At any given time, technology does two things for employment: it replaces traditional jobs, and it creates new lines of work. Machines replace farmers, but enable aeronautical engineers to survive. So, if technology creates new jobs, who gets them? How well do they pay? How long do new jobs stay new before they become another common task that any employee can do?
A new study of US employment led by MIT labor economist David Autor sheds light on all these matters. In postwar America, as Autor and his colleagues have shown in detail, new forms of work have benefited college graduates under 30 more than anyone else.
“We’ve never seen before who is doing the new work,” says Autor. “It’s more done by young and educated people in urban settings.”
The study also contains a powerful insight on a larger scale: A lot of innovation-based new work is driven by demand. In response to World War II, government-supported expansion of research and manufacturing in the 1940s led to large amounts of new work and new forms of expertise.
“It says wherever we make new investments, we get new expertise,” Autor says. “If you create a large-scale activity, there will always be opportunity for new specialized knowledge that will be relevant to it. We thought that was exciting to see.”
The paper, “What Makes New Work Different from More Work?” upcoming in annual review of economics. Writers are writers; Caroline Chin, a doctoral student in MIT’s Department of Economics; Anna M. Salomons, Professor at the Department of Economics at Tilburg University and the School of Economics at Utrecht University; and Brian Seegmiller PhD ’22, assistant professor at Northwestern University’s Kellogg School of Management.
And yes, learning about new work, and the types of workers who get it, may be relevant to the spread of artificial intelligence — though, in Autor’s estimation, it’s too early to tell how AI will impact the workplace.
Autor believes, “People are really worried that AI-based automation will destroy specific tasks more rapidly.” “Eliminating jobs is not the same thing as eliminating jobs, because many jobs involve a lot of tasks. But we’re all saying: Where will the new work come from? It’s very important, and we know very little about it. We don’t know what it will be, what it will look like, and who will be able to do it.”
“If everyone is an expert, no one is an expert”
The four co-authors also collaborated on a previous major study of new work published in 2024, which found that nearly six in 10 jobs in the US from 1940 to 2018 were in new specialties that had grown widely since 1940.
To do this, the researchers used U.S. Census Bureau data from 1940 to 1950, as well as the Census Bureau’s American Community Survey (ACS) data from 2011 to 2023. In the first case, because Census Bureau records become fully public after about 70 years, scholars can examine individual-level data about occupation, wages, and more, and track down the same workers who changed jobs between the 1940 and 1950 censuses. Counting
Through a collaborative research arrangement with the US Census Bureau, the authors also received secure access to individual-level ACS records. These data allowed them to analyze the earnings, education, and other demographic characteristics of workers in new occupational specialties and compare them with workers in long-standing specialties.
Autor believes that new work is always associated with new forms of expertise. First, this expertise is rare; Over time, this may become more common. In any case, specialization is often associated with new forms of technology.
“It requires mastery of some ability,” says Autor. “What makes labor valuable is not just the ability to do the work, but the specialized knowledge. And that’s often what differentiates high-paying work from low-paying work.” Furthermore, he adds, “It must be rare. If everyone is an expert, then no one is an expert.”
By examining census data, scholars found that in 1950, about 7 percent of the workforce had jobs in occupations that had emerged since the 1930s. Most recently, about 18 percent of workers in the period 2011–2023 were in jobs started since 1970. (That’s about the same share of new jobs per decade, although Autor doesn’t think it’s a hard-and-fast trend.)
Over these time periods, new jobs have emerged more frequently in urban areas, with people under 30 benefiting more than any other age group. Being employed in a new job appears to have a lasting effect: people employed in a new job in 1940 were 2.5 times as likely to be in a new job in 1950 as the general population. College graduates were 2.9 percentage points more likely to engage in new work than high school graduates.
New work also has a wage premium, that is, better pay overall than pre-existing types of work. Yet, as the study shows, the wage premium also diminishes over time, as particular expertise in a variety of new tasks becomes more widely understood.
“The value of scarcity diminishes,” says Autor. “It becomes common knowledge. It becomes automatic. New work becomes old.”
After all, Autor points out, driving a car was once a rare form of expertise. For that matter, in the 1990s it was possible to be able to use word-processing programs such as WordPerfect or Microsoft Word. However, after some time, being able to handle word-processing tools became the most elementary part of using a computer.
Let’s go back to AI for a minute
Studying who gets new jobs has led scholars to conclusions about how new work is created. Examining county-level data from the World War II era, when the federal government was supporting new manufacturing in public-private partnerships across the US, the study shows that counties with new factories had more new jobs, and from 1940 to 1950 85 to 90 percent of new jobs were technology-driven.
In this sense there was a lot of demand-driven innovation at that time. Today, public discussion about innovation often focuses on the supply side, namely innovators and entrepreneurs who are trying to create new products. But studies show that the demand side can significantly influence innovative activity.
“Technology is not like ‘Eureka!’ Where it just happens,” says Autor. “Innovation is a purposeful activity. And innovation is cumulative. If you get far enough, it will have its own momentum. But if you don’t do that, it’ll never get there.”
Which brings us back to AI, the topic many people are focused on in 2026. Will AI create good new jobs, or will it take away jobs? Well, it probably depends on how we implement it, believes Autor. Consider the vast health care sector, where a variety of tech-driven new work could occur if people were interested in creating jobs.
“There are different ways we can use AI in health care,” Autor says. “One is simply to automate people’s jobs. The other is to allow people with different levels of expertise to do different tasks. I would say the latter is more socially beneficial. But it’s not clear where the market will go.”
On the other hand, perhaps with government-driven demand in various forms, AI can be implemented in ways that will ultimately boost the productivity of the health care sector, resulting in the creation of new jobs.
Autor believes, “More than half of the dollars in health care in America are public dollars.” “We have a lot of leverage there, we can move things in that direction. There are different ways to use it.”
This research was supported in part by the Hewlett Foundation, the Google Technology & Society Visiting Fellows Program, the NOMIS Foundation, the Schmidt Sciences AI2050 Fellowship, the Smith Richardson Foundation, the James M. and Kathleen D. Stone Foundation, and the Institut Guck.
