The Paradox of Energy Efficiency

Why greener technology doesn't translate into reduced energy consumption


Automobile manufacturers have been working for decades on improving fuel efficiency. So why aren't the cars we drive today getting dramatically better gas mileage? Underlying that question is a fascinating paradox about energy consumption. 

A study by MIT economist Christopher Knittel in the December 2011 issue of the American Economic Review found that since 1980 the average fuel economy of American vehicles has increased only slightly, from 23 miles per gallon to 27. Yet Knittel found that fuel efficiency—the amount of power an engine produces per gallon of fuel burned—increased by 60 percent during that period. What's going on here? Cars and trucks have become bigger and more powerful: The average weight of passenger vehicles has increased 26 percent since 1980, while their horsepower has risen by 107 percent. Most of the gains in fuel efficiency have gone into compensating for the extra size and thrust. 

Automobiles are not the only category in which greater efficiency has failed to translate into reduced energy consumption. Lighting efficiency has improved during the last three centuries by many thousand-fold, from sputtering candles to modern LEDs, as Jeff Tsao and his colleagues from the Sandia National Laboratory note in the July 2012 issue of the journal Energy Policy. But the result "has been an increase in demand for energy used for lighting that nearly exactly offsets the efficiency gains." The authors note that "when lighting becomes cheaper, economic agents become very creative in devising new ways to use it," such as illuminating office ceilings with LED virtual skies. In coming decades, Tsao et al. predict, increased demand for lighting probably will again swallow up any new gains in energy efficiency.

In another recent study, reported in the July 2012 issue of the journal Sustainability, Graham Palmer, technical director of an Australian heating and cooling company, looked at trends in space heating efficiency during the last 50 years in Melbourne. Modern houses are up to 10 times more energy efficient, Palmer found, yet Australians are collectively using just as much energy to heat their homes as they did a half-century ago. Why? New houses are much bigger, people heat larger areas for longer, and fewer people live in each dwelling. Of course, modern Australians are much more comfortable in the winter than their grandparents were. 

Similarly, a 2006 study commissioned by the U.S. Environmental Protection Agency found that homes in Phoenix, Arizona, that qualify for the EPA's Energy Star designation use 12 percent more energy than homes that don't. Owners of Energy Star houses may use 16 percent less energy per square foot to keep their indoors livable, but they spend those gains on bigger houses. 

This energy "rebound effect" has important implications for efforts to restrain climate change through conservation. Various studies have suggested that improvements in efficiency could reduce energy consumption enough to cut global carbon dioxide emissions by as much as 25 percent during the next four decades. But this is a highly controversial area of scholarship.

In a 2007 Science article, Princeton University researchers Robert Socolow and Stephen Pacala calculated that seven "stabilization wedges" could prevent global carbon dioxide atmospheric concentration from rising to more than twice its pre-industrial level by 2050. "Improvements in efficiency and conservation probably offer the greatest potential to provide wedges," they argued. One wedge (equaling one-seventh of the necessary reduction) could be achieved either by doubling the miles-per-gallon performance of the planet's projected 2 billion automobiles or by cutting in half the distance they travel each year. Another wedge, they said, could be achieved by boosting the efficiency of coal-burning electricity plants from 40 percent to 60 percent.

California State University at Fullerton economist Robert Michaels tackles the thinking behind such conservation projections in a new report for the pro-market Institute for Energy Research (IER) titled "The Rebound Dilemma." Michaels' analysis divides rebound effects into four categories: direct, indirect, embedded-energy, and economy-wide. 

The Melbourne heating case illustrates the direct rebound effect: Better insulation and more-efficient heaters did not reduce energy use because people spent the gains on bigger, warmer houses. Another example: When cars get more mileage per gallon, driving becomes cheaper, so people tend to drive more. An indirect rebound occurs when efficiency improvements raise the productivity of other goods, thereby boosting the demand for energy. The demand for tires, for example, goes up as people wear out tires driving their energy-efficient cars more, so the tire industry uses more energy. Embedded energy is the extra power used to produce, distribute, and maintain energy-efficient goods such as high-efficiency insulation. And economy-wide rebounds, which include indirect and embedded rebounds, result from the ways in which people use their savings on energy to purchase other goods and services that also require energy to produce. 

Conservation proponents say direct rebound effects are often much smaller than the energy saved by increased efficiency. A classic 1992 Energy Journal study by David Greene, an environmental engineer at the Oak Ridge National Laboratory, found that the direct rebound effect of increased automobile fuel efficiency raised energy consumption by only 5 percent to 15 percent because people drove that much more. The new MIT study, however, suggests that most of the improvement in fuel efficiency during the last few decades has been spent on bigger and more powerful cars, a phenomenon that Greene did not consider.

Robert Michaels' review of research on the rebound effect associated with increases in household energy efficiency turned up varying results, but a quick look at the numbers shows a direct rebound rate hovering around 30 percent. If an appliance that used 100 kilowatt-hours per month is replaced by one that uses just 50 kilowatt-hours, a 30 percent rebound implies that energy consumption would fall not to 50 kilowatt-hours but to 65 (i.e., 30 percent more than 50), for a total reduction of 35 kilowatt-hours. Still not bad, since the consumer gets equivalent (or more) benefits from the new appliance while saving on electricity. 

Economy-wide rebounds are much harder to calculate. The money saved from driving a fuel-efficient car, for example, may now be spent on flying to a Caribbean beach vacation. Compounding indirect rebounds throughout the economy can lead to so much additional energy use that the net result of improved efficiency is higher consumption.

In the IER study, Michaels cites 11 econometric analyses that found economy-wide rebounds ranging from 23 percent to 177 percent; five of the studies reported economy-wide rebounds of more than 100 percent. In other words, the long-run result is higher energy consumption.

The upshot of all these studies is that energy efficiency mandates probably will fall far short of expectations for mitigating man-made global warming. "Instead of imposing energy efficiency mandates," Michaels concludes, "energy policy should embrace market prices and disruptive innovations to guide energy to its most valuable uses." After all, the point of improved energy efficiency is not to forgo the use of power but to boost its productivity as a way to provide people with more of the goods and services they want.