Jevons Paradox: Why Efficiency Rarely Means Less

Published On: July 28, 2026

Written by: Ben Atwater and Matt Malick

In 1865, Britain faced a quiet crisis. Coal fueled its factories, railroads, steamships, and steel mills. It heated homes and powered nearly every machine driving the Industrial Revolution. But the coal seams were finite, and many leaders worried that the nation’s prosperity was built on a resource destined to run out.

Engineers seemed to offer a solution. New steam engines extracted more work from every ton of coal. The logic felt straightforward: if each machine burned less fuel, Britain would consume less coal and extend its future.

William Stanley Jevons saw the flaw.

In The Coal Question, he argued that greater efficiency would not reduce coal consumption – it would increase it.

His reasoning was deceptively simple. When something becomes cheaper to use, people find more ways to use it. Lower the cost of energy, and manufacturers build more factories. Railroads run more trains. Steamships sail more routes. Entrepreneurs launch ventures that once made no economic sense. Each gain in efficiency expands opportunity, and opportunity, in turn, fuels demand.

History proved him right. Britain’s coal consumption did not decline. It surged.

Economists now call this the Jevons Paradox: greater efficiency often leads to more consumption, not less.

The pattern extends far beyond coal.

The automobile did not reduce travel – it made travel so affordable that suburbs spread outward, and millions began commuting daily. Computers did not eliminate paperwork – they enabled businesses to create, analyze, store, and distribute information on a scale that earlier generations could scarcely imagine. The internet did not reduce communication. It unleashed billions of emails, messages, video calls, and transactions every day. LED lighting lowered the cost of illumination – homes, offices, warehouses, streets, and stadiums responded by using more light, not less.

Human wants have no natural endpoint. Lower the cost of satisfying one need, and another quickly emerges. Investors often recognize this pattern – but many stop their analysis too soon.

Jevons tells us demand will grow. He says nothing about profit, which is an entirely different question.

Railroads transformed commerce, but they also attracted vast amounts of capital. Competition intensified. Rates fell. While the country prospered, many investors lost fortunes.

The airline industry made global travel routine. Warren Buffett once joked that a farsighted investor should have shot down the Wright brothers. Air travel became indispensable; airline shareholders rarely shared that success.

The internet reshaped the global economy. In the late 1990s, telecommunications companies spent staggering sums laying fiber-optic cable beneath cities and oceans. Their vision proved correct—the world eventually needed nearly every mile they built. Many of the companies that financed that expansion never survived to see the payoff.

Society benefited. Early investors often did not.

John Templeton built his career on a related insight: investors lose money when they mistake an extraordinary idea for an extraordinary investment. Revolutionary technologies tend to attract so much capital that competition erodes the very profits investors anticipate.

Artificial intelligence may be the latest example.

As computing costs fall, businesses will almost certainly discover new uses for AI, many of which are difficult to imagine today. Jevons predicted exactly this pattern: cheaper intelligence should lead to greater demand for intelligence.

Many people fear AI will destroy jobs. History disagrees.

The same fear gripped Britain in 1865. Machines threatened jobs then too. But cheaper power didn’t shrink the workforce. It grew industries nobody had imagined yet. Factories multiplied. Railroads needed conductors, engineers, and clerks. New jobs replaced old ones, and more followed.

The automobile killed the horse-and-buggy trade. It also built Detroit. Assembly lines needed workers. Gas stations needed attendants. Highways needed builders. One industry died. Ten more rose from its ashes. Computers threatened to wipe out the office worker. Instead, they created whole professions—programmers, analysts, an app economy nobody could picture thirty years ago.

AI will kill some jobs. Jevons would call that the wrong question. Cheaper intelligence doesn’t mean less work. It means more work, aimed at problems that were too expensive to touch before. The real question isn’t whether jobs will disappear. They always do. The question is what replaces them, and who captures the value when they do.

But the harder investment question lies elsewhere.

Technology companies are now spending hundreds of billions of dollars building data centers, designing chips, expanding electrical capacity, and developing software. They are racing to secure market share before the industry matures.

Will demand grow? History strongly suggests it will.

Will every company making these investments earn attractive returns on capital? History offers a far less certain answer.

Warren Buffett has long distinguished between a wonderful product and a wonderful business. Investors should apply the same distinction to artificial intelligence. A technology can reshape civilization while disappointing many of those who financed its early growth.

This is precisely why broad-based diversification matters more, not less, during periods like this one. Few investors picked the winning railroads in the 1860s, the winning automakers in the 1910s, or the winning dot-coms in the 1990s. Most investors who tried picking got it wrong, even when they correctly saw where the world was headed.

Jevons left us with a lesson that reaches far beyond nineteenth-century coal mines.

Efficiency lowers costs. Lower costs create opportunity. Opportunity drives demand. Demand attracts capital. Capital invites competition.

The first four steps build prosperous economies. The final step decides whether investors prosper.

Disclosure: Regulators could view this communication as marketing / advertising. This commentary is written by Ben Atwater and Matt Malick and reflects only their opinions and viewpoints. Atwater Malick, LLC sources facts, figures, quotations, etc. from what they believe are reliable sources, but they cannot guarantee their reliability. In addition, this essay makes no claims as to investment performance – past, present, or future. For additional important disclosures, please click here.