The New York Times got the headline wrong in “The Problem With Energy Efficiency,” an October 8th op ed by Michael Shellenberger and Ted Nordhaus, but the authors are right that the rebounds in energy demand triggered by efficiency improvements are real, typically significant, and should force a careful rethinking of the role of energy efficiency in global climate mitigation efforts.

Rebound effects are only “a problem” for energy efficiency if you believe efficiency’s unalloyed goal is to cut energy consumption. But that’s hardly the case.

Think of increased efficiency as an improvement in “energy productivity,” and we would expect and even welcome “rebounds” in demand for energy, just as we do for labor productivity. (No one ever bemoaned a rebound in labor demand following productivity improving factory upgrades, for example!).

Improving the productivity at which we use energy resources simply means we are getting more energy services out of our resources than ever before. That makes energy even more valuable, and it makes perfect since that we would want to make even more use of this valuable resource.

Overall, improving energy productivity is thus great news, regardless of whether it cuts energy use, helps us get even more value out of the same amount of energy use, or some combination thereof. And that is exactly why taking rebound effects seriously should do nothing to undermine the business case for energy efficiency.

If you’ve been following the online kerfuffle over the “Problem” op ed last week and keeping score at home, you’ll know that I’m leveling an explicit critique here at the headline of the column (although the authors themselves did not argue efficiency was “a problem,” and to the contrary, lauded it’s contributions to expanding energy access and fueling global economic productivity and development).

But this is also a clear criticism of those who see efforts to take rebound seriously as a threat and try to dismiss the clear evidence for significant rebound effects at every turn.

Energy efficiency advocates should also take note: don’t build your case for efficiency by arguing its value is solely in reducing energy use. That’s just not how efficiency (aka productivity) works, nor should it be.

I’ve reviewed nearly 100 peer-reviewed and academic articles on the topic, and the consensus is clear: while rebound effects don’t undermine the case for efficiency, taking rebound seriously does force us to rethink the role of energy efficiency in confronting climate change.

The public debate over rebound effects is complicated by the fact that the magnitude of rebound varies from context to context, so it’s difficult to generalize about the scale of rebound, but I’ll try to provide the best summary I can here…

In rich, developed countries, where energy is plentiful and demand for energy services widely fulfilled, rebound effects most likely erode 20-70% of the original energy savings from various efficiency measures. (Even then, there are some outliers on either end of that scale).

According to the Intergovernmental Panel on Climate Change’s survey of the rebound literature included in the 2014 Fifth Assessment Report, “the majority of studies” show rebound effects for end-use energy services like heating, cooling, and lighting, “in the region of 20-45% … meaning that efficiency measures achieve 65-80% of their original purposes.” For transportation, the IPCC notes that “there are some studies that support higher rebounds,” with one study finding rebounds in transportation eroded more than half of the original energy savings. In industry, rebounds can also be significant, with one study finding a range of 20-70 percent across various industries.

Those figures are for the immediate rebound in demand for more efficient energy services (aka the “direct rebound”), and they do not include more indirect effects, such as the impact of spending any energy savings on other energy-consuming goods or services or the impact of improving energy productivity on economic growth (and thus energy use) nationwide. Add those factors in, and the total impact of rebound on overall energy demand rises further.

Yet when we think about the importance of rebound effects for climate strategies, the real story is in the developing world. More than 90 percent of energy demand growth over the next two decades will be fueled by the world’s emerging economies, and that’s where demand for energy services is far, far from saturated.

As the IPCC reports, there is thus “evidence to support the claim that rebound effects can be higher in developing countries.”

In the studies I’ve reviewed of rebound after end-use consumer energy services in developing nations (and there are comparatively few for the developing world, a gap that should be a major research priority), the direct rebound effects alone were much higher than in richer nations, on the order of 40-80%.

We should expect – and welcome! – larger rebounds in developing economies, because demand for energy services is far from saturated, demand is far more elastic (responsive to changes in price), and the cost of energy services is often a key constraint on the enjoyment of energy services.

Since expanding the supply of energy services is also a key constraint on economic activity in developing nations, the macro-economic impact of efficiency improvements in developing economies is also likely to be more significant, helping developing economies grow faster (and thus consume more energy).

This is exactly what Shellenberger and Nordhaus argue in the Times, as they ask us to consider the way ultra-efficient LEDs might “allow poor people to bring modern lighting into their homes much faster than they otherwise would. And … result in faster growth in energy demand globally.”

So how big a deal are rebound effects overall for global climate mitigation efforts?

The best study I’ve seen on this question is by University of Cambridge climate researcher Terry Barker and colleagues. They reanalyzed an efficiency plan from the International Energy Agency and found that rebound effects would erode 52 percent of the energy demand reductions by 2030. By downplaying rebound effects — the I.E.A. assumed rebounds were only about 10 percent — the influential agency overstated the contribution of efficiency to its climate plans by 88 percent.

There’s also reasons to believe this paper underestimates rebound effects overall. Most importantly, they assume the same, fairly modest degree of direct rebound for rich and poor countries alike, and that’s a poorly supported assumption. So consider this a conservative estimate. The real scale of global rebound from the kind of efficiency measures included in most climate plans is likely to be higher then, perhaps eroding 60 percent or more of expected energy reductions.

To quote my favorite Vice President, that’s “a big f-ing deal.”

The I.E.A. counts on efficiency to deliver the largest share of carbon dioxide reductions in their climate plan—more than all renewable energy sources combined, for example.

The I.E.A. is far from alone in banking big on efficiency as a climate tool. I recently completed a paper with Peter Loftus, Armond Cohen, and Jane Long that’s been accepted for publication in WIREs: Climate Change later this fall. Our paper reviews 17 global decarbonization scenarios from a range of sources, from individual academics to major international research efforts to environmental groups like Greenpeace and WWF. See the graphic below, which shows the primary energy supply mix for the final year of these scenarios for which the data was available.

Source: Loftus, Cohen, Long & Jenkins (2014), WIREs Climate Change (in press).

Compare the total energy demand in the I.E.A. reference case to each of these scenarios. The gap between total demand in the reference case and each of these other studies is the contribution each plan expects from energy efficiency improvements above and beyond what we would expect in BAU (note that a fair amount of LED adoption and other energy productivity improvements are going to happen as part of any future business-as-usual scenario, precisely because they make such good economic sense!)

Two things are notable:

First, the folks in “Group 4”, including the ones that try to show how we can get to deep decarbonization with renewables alone (i.e., Greenpeace/EREC & Jacobson & Dellucchi) depend on efficiency to keep global energy use from growing at all over the next 50 years, despite massive increases in GDP and 2 billion more people on the planet by then. That’s a heroic assumption, to say the least!

If you take rebound seriously (these studies don’t consider it at all), you quickly come to the conclusion that we really can’t bet the planet on renewables and efficiency alone. A more balanced portfolio is needed, and yes, that means making hard choices about nuclear, carbon capture and storage (CCS), etc…

But taking rebound seriously doesn’t just amount to saying “energy efficiency is imperfect and we can’t get the climate job done with efficiency and renewables alone.”

If you look at all of the other scenarios (except the first EMF22 scenario and Brook study at the top), the energy demand reductions from efficiency are larger than the energy contribution of any single zero-carbon energy source. In many cases much larger. Factoring in a significant rebound effects to most of these scenarios would be like erasing the entire contribution from solar and wind power combined, for example!

Ignoring rebound effects could thus lead global climate efforts to fall far short.

That’s the real motivation for me (and others) to make sure rebound effects are being taken seriously and carefully.

The implications are manifold: we may be significantly underestimating the contributions we need from renewables, nuclear, CCS, etc., and that may affect everything from the allocation of research and demonstration funding to deployment subsidy programs to how we design carbon pricing policies, etc. not to mention how climate advocates allocate their limited political capital and what kind of organizing strategies they adopt.

So let’s be clear: rebound effects are not a problem for energy efficiency. But failing to take rebound seriously would be a huge problem for climate mitigation. And when our planetary future is at stake, that’s a problem we can’t afford to ignore.