Approaching Backfire

Energy consumption is going to explode in poor countries this century –– over 90 percent of the growth in energy consumption through 2050 will occur in non-OECD countries. These countries are also where the International Energy Agency (IEA) hopes to reduce future demand growth the most in the name of mitigating climate change –– 77 percent of the modeled demand reductions in the IEA’s 450pmm scenario come from non-OECD countries.

If we’re counting on energy efficiency to such a great degree in poor countries, then we want to be sure we really understand efficiency rebound in that context. Unfortunately, that’s not the case. Rebound is overwhelmingly discussed and studied in the context of consumer end-uses in rich countries. Productive sectors, emerging economies, and economy-wide rebound get much less attention, despite evidence that rebound is much higher in these contexts.

The upshot is that efficiency rebound is highest, and least understood, precisely in the contexts where it matters most: in industrial sectors and emerging economies.

Several recent analyses have concluded that rebound effects have a wide range of magnitude, with at least three different major assessments claiming that rebound effects rarely exceed 60 percent. [IPCC, IEA, Gillingham et al.] But a quick examination of the literature reveals many instances where rebound above 60 and even 100 percent has been measured. Here is a sampling:

• Broberg et al 2015 found rebound in the range of 40-70 percent in the Swedish industrial sector.
• Lin and Liu 2013 (a) found rebound between 91 and 107 percent in the Chinese transport sector.
• Lin and Liu 2013 (b) found rebound between 132 and 165 percent for residential electricity in China.
• Roy 2000 provided a seminal framework finding efficiency backfire (rebound above 100 percent) for lighting consumption in rural India.
• Lin and Li 2014 find rebound in Chinese heavy industry at 74 percent.
• Wang and Lu 2014 found a range of rebound effects in Chinese freight transport between 52 and 84 percent.
• Saunders 2013 finds rebound ranging between 19 and 120 percent in 30 US productive sectors.
• Tsao and Saunders 2011 and Nordhaus 1998 examine the history of lighting and conclude that efficiency improvements have led to continual backfire.
• Dahmus and Gutowski 2014 consider 10 activities (iron production, fertilizer production, electricity generation, etc.) and found that energy consumption in all sectors rises in the long term.
• Semboja 1994 modeled rebound effects between 170 and 350 percent at the economy-wide scale in Kenya.
• Dufournaud 1994 modeled economy-wide rebound effects between 47 and 77 percent in Sudan.
• Grepperud and Rasmussen 2004 found rebounds exceeding 100 percent for electricity in Norway.
• Glomsrod and Wei 2005 find evidence for potential backfire in Chinese coal use.
• Hanley et al 2005 find rebound above 130 percent for energy in Scotland.

Understanding where these high rebounds occur is vital to meeting future energy demand and accurately describing the decarbonization challenge. In general, instances of high rebound are found in poor or industrializing countries, where demand for energy is far from saturated, and in productive sectors of the economy, where demand elasticities are relatively high.

High rebound and backfire may not be the norm, particularly in the rich world. The plateauing growth rates of electricity consumption in the United States suggest that energy demand can saturate. But nor are high rates of rebound and backfire “a distraction,” as Kenneth Gillingham and his coauthors insist. Rather, backfire is a real phenomenon worth considering and studying for the contexts in which it is relevant. “Backfire appears to be a real phenomenon,” Harry Saunders wrote recently, “especially in developing countries.” Likewise, energy expert Steve Sorrell wrote that “while the evidence in favour of ‘Jevon’s Paradox’ is far from conclusive, it does suggest that economy-wide rebound effects are larger than is conventionally assumed and that energy plays a more important role in driving productivity improvements and economic growth than is conventionally assumed.”

Too often, the discussion of rebound effects has been deliberately narrowed, either to focus disproportionately on low estimates of rebound or to elide discussions of energy efficiency’s impact on productivity, sectoral reallocations, innovation, and economic growth. The real, positive impacts of high rebound and backfire need to be better understood, not ignored.