Why do Americans work so much? The conventional answer is that it’s cultural. We’re a nation of workaholics. This perspective points to the Protestant work ethic or the power of the American dream as a root cause. The argument also gets circular. Sometimes, evidence of long working hours and the status value of being at the office all the time are offered to support the cultural explanation. But it doesn’t hold water.
The main reason is that for many decades, the United States was a place where people worked less. Before 1900, American hours were lower than in a number of European countries, such as Belgium, France, Germany, the Netherlands, and Italy. The U.S. was first to go to the five-day week. In 1950, Germany, France, the U.K., Italy, and Spain all had longer hours. Even through the 1960s, work schedules in Europe exceeded those in the U.S. Then the two regions took different paths. U.S. hours stagnated and rose. Europeans continued a century-long trajectory of reducing work time.
It is hard to credit “culture” when the thing to be explained departs from the nation’s history and is of recent vintage. As one influential paper on the differences between European and U.S. working patterns asks, “Why did ‘culture’ start diverging in the early seventies across the Atlantic so dramatically?” Why indeed? The authors find that unionization and regulation of mandated holidays account for the differences. In fact, the influence of unions raises a fundamental question about how labor markets operate.
The standard model says that workers have preferences between work and leisure and that employers offer jobs that match those preferences. The key assumption here is that employers are perfectly happy to let workers choose their hours. The other way to express this is that there’s a “market in hours.” Workers can “buy” more or less leisure according to their preferences.
But that’s not how labor markets work. Employers actually care a lot about hours and have their own reasons for preferring particular schedules. In the U.S., if you want a job with low hours, you have traditionally had to sacrifice benefits and a career ladder to get it. Companies offer what economists have called “tied wage-hours” offers. The wage and the hours go together, and traditionally, to change hours, workers have had to change jobs. This is why unions matter. They can be powerful enough to overcome what employers want.
Even through the 1960s, work schedules in Europe exceeded those in the United States. Then the two regions took different paths.
And what do employers want? In The Overworked American, I argued that they want long hours. I identified three reasons why. Reason number one is that in the U.S., employers offer health care, which is paid by the person, rather than prorated by hours worked. It functions like a tax on employment, giving employers an incentive to hire fewer people for more hours. This was an accidental and unfortunate pairing; during World War II, employers began offering health insurance to attract workers because wages were controlled by the government to keep wartime inflation at bay. Little did anyone expect this would distort the labor market eighty years later.
There are also some employer taxes, such as worker compensation and unemployment insurance, that have a similar incentive structure, because they are capped at a certain level. Hours above that level become cheaper for the company. To continue the contrast, this makes the U.S. system different from the setup in most European countries, where health insurance is financed independently of employment, thereby avoiding this disincentive. Employer-provided health care also helps explain why benefits aren’t paid for part-time work.
The second reason employers prefer long hours goes back to the cost of job loss. Longer hours (with more pay) raise the cost of job loss. That makes the worker more manageable, cooperative, or productive—pick whichever of those words you like. The easiest way to see this is to think about the difference between having two twenty-hour-a-week jobs and one at forty. Challenging the boss is more costly in the forty-hour job, because getting fired from it entails twice as much income loss as from one of the twenty-hour positions. My research found that work effort and the propensity to strike move up and down with the cost of job loss.
The final reason is the practice of paying by salary, rather than by the hour. In 2022, 44 percent of U.S. employees were paid on salaries, a fraction that has grown over time. Because additional hours are “free” to the company, these workers end up with elastic hours. (Have your hours crept up over time in a salaried job?) Years ago, I estimated that merely switching from being an hourly to a salaried worker would increase annual hours by a hundred a year or more. Together, these three factors serve as structural disincentives for work-time reduction in the U.S. economy.
There are powerful structural factors operating to keep hours high. We saw this in the Industrial Revolution. Those technological breakthroughs led to longer, not shorter, hours of work.
Other researchers have found an additional structural feature that has been driving up hours in the U.S.—the growth of inequality. As income flowed to the top of the distribution, that created competitive pressures for people to work more to keep their incomes up as well as to avoid falling further behind. Given that U.S. inequality was already high and has increased more than in Europe, this also helps explain the U.S./European divergence.
The rapid development of digital technology and AI, and the efficiency gains that come with it, is an opportunity to reverse that trend. But will it? There are powerful structural factors operating to keep hours high. We saw this in the Industrial Revolution. Those technological breakthroughs led to longer, not shorter, hours of work. In recent decades, digitization has transformed work in many occupations and industries, but in the U.S. hours haven’t fallen. I’ve argued that’s due to biases in the economy that have operated against hours reductions. Europe has some of these biases, but stronger unions and welfare states and a more equal income distribution have reduced those pressures, so European countries have continued to translate productivity growth into free time. Since 1973, I’ve calculated that the U.S. has taken less than 8 percent of its increased productivity to reduce hours, while western European countries have taken much more—generally three to four times that amount.
This history suggests that we can’t just sit back and let the market determine the impact of AI on our societies. Yes, there are structural pressures going in the direction of unequal benefits, unemployment, and maintaining long hours of work. But those outcomes are far from inevitable. As Daron Acemoglu and Simon Johnson argue in Power and Progress, whether new technologies bring widespread prosperity “is an economic, social, and political choice.”
Our U.K. collaborators at the Autonomy Institute have generated estimates of how many U.S. workers could be enjoying a four-day week within a decade as a result of AI. They estimate that the productivity gains of this technology could enable 28 percent of the U.S. workforce, or 35 million workers, to transition to thirty-two hours by 2033. A less stringent, 10 percent work-time reduction is feasible for 128 million, representing 71 percent of the workforce. We shouldn’t lose this opportunity.
“Why can’t these technologies, if they’re able to do 20 percent of what a human does—or did—why can’t that 20 percent be given back to the employee?”
Among my conversations with senior managers, only Matt Juniper of Praxis raised the specter of AI. “For a hundred years technologies allowed employees to work faster and more efficiently. Never at any point has that been given back to employees. The productivity enhancements have just been baked into employers’ bottom line.”
Matt sees AI as a chance to escape that trap. “Why can’t these technologies, if they’re able to do 20 percent of what a human does—or did—why can’t that 20 percent be given back to the employee? Maybe that’s a philosophical debate. But I feel it’s an important one.”
The impact of AI on work is ultimately a question about control—at many levels. Control over how the technology is used. Control over who reaps its rewards. Control over who has access to jobs and hours of work. And control over who is controlled—and surveilled—by it. Will the benefits of AI be shared by companies and workers alike? It’s up to us to decide.
From the book: Four Days a Week: The Life-Changing Solution for Reducing Employee Stress, Improving Well-Being, and Working Smarter by Juliet Schor. Copyright © 2025 by Juliet Schor. Reprinted courtesy of Harper Business, an imprint of HarperCollins Publishers.
