind and solar energy technologies to safer new nuclear power plants. A set of challenges must be overcome to ensure that each of these technologies can truly scale to meet the needs of an energy-hungry planet, and new technologies must be readied for market, particularly in the transportation sector. National governments and industry leaders must now invest the necessary resources in advanced energy innovation and deployment.
Global energy consumption will likely more than double by mid-century as the population expands towards 10 billion people and billions more global citizens are lifted out of poverty. If we are to stabilize carbon emissions at or below 450 parts per million, virtually all of that new energy must be clean. If human civilization can reduce energy intensity faster than it has over the last century, then this effort will be made all the easier. But if it turns out that efficiency’s potential has been exaggerated, then we will be glad that strong efforts were made to move rapidly towards cleaner sources of energy. 
Editor’s note: The views expressed here are those of the author and not necessarily of Ensia. We present them to further discussion around important topics. We encourage you to respond with a comment below, following our commenting guidelines, which can be found here. In addition, you might consider submitting a Voices piece of your own. See Ensia’s “Contact” page for submission guidelines.
Ensia shares solutions-focused stories free of charge through our online magazine and partner media. That means audiences around the world have ready access to stories that can — and do — help them shape a better future. If you value our work, please show your support today.
Yes, I'll support Ensia!
To me, the bottom line here is that large energy efficiency gains are physically possible, and are probably necessary steps to addressing our energy and climate challenges. But, as this essay and partner report show, they are not going to be as easy or as automatic as many have suggested before.
Okay, that's good to know. Hard work is needed here. And we may need to work harder on decarbonizing energy sources in parallel with this effort.
But I would love to understand why certain efficiency efforts have worked better than others, and what we can learn from them? And what policy and market instruments could help accelerate efficiency gains?
I'd really hate to see us give up on possible efficiency gains. What would it take to realize them? To me, that's the big question.
Thanks for the thoughtful article.
One has to wonder how often and how long a consultant can be wrong before people stop paying attention and stop granting the "energy guru" status.
There have been only a few times in the past 150 years when energy use actually declined, and each of those times was associated with terribly economic consequences. Recessions effectively reduce energy demand growth, but that is a strategy that only a misanthrope can love.
We know how to produce safe nuclear power plants; we've been doing it since before Lovins published his "anything but nuclear" article.
I often remind people of a quote from a Lovins appearance on Democracy Now! in July 2008:
"You know, I’ve worked for major oil companies for about thirty-five years, and they understand how expensive it is to drill for oil."
I am pretty sure that Lovins is clever enough to know that energy efficiency will never work as well as he claims. I am also quite confident that his clients in the oil and gas industry are happy as long as his exaggerated efficiency predictions discourage people from taking the nuclear-energy-enabled decarbonization path that the French and the Swedes proved works fine and lasts a long time.
After all, countries that have built a lot of nuclear power stations have seen a permanent drop in their fossil fuel demand.
I want to pick up on one key point in particular:
"To be clear, rebound effects, particularly in emerging economies, mean consumers and firms are using energy efficiency to enhance their economic welfare, getting more energy services out of the same or less overall energy use. That’s a fundamentally good thing. That said, the developing world is projected to account for virtually all energy demand growth in the coming decades. ..."
Reducing poverty is an intrinsically good thing. Regarding the concerns with carbon emissions, it also yields some collateral benefits, even if there is some rebound effect from one form of energy consumption to others.
Economists and demographers have long recognized that increasing prosperity is associated with a "demographic transition" to significantly lower fertility rates. So even as per-capita consumption grows, there also will be, over time, fewer 'capita' and thus a damping effect on gross demand.
Indeed, fertility rates in most industrialized nations now are below replacement level -- pointing toward declining populations in the absence of increased immigration. Slower or negative population growth tends too to have a parallel effect on GDP growth as a result of fewer workers to produce output. Less economic product implies relatively less carbon emissions, even with technology constant.
Moreover, greater prosperity makes investment in and acquisition of alternatives to fossil fuels -- whether via efficiency or alternative energy sources -- more affordable.
I agree with Carol that we need to do 'all of the above'.
In reply to Jon's question, I'm afraid the answer might be higher energy costs. Because we can't seem to change peoples behavior regarding electricity consumption, we have embarked on an experiment called real time pricing. When the demand for electricity is the highest, your local utility charges the most for it, thereby avoiding (hopefully) the cost of building a new power plant.
First, I think this is a very regressive policy, and will harm the lower income people the most, and second, in the complicated world we live in, who has the time to watch a consumption meter when they get home from work-and sit in the dark when demand gets high. Not sure that will get us where we need to go either.
The ~500 billion tons of unnaturally emitted carbon now in air & seas will require extraordinary measures worldwide -- measures requiring energy, but which must combat changed world chemistry. We have years, not decades, to do this.
.
1) Certainly efficiency installed does not historically deliver promised value if it is not maintained. Installing a measure and performing proper M&V checks to verify maintained performance, is a waste of money. At the corporate level, the operational mentality may not include the proper maintenance measures in the financial model, following installation efficiency improvement countermeasures.
2) The negative effects of climate change on the economics of doing business as usual are not accounted for. That is assuming the EUI growth rate difference of both scenarios have a certain level of negative feedback on economic models. These are not well understood.....especially as we achieve higher levels of GHG concentrations. Uncertainty bands widen as we see exponential compounding effects. So there is a non-linear relationship that many bean counters cannot begin to comprehend.
3) The economics of externalized costs appear neglected in the computations. This accounting is difficult for many to fully model, as the extent of damage is deep and wide. These obviously have negative feedback effects on economies both locally and globally....and cannot be discounted or the computations are based on non-reality.
4) Article does not express the dynamics in terms of complex energy usage mathematical models and only considers first (or maybe second) level effects.....and in fact the equation likely involves levels of variables of many more degrees of complexity.
5) Decarbonization is trivialized, as only a reduction of our current feedstocks seem considered. The surface of the problem is scratched, and traditional RE sources of wind/solar/nuke solutions are mentioned. The analysis does not mention the embodied energy "costs" of each, compared to other less intensive technologies capable of delivering energy on a constant basis...with less embodied energy.
6) Migration away from combustion of fossils is not considered.....and it is the key element holding us back, in a stuck in the stone age mindset. This is an implementation/transition long term scenario.
7) Harnessing of Hydrogen as a new fuel feedstock....involving no carbon, is not even mentioned and is key to getting us OUT of the stone age mindset. I am speaking about reacting the hydrogen and not combusting it. This is our best long term sustainable bridge to until nuclear fusion, which will not poison our atmosphere further. The key to Hydrogen is that is enables RE integration and scalable storage for long term baseload energy delivery.
8) Energy efficiency measures implementation relying on government support for the "economics" of a project are a fallacious position, in which we cannot do anything well without incentives. This is a truly limited mentality, which keeps us stuck because we could not possibly know enough to do any bold improvements without their backing. I like the "build it and they will come" perspective.