The “lump of labor fallacy” explained

Right now, a good chunk of Americans feel threatened by artificial intelligence. A recent Reuters/Ipsos poll found that more than half of Americans are worried that AI could put them or someone in their household out of work.

One economic concept that can help explain how new technologies tend to impact the labor market is known as the “lump of labor fallacy.”

“The lump of labor fallacy is simply the notion that there's a fixed amount of work to be done,” said Edouard Wemy, an associate professor of economics at Clark University. “And if one person [were] to gain one job, then it must be at the expense of somebody else.”

Catalina Amuedo-Dorantes, a professor of economics at the University of California, Merced, said in that scenario, the economy would be like a big game of “musical chairs,” where one player equals one worker and one chair equals one job. “So if one more worker is going to enter the labor market, someone else must lose a seat.”

But history tells us that the economy is not, in fact, a game of musical chairs. “Technology does not simply eliminate labor; it changes basically what workers do, and it can raise productivity, lower cost, and create new products, firms, and occupations in general,” said Amuedo-Dorantes.

That’s the fallacy part of the “lump of labor fallacy.”

If you think about the new technologies of the past, like mechanized agriculture, sewing machines, or word processors, those technologies did not lead to mass unemployment. Instead, they freed people up to work in other industries, create new businesses, and make new jobs.

“It's actually that they bring more chairs,” Amuedo-Dorantes said. “Some jobs will be lost,” said Wemy. “But more jobs overall are going to be created in terms of total employment.”

Over time, the total number of jobs in this economy tends to grow.

That is the lump of labor fallacy in action. But that thing about some jobs being lost when a new technology is introduced matters too. “Individually, it's gonna hurt,” said Wemy. “But on a macro level, more jobs are going to be created.”

So let’s bring it back to artificial intelligence.

“AI, as a technology itself, is different from technologies in the past,” said Nigel Melville, an associate professor at the University of Michigan who studies the socio-technical implications of AI. He said agentic AI’s capacity to mimic human behavior does separate it from previous technologies, but it’s hard to know how that will impact the labor market.

A new report from Challenger Gray and Christmas found that companies cited AI in more than 100,000 layoff announcements this year.

“I call it ‘the gap,’” Melville said. “So this is the difference between the people that are being laid off or being let go because of AI, and then the new jobs that are created because of AI.”

According to the Stanford AI Index, 88% of organizations now use artificial intelligence for at least one business function. “We don't know exactly what the gap will be,” Melville said, “but we do know that we're approaching level four or level five rapids.”

It hurts when a new technology takes your job. But the lump of labor fallacy tells us that, eventually, the economy will adjust and the total number of jobs will grow.

“So I would not say that history proves that AI will have no negative employment effects, but I would be skeptical of the claim that every automated task translates directly into a permanent lost job,” Amuedo-Dorantes said.

“There will be more chairs,” said Wemy. The big question is how long it will take, and whether workers will have the skills they need to sit in those new chairs when they arrive.