The Efficiency Dilemma

Original source

Summary

A comprehensive New Yorker feature tracing the Jevons paradox from its 1865 origins through its modern implications for energy efficiency policy. Owen argues that efficiency improvements, far from reducing consumption, consistently increase it — through proliferating appliances, expanding markets, and enabling new consumption patterns. The piece is the most thorough popular treatment of the Jevons paradox available: it covers the original coal argument, the rebound/backfire distinction, the refrigerator and air-conditioning case studies in granular detail, the lighting cost history from Babylonia to fluorescents, the Model T driving productivity argument, and the political economy of why efficiency is popular precisely because it demands no sacrifice.

Key Points

• Jevons Paradox originated in William Stanley Jevons’ 1865 book The Coal Question. Jevons was 29 when he published it. The book was admired by John Stuart Mill and William Gladstone and prompted a government investigation.
• Jevons’ original example: British blast furnaces. If technology made iron production require less coal, profits would rise, investment would flow in, iron prices would fall, demand would grow, and “the greater number of furnaces will more than make up for the diminished consumption of each.”
• Lighting cost history (William Nordhaus, Yale): An ancient Babylonian needed 41+ hours of labor for 1,000 lumen-hours of light. A contemporary of Thomas Jefferson’s needed ~5 hours 20 minutes (tallow candles). By 1992, an average American with compact fluorescents needed less than half a second. Yet total energy spent on illumination has only grown. “We now generate light so extravagantly that darkness itself is spoken of as an endangered natural resource.”
• Refrigerators: The 35-year period during which new refrigerators plunged in electricity use is also the period during which the global refrigeration market burgeoned and total energy consumption from keeping things cold climbed. Owen’s family: the old fridge moved to the basement (stayed plugged in 25 years), was joined by a stand-alone freezer and a bar icemaker. Modern kitchens: enormous side-by-side refrigerator + side-by-side freezer + under-counter mini-fridge. Gas stations now have almost as much refrigerated shelf space as 1960s grocery stores.
• Air conditioning: Between 1993 and 2005, energy efficiency of residential A/C improved 28%, but energy consumption for A/C by average air-conditioned household rose 37%. The US now uses roughly as much electricity to cool buildings as it did for all purposes in 1955. A/C ownership flipped from 12% (1960) to 84% (2005). A Las Vegas resident described cars as “devices for transporting air-conditioning between buildings.”
• Food waste connection: Since mid-1970s, per-capita food waste in US increased by half — we throw away 40% of edible food produced. More than a quarter of US freshwater use goes into producing food that is later discarded. Refrigeration enabled this waste by conveying a false sense that food will last longer.
• Model T comparison: Fuel economy ranged 13-21 MPG (1908-1927). Some modern vehicles do worse. But focusing on MPG misses the point — a modern driver gets vastly more benefit per gallon (comfort, reliability, food/lodging access, entertainment). Driving productivity has soared; so has consumption.
• Rebound vs. backfire: Economists distinguish “rebound” (efficiency gains partially offset by increased use) from “backfire” (increased consumption more than cancels savings). Harry Saunders (1992): “With fixed real energy price, energy efficiency gains will increase energy consumption above where it would be without these gains.”
• The political economy of efficiency: Steven Chu’s Clean Energy Ministerial efficiency discussions were “never contentious” — which Owen finds dismaying. Efficiency advocacy involves virtually no political risk, unlike capping emissions, pricing carbon, or increasing energy taxes. “Improving efficiency is easy to endorse: we’ve been doing it, globally, for centuries. It’s how we created the problems we’re now trying to solve.”
• The Schipper counterargument: Lee Schipper (Stanford) argues rebound is real but small in modern economies because energy is only 6-8% of GDP. Jevons effects were strongest in the 19th century when energy was a larger share of economic activity.
• The Alcott reductio: Blake Alcott’s counterargument to efficiency optimists — if efficiency reduces consumption, then inefficiency should increase it. But engines in 1800 were far less efficient than today’s; clearly a world frozen at 1800 efficiency levels would not consume more energy than we do now.
• Len Brookes revived the Jevons argument during the 1970s oil crisis, only later discovering Jevons had anticipated him by over a century.
• Jevons biography: Born Liverpool 1835, assayer at Sydney Mint, University College London, published The Coal Question at 29, Theory of Political Economy (1871) — founding text of mathematical economics. Drowned in English Channel at 46. Keynes called him “one of the minds of the century.”

Newsletter Angles

• The most citable Jevons source available: This piece is the gold standard for a general-audience explanation of Jevons paradox. The Nordhaus lighting history (Babylonian → Jefferson → fluorescent), the refrigerator family saga, the A/C ownership flip, and the Model T productivity argument are all ready-made examples for a Concept Decoder piece.
• The political economy angle is the sharpest: “Improving efficiency is easy to endorse: we’ve been doing it, globally, for centuries. It’s how we created the problems we’re now trying to solve.” This applies directly to AI efficiency claims — every tech CEO citing Jevons is making the politically costless argument.
• The food waste connection is an underappreciated second-order Jevons effect — efficient refrigeration didn’t just multiply fridges, it multiplied waste. This is the kind of “unexpected connection” that earns Marcus’s attention.
• Schipper vs. Brookes/Alcott maps directly onto the AI debate: Schipper says rebound is small because energy is a small share of GDP. But AI compute is a growing share of tech costs — more like Jevons’ coal than Schipper’s modern economy. The conditions that make Jevons strongest (rapid industrialization, energy as a large share of cost) are exactly what the AI buildout resembles.

Entities Mentioned

• William Stanley Jevons — biographical detail: born Liverpool 1835, Sydney Mint assayer, published The Coal Question at 29, drowned at 46
• Steven Chu — Secretary of Energy, efficiency evangelist, cited refrigerator improvements
• Amory Lovins — “a lunch you’re paid to eat” formulation; influenced Chu; Owen’s piece implicitly challenges Lovins
• Len Brookes — revived Jevons argument during 1970s oil crisis; key modern proponent
• Lee Schipper — Stanford; edited 2000 Energy Policy rebound issue; argues rebound effects are small
• Blake Alcott — ecological economist; reductio ad absurdum argument for Jevons
• Harry Saunders — 1992 paper on rebound; concise formulation of the basic idea
• William Nordhaus — Yale; 1998 paper on lighting cost throughout human history

Concepts Mentioned

• Jevons Paradox — the central subject; traced from 1865 through modern manifestations
• Rebound Effect — the general case; efficiency gains partially offset by increased use
• Backfire — the extreme Jevons case; increased consumption more than cancels savings
• Decoupling — the claim that economic growth is separating from energy use; Owen/Nørgård argue this is “largely a statistical delusion”

Quotes

“It is wholly a confusion of ideas to suppose that the economical use of fuel is equivalent to a diminished consumption. The very contrary is the truth.” — William Stanley Jevons, The Coal Question (1865)

“With fixed real energy price, energy efficiency gains will increase energy consumption above where it would be without these gains.” — Harry D. Saunders, 1992

“We now generate light so extravagantly that darkness itself is spoken of as an endangered natural resource.” — David Owen

“Improving efficiency is easy to endorse: we’ve been doing it, globally, for centuries. It’s how we created the problems we’re now trying to solve.” — David Owen

“A resident of Las Vegas once described cars to me as ‘devices for transporting air-conditioning between buildings.‘” — David Owen

“I am convinced that he was one of the minds of the century.” — John Maynard Keynes on Jevons (1905 letter to Lytton Strachey)

Notes

• Published December 20, 2010 — predates AI entirely but the framework is perfectly portable. The AI application is the reader’s synthesis, not Owen’s.
• This is the primary source for the refrigerator example used in the Jevons Paradox concept page and the Concept Decoder outline. The wiki previously attributed it to “David Owen, New Yorker, 2010” without a dedicated source page.
• The piece is exceptionally well-sourced for a magazine feature: Nordhaus (Yale), Schipper (Stanford), Brookes, Alcott, Saunders, and Cox are all cited by name with specific studies referenced.
• Owen’s political economy argument (efficiency is popular because it demands no sacrifice) is the strongest framing for the “half the story” thesis in the Concept Decoder piece — it explains why tech CEOs cite Jevons: it’s the politically costless argument.
• The Schipper counterargument (rebound is small because energy is 6-8% of GDP) is genuinely important and should be represented fairly in any piece that argues for large-scale Jevons effects. The counterargument is that AI compute may resemble Jevons’ coal (high share of cost in a rapidly growing sector) more than Schipper’s modern economy (energy as a small share of mature GDP).