"Full Marks for Imagination": Reply from Amory Lovins

Thanks for writing. A few brief comments are below. May we post this exchange on RMI's Jevons blog, http://blog.rmi.org/blog_Jevons_Paradox?

Amory B. Lovins
Chairman and Chief Scientist
Rocky Mountain Institute
1739 Snowmass Creek Road
Snowmass CO 81654, USA
Executive Assistant (Patty Heydenberk)
Special Aide (Clay Stranger)

On 23 Mar 2012, at 15:16, Tom Walker wrote:

Dear Dr. Lovins,

I have read (and replied to) your responses to David Owen regarding the Jevons Paradox at the New York Times, Room for Debate and the Rocky Mountain Institute blog. I have also written a blog post at Ecological Headstand and EconoSpeak, which has received over 700 views so far today. I remain unsure about what I perceive as inconsistencies in your position and would appreciate clarification.

At the RMI blog you referred to the energy rebound argument as "this old canard":

"There is a very large professional literature on energy rebound, refreshed about every decade as someone rediscovers and popularizes this old canard."

In 2005, however, you wrote, "Far from dampening global development, lower energy bills accelerate it."("More Profit with Less Carbon," Scientific American 293, 74 - 83). By "lower energy bills" I assume you were referring to lower prices resulting from improved energy efficiency

I was referring to lower bills, which are the product of price times consumption—in this case, impliedly, the net bill, namely the value of the energy saved minus the cost of saving it.

and by "global development" I assume you meant economic growth.

Loosely yes, although, as you know, the metrics are controversial, and Gross World Product is not a meaningful one for many reasons.

Although the increased efficiency would lower the energy intensity of each dollar of GDP, the "accelerated" development would inevitably offset at least some of those efficiency gains with respect to total consumption. Leaving aside the proportions between energy saved through efficiency and energy consumed through accelerated development, that sounds to me like a rebound argument. If it walks like a paradox, swims like a paradox and quacks like a (Jevons) paradox, can we not conclude that it is a version of "this old canard"?

I've agreed for the decades I've participated in these discussions that rebound is real. It is just small. There is no inconsistency. I've never said rebound does not exist—only that it's unimportant. The "old canard" I referred to is the notion, as various advocates claim about every decade, that based on coal use after Watt's steam engine, it is generally true that energy efficiency is largely, wholly, or more than wholly ("backfire" or "Jevons paradox") offset by extra economic activity caused by energy efficiency. The literature provides no evidence for such claims.

It is common ground that disruptive technologies, some of which happen to save energy, can disproportionately boost economic activity. This does not make the general backfire thesis valid. As my response to David Owen's 19 March 2012 NY Times posting, at blog.rmi.org/blog_Jevons_Paradox, notes:

James Watt's more-efficient steam engine did spark an industrial revolution that (as Stanley Jevons observed) created great wealth and burned more coal. But this is no proof that energy efficiency generally triggers economic growth that devours its savings (or more)—a "backfire" effect never yet observed. Rather, it shows that many disruptive technologies stimulate economic growth and wealth, sometimes sharply. Some disruptive technologies, like microchips and the Internet, incidentally save net energy even though they are not meant to be energy technologies; some disruptive energy technologies, like automobiles and jet airplanes, increase energy use, while others, like electric motors, probably decrease it, and still others, like electric lights, could do either depending on technology and metrics (which Owen's cited lighting analysis muddles); still other disruptive technologies that Owen doesn't criticize, like key advances in public health, mass education, and innovation, enormously increase wealth and have complex and indeterminate energy effects. Blaming wealth effects on energy efficiency has no basis in fact or logic.

In the posted Times comments, Dr. Jon Koomey at Stanford (formerly LBNL) correctly notes:

Many people on this thread (including Owen) confuse the effects of getting wealthier with the effects of efficiency. Rebound, properly defined, concerns the latter. The problem of wealth causing people to buy more stuff and build bigger houses is only peripherally related to efficiency (efficiency has a minor impact on the total cost of owning and operating most goods and services). And in addition, there are saturation effects, so that the additional time that I might drive because driving is cheaper is limited by how much time I want to spend in the car. *** Getting wealthy is caused by many factors, and [energy] efficiency is only one such factor (and a comparatively small one at that).

I understand your position is that potential energy efficiency gains are large enough to more than compensate for increased energy demand from economic growth, thus implying the possibility of an absolute "decoupling" of growth and resource consumption.

Yes (in the sense that resource use and GDP could long sustain changes in opposite directions) for energy, generally yes for water and minerals, but with some potential exceptions, notably phosphorus. Also, we've analyzed U.S. energy-saving prospects only for the first 6x improvements (www.reinventingfire.com) vs 1975 aggregate intensities, although clearly there's even more after that, probably with good economics.

The index you cite in support of this contention expresses the energy intensity of GDP. However, I would suggest that a more salient index would measure the energy intensity of employment. I make this suggestion because Jevons in his original statement of the paradox referred to "a principle recognised in many parallel instances," and cited specifically the eventual expansion of employment after "the introduction of new machinery, for the moment, throws labourers out of employment." Thus the energy rebound argument is explicitly based on the parallel idea of an employment rebound. Nor is this a merely conceptual analogy because fuel is consumed running the machines that, according to the principle, displace labor (momentarily) but eventually expand the demand for it. The fuel consumption, economy of fuel, machines, labor demand and economy of labor are thus united in a continuous cycle.

I don't recall anyone else who has proposed or analyzed energy intensity of employment. It's an intriguing concept in theory, but I have trouble understanding how it could be meaningful in practice. Even if you did it properly and tried to analyze an elasticity of substitution between energy and labor, you'd encounter serious problems of measurement and uncontrolled fluctuations in both variables, due e.g. to weather and interfuel substitutions for energy and to compositional, labor-market, wage, demographic, productivity, and business-cycle variations (among others) for labor.

I'll share this with a couple of colleagues more expert than I in econometrics to see if they have any further thoughts.

I have extensively researched the original principle upon which Jevons's "parallel" paradox is based and have traced it to a pamphlet, "Thoughts on the Use of Machines in the Cotton Manufacture" written in 1780 by a Lancashire magistrate, Dorning Rasbotham, esq. I must confess that I find the original conception to have been a hasty generalization and a post hoc fallacy, so I would have to concede that the Jevons version must indeed be somewhat of an "old canard", as you call it. But, as Cecil Pigou pointed out a hundred years ago, "If it were a good ground for rejecting an opinion that many persons entertain it for bad reasons, there would, alas, be few current beliefs left standing!" Therefore the investigation must be carried further, specifically to the linkage between the energy rebound argument and the employment rebound.

I'm not clear how this would illuminate rather than obscure the existing conversation about energy rebound. However, you deserve full marks for imagination.

I have compiled indexes based on U.S. employment and energy consumption data from 1949 to 2009 and world data from 1980 to 2006. My calculations show that the energy intensity of per person employed in the U.S. increased substantially from 1949 to 1973, decreased from 1973 to 1986 and hovered around the 1986 level for the next 23 years. Globally, energy intensity per labor force participant (employment only data was not available) was around five percent higher in 2006 than it was in 1980. Thus, when examining the energy intensity of employment, there has been no sustained relative decoupling of employment from energy consumption. Without evidence of sustained relative decoupling, it is difficult to imagine how absolute decoupling could be achieved without a radical, fundamental change in economic paradigm.

That's a big logical leap. Analogously, U.S. primary intensity fell faster than GDP grew in nine of the past 35 years, just as you show a decline in energy intensity of employment during the energy-saving boom years of 1973-86, but you can't validly infer from these results' not occurring every year that they could never do so. I see no basis for that conclusion. — ABL

I have posted a chart at my blog showing my calculation of the energy intensity of employment and would welcome any criticism you might have of my analysis.

http://ecologicalheadstand.blogspot.com/2012/03/lovins-paradox-this-old-canard.html

Yours sincerely,

Tom Walker