Jevons paradox
Based on Wikipedia: Jevons paradox
The Curse of Getting Better
Here's a puzzle that haunts economists, environmentalists, and anyone trying to save the planet: What if becoming more efficient actually makes things worse?
It sounds absurd. If your car gets twice as many miles per gallon, surely you'll use less gas. If your factory needs half the coal to produce the same output, surely coal consumption will drop. If your irrigation system wastes less water, surely the aquifer will thank you.
But history tells a different story. A maddening, counterintuitive story that a Victorian economist named William Stanley Jevons first noticed in 1865, and that we're still grappling with today.
The Coal Paradox
In the mid-nineteenth century, Britain ran on coal. Coal heated homes, powered factories, and drove the locomotives and steamships that were remaking the world. And Britain was worried it was running out.
Many experts of the day offered reassurance. Technology, they argued, would solve the problem. James Watt had already revolutionized the steam engine, making it vastly more efficient than Thomas Newcomen's earlier design. As engines continued to improve, each task would require less coal. Problem solved.
William Stanley Jevons thought this was dangerously wrong.
Jevons was a careful observer of industrial economics, and he noticed something peculiar. After Watt's improvements made steam engines more efficient, coal consumption in Britain didn't fall. It soared. The more efficiently coal could be used, the more coal got used.
This wasn't a failure of technology. It was a feature of human behavior.
"It is a confusion of ideas to suppose that the economical use of fuel is equivalent to diminished consumption. The very contrary is the truth."
That's Jevons writing in his book The Coal Question, published in 1865. He was warning his contemporaries that their faith in efficiency was misplaced. Better technology would not save Britain's coal reserves. It would deplete them faster.
Why Efficiency Backfires
The logic, once you see it, is painfully simple.
When something becomes more efficient, it becomes cheaper to use. Not cheaper in absolute terms—the price of coal per ton might stay the same—but cheaper in terms of what you can accomplish with it. If your steam engine suddenly needs half the coal to do the same job, you've effectively cut the cost of that job in half.
And when things get cheaper, people use more of them.
Before Watt's innovations, steam engines were too expensive to operate for many applications. They were worth it for pumping water out of mines, where the alternative was flooded tunnels and dead miners. But for manufacturing textiles? For transportation? For countless other uses? The economics didn't work.
Watt changed the economics. His more efficient engine made steam power viable for industries that couldn't have afforded it before. Factories adopted steam. Railways spread across the countryside. Steamships crossed the oceans. Each individual application used coal more efficiently than before, but the explosion of new applications overwhelmed any savings.
Total coal consumption rose, not despite the efficiency gains, but because of them.
The Modern Rebound
Economists today call this the "rebound effect," and they've found it everywhere they've looked.
Consider your car. When fuel efficiency improves, driving becomes cheaper per mile. So what do people do? They drive more. They take jobs farther from home. They go on more road trips. They move to the suburbs. Some of the fuel savings from the efficient engine get eaten up by the additional miles traveled.
The question is: how much gets eaten up?
If the rebound effect is small—say, people drive ten percent more when their car gets ten percent more efficient—then improved efficiency still reduces total fuel consumption. The efficiency gains outweigh the increased use.
But if the rebound effect exceeds one hundred percent—if people respond to better efficiency by increasing their use so much that total consumption actually rises—then you've got the Jevons paradox in its full, frustrating glory.
Whether the paradox actually occurs depends on something economists call price elasticity of demand. This is a measure of how sensitive consumption is to price changes. When demand is "elastic," a small drop in effective price causes a large increase in consumption. When demand is "inelastic," consumption doesn't change much regardless of price.
Food is a classic example of inelastic demand. The agricultural revolution of the twentieth century—mechanization, synthetic fertilizers, improved crop varieties—made farming dramatically more productive. Food became cheaper. But people didn't respond by eating five times as much. There's only so much food a person can consume. Demand was inelastic, so the Jevons paradox didn't occur. Instead of everyone eating more, fewer people needed to be farmers. American agricultural employment dropped from forty percent of the workforce in 1900 to less than two percent today.
But travel? Energy use? These seem to have much more elastic demand. Make them cheaper, and consumption can expand almost without limit.
The Khazzoom-Brookes Postulate
In the 1980s, two economists—Daniel Khazzoom and Leonard Brookes—extended Jevons's observation to modern energy policy. Brookes, working as chief economist at the United Kingdom Atomic Energy Authority, argued that the whole premise of energy efficiency policy was flawed. Governments assumed that improving efficiency would reduce energy consumption. Brookes said it would do the opposite.
Khazzoom, meanwhile, was scrutinizing the California Energy Commission. The commission was setting mandatory efficiency standards for household appliances—refrigerators, air conditioners, washing machines—on the theory that more efficient appliances would reduce California's energy consumption. Khazzoom pointed out that they were ignoring the potential for rebound.
In 1992, economist Harry Saunders gave this idea a name: the Khazzoom-Brookes postulate. And he showed that standard economic growth theory actually predicted it would occur.
The argument works on two levels.
First, there's the direct rebound effect—the same phenomenon Jevons observed. When energy becomes more efficient to use, its effective price falls, and people use more of it.
But there's also an indirect effect, and it might be even more important. When energy efficiency improves, real incomes rise. You're getting more output for the same input, which means the economy is wealthier. And wealthier economies consume more of everything—including energy.
This is the trap. Efficiency improvements don't just change how we use energy for specific tasks. They change the entire economy, accelerating growth, increasing production, raising living standards. All of which increases energy demand.
Saunders argued that when you add up the direct and indirect effects—the microeconomic and macroeconomic impacts—technological progress that improves energy efficiency tends to increase overall energy use.
A Paradox, Not a Failure
Before we despair, it's worth pausing on an important point. The Jevons paradox, if it occurs, doesn't mean efficiency improvements are worthless. It means they're not sufficient for conservation.
Those are very different things.
The reason coal consumption exploded after Watt's improvements wasn't a catastrophe. It was the Industrial Revolution. The cheap power enabled by efficient steam engines didn't just go to waste—it transformed human civilization. It enabled the production of goods that improved material living standards for billions of people. It powered transportation networks that connected the world.
If your goal is to use less coal, efficiency won't help. But if your goal is human flourishing, efficiency has been spectacularly successful.
This distinction matters for policy. Climate advocates sometimes assume that the goal of efficiency policy is to reduce energy consumption. But efficiency improvements raise living standards, accelerate economic growth, and enable activities that weren't previously possible. That's not a bug—that's the point.
The problem arises only if you assume that reduced energy consumption will automatically follow. It won't.
The Policy Implications
So what do you do if you actually want to reduce resource consumption? The Jevons paradox suggests that efficiency improvements alone won't get you there. You need something more.
Several economists have proposed a straightforward solution: combine efficiency improvements with policies that keep the effective cost of using a resource the same or higher. If you make cars more fuel-efficient, but also raise fuel taxes proportionally, you've eliminated the price drop that triggers the rebound effect. People don't drive more because driving hasn't gotten cheaper.
This is the logic behind carbon taxes, cap-and-trade systems, and green taxes generally. They're not alternatives to efficiency improvements—they're complements. Efficiency improvements make goods and services cheaper, which threatens to increase consumption. Carbon taxes make them more expensive, which restrains consumption. Combined, you might actually achieve reduced emissions without sacrificing the productivity gains that efficiency provides.
The ecological economists Mathis Wackernagel and William Rees have proposed an even more direct approach. They suggest that any cost savings from efficiency gains should be "taxed away or otherwise removed from further economic circulation." The captured funds could then be reinvested in ecological restoration—rebuilding the natural capital that economic activity depletes.
This might sound draconian, but there's a gentler way to frame it. Efficiency improvements make conservation policies more palatable. If a carbon tax would otherwise raise energy costs and slow economic growth, efficiency improvements can offset some of that pain. People are more willing to accept policies that keep costs stable than policies that raise them. Efficiency is the spoonful of sugar that helps the medicine go down.
Water, Land, and Artificial Intelligence
The Jevons paradox isn't limited to energy. It shows up wherever efficiency improvements reduce the effective cost of using a resource.
Consider water. Engineers build a reservoir to protect a region from drought. Water becomes more reliably available. So developers build more homes, farmers plant more water-intensive crops, and industry expands. Water consumption rises to match the new supply. When the next drought comes, the region is more vulnerable than before—not less. Hydrologists call this the "reservoir effect" or the "safe development paradox." More infrastructure creates a false sense of security that encourages more risk-taking.
Or consider agricultural land. If crop yields improve—more bushels per acre—you might expect farmers to use less land. But if growing wheat becomes more profitable, farmers might convert more land to wheat production. The efficiency gain doesn't reduce land use; it transforms it.
And then there's artificial intelligence.
Microsoft's chief executive, Satya Nadella, has explicitly invoked the Jevons paradox when discussing AI. The concern is that as AI makes certain tasks more efficient, it might not eliminate the need for human workers—it might increase demand for the work those humans do.
The economist Erik Brynjolfsson argues that for some occupations, AI improvements could meet all three conditions for the Jevons paradox: technology improves productivity, the improvement reduces effective prices, and demand is elastic enough that consumption rises faster than productivity. Radiologists, translators, and programmers might find that AI doesn't replace them—it increases demand for their services by making those services cheaper and therefore more widely used.
This might sound optimistic for workers worried about automation. And maybe it is. But it's also a reminder that efficiency improvements rarely play out the way we expect.
The Debate Continues
Not everyone accepts that the Jevons paradox is a serious concern for energy policy. Many economists argue that the empirical evidence for large rebound effects is limited, especially in mature markets like oil consumption in developed countries. In these markets, demand may be relatively inelastic—people can only drive so much, heat their homes so much, fly so much. The rebound effect exists but typically falls short of one hundred percent. Efficiency improvements, on net, still reduce consumption.
Others counter that this analysis ignores the macroeconomic effects. Even if individual markets show limited rebound, the economy-wide impact of cheaper energy is harder to measure and potentially larger. When energy gets cheaper, economic growth accelerates, and that growth eventually translates into higher energy consumption—perhaps decades down the line, in forms we can't predict.
The debate matters because it shapes policy. If the Jevons paradox is rare or small, then efficiency standards—for cars, appliances, buildings, industrial equipment—are an effective tool for reducing emissions. If the paradox is pervasive, then efficiency standards are at best insufficient and at worst counterproductive. They need to be paired with consumption caps, carbon prices, or other policies that directly limit resource use.
There's a pessimistic version of this argument, often raised by opponents of environmental regulation. If efficiency improvements lead to more consumption, why bother? Why impose fuel economy standards if people will just drive more? Why push for efficient appliances if the savings will be spent on bigger houses and more gadgets?
But this argument proves too much. It implies we shouldn't pursue efficiency improvements at all—which no serious economist advocates. Even if the Jevons paradox means efficiency alone won't solve climate change, efficiency still provides enormous benefits: lower energy costs, higher productivity, improved living standards. The paradox is an argument for complementary policies, not for giving up.
The Lesson
William Stanley Jevons died in 1882, drowning while swimming at the age of forty-six. He never saw the age of petroleum, the internal combustion engine, or the climate crisis. But his paradox haunts us still.
The lesson isn't that efficiency is bad. It's that efficiency isn't magic. Making things cheaper to use makes people use more of them. If you want to reduce consumption, you have to deal with that reality, not wish it away.
This is uncomfortable for technologists who believe that innovation will solve our problems. It's uncomfortable for policymakers who want to offer voters painless solutions. It's uncomfortable for all of us who'd rather have better without giving up more.
But the Jevons paradox isn't a reason for despair. It's a reason for realism. Efficiency improvements are valuable—perhaps essential—but they need to be paired with policies that prevent the rebound from overwhelming the gains. Carbon taxes, cap-and-trade systems, consumption limits, green taxes—these aren't obstacles to progress. They're what make progress possible.
In the end, the Jevons paradox is about human nature. Give us more for less, and we'll take more. That's not a flaw in economics. It's a feature of being human. The challenge is building systems that channel that impulse toward outcomes we actually want.