Study Shows Which Workers Get New Tech Jobs - and How Long the Pay Premium Lasts
A new MIT study of postwar U.S. employment reveals a consistent pattern: college graduates under 30, living in cities, have historically captured most new jobs created by technology. The finding matters now because the same question looms over artificial intelligence - who will fill the jobs AI creates, and who will be left behind?
The research, led by MIT labor economist David Autor, analyzed Census Bureau records from 1940 to 1950 and recent data from 2011 to 2023. It shows that new forms of work emerge regularly, but they don't distribute evenly across the workforce.
Who Gets New Work
In 1950, about 7 percent of workers held jobs in occupations that had emerged since 1930. By the 2011-2023 period, roughly 18 percent worked in fields that had developed since 1970. That's roughly the same share per decade, though Autor cautions against treating it as a hard rule.
The pattern is stark. College graduates were 2.9 percentage points more likely than high school graduates to enter new work. People under 30 benefited more than any other age group. Urban areas saw more new work than rural ones.
If someone held a new job in 1940, they were 2.5 times more likely to be in new work in 1950 compared to the general population. Early entry into emerging fields appears to create a lasting advantage.
The Wage Premium Fades
New work pays better than established work - at first. But that premium erodes as the specialized knowledge becomes common.
Driving a car was once scarce expertise. Using WordPerfect or Microsoft Word was still a marketable skill in the 1990s. Today, word processing is elementary. "The scarcity value erodes," Autor said. "It becomes common knowledge. It itself gets automated. New work gets old."
This pattern has clear implications for workers entering fields built around new technology. The higher pay reflects scarcity, not permanent advantage.
Demand Drives Creation of New Work
The study examined county-level data from World War II, when the federal government backed manufacturing expansion across the country. Counties with new factories saw more new work emerge. Between 1940 and 1950, 85 to 90 percent of new work was technology-driven.
The finding challenges how innovation is typically discussed. Public discourse focuses on the supply side - entrepreneurs and inventors creating new products. But this research shows demand matters enormously. When government or industry invests in a new activity at scale, specialized knowledge requirements follow.
"Wherever we make new investments, we end up getting new specializations," Autor said. "If you create a large-scale activity, there's always going to be an opportunity for new specialized knowledge that's relevant for it."
What This Means for AI
The study doesn't directly address AI, but its implications are relevant. AI adoption could destroy jobs or create them - the outcome depends partly on how organizations choose to deploy it.
Health care illustrates the choice. AI could automate away existing jobs. Or it could allow workers with different skill levels to perform different tasks, boosting productivity while creating new roles. "The latter is more socially beneficial," Autor said. "But it's not clear that is where the market will go."
Government spending offers leverage. More than half of U.S. health care dollars are public money. Policy choices about how to deploy AI in that sector could shape whether new jobs emerge or existing ones disappear.
The broader lesson: new work creation isn't inevitable. It depends on investment decisions made now.
For HR professionals managing talent strategy, the research suggests three takeaways. First, new job categories will emerge, but they'll concentrate among younger, college-educated workers in urban areas unless deliberate effort widens access. Second, the wage premium attached to emerging roles is temporary - as skills spread, pay converges with established work. Third, where organizations and government invest resources, new specialized roles follow. Workforce planning should account for this relationship between investment and job creation.
The paper, "What Makes New Work Different from More Work?" will appear in the Annual Review of Economics.
Your membership also unlocks:
