America is getting better at getting things done.
Take Vic Viktorov, a gym owner who increased revenue at his Boston business in 2024 by 30% without adding a single salesperson to the two already on staff. Instead, he has been using an artificial-intelligence model loaded with company documents, sales materials and other information. Now, he can complete in just minutes work that used to take hours, such as writing marketing plans, email drafts and social-media posts.
“It allows us to be lean, nimble and fast,” said Viktorov.
Productivity in the U.S., as measured by how much the average worker gets done in an hour, has been on the rise. That matters because the faster that productivity grows, the faster the economy can grow as well. The success of the U.S. economy, and why it has grown so much compared with other countries over the past century and more, has hinged on its productivity.
Productivity—the total output of the economy divided by hours worked—rose 2% in the third quarter compared with a year earlier, according to the Labor Department. That marked the fifth quarter in a row with an increase of 2% or better. In the five years before the pandemic, there were only two such quarters.
The gains in part reflect massive changes in the U.S. economy since the onset of Covid-19. Companies learned new ways of doing things and adopted new technologies, while an upheaval in the labor market moved workers into more productive jobs.
Another big change in the American labor force—a massive influx of immigration—might also have played a role. Immigrants are often slotted into manual-intensive jobs, which could allow other workers to move up to more highly skilled jobs.
Businesses learned new ways to operate: QR codes instead of paper menus at restaurants, for example, or a videoconference instead of a time-consuming trip out of town. There has also been a big and continuing jump in the number of new businesses getting started.
And workers, for their part, moved themselves into better-paying and higher-skilled jobs. When restaurants, hotels and retailers reopened after briefly shutting down, they struggled to find workers and were more inclined to offer bonuses or promotions. That made it easy for, say, a cashier at a poorly run store to get work at a well-run one—where he might earn more money, have more responsibilities and get more done.
Of course, increased productivity isn’t always good news for workers: One way that companies get more productive is by laying off employees. New technologies such as AI can create new jobs and make workers more efficient—or take their jobs.
The recent dockworkers strike was fueled in part by port employers’ desire to expand the use of automated machinery on docks. President-elect Donald Trump threw his support behind the dockworkers, saying in December that automation threatened jobs.
And it isn’t clear that the move up in productivity growth will last. The figures are both volatile and subject to revision. The wave of job switching after the pandemic hit has run its course. And so far, productivity isn’t experiencing anything like the boom in the 1990s, when the wide-scale adoption of the personal computer and the advent of the internet reshaped the economy.
But at the least, it looks better now than before the pandemic, when economists worried the U.S. was stuck in a low-productivity funk.
With labor scarce in recent years, Novae, a Markle, Ind.-based maker of trailers for pickup trucks, built a state-of-the-art factory that opened six months ago. It cost $35 million, about seven times more than typical plants in the industry, and output is already 35% higher per worker, according to Chief Executive Manish Bhandari. He expects even better results over time, partly because the new factory helps the company retain its skilled workers.
At the plant, Novae automated some processes and incorporated improvements suggested by workers. One employee designed a bin that hangs 3 inches away from assemblers’ hands and holds a fastener used in the trailer’s frame.
The company also worked with Streamliners, an operations consulting firm, on an older factory near Minneapolis, with a goal of increasing productivity by 70%. Lacking additional room to expand, the team designed a whole new layout for the existing space.
“There is no silver bullet here,” Bhandari said. “It’s hundreds of small decisions.”
‘They don’t have anything to lose’
The stakes are high. Economic growth fundamentally relies on how many people are working and how much they can produce while they are on the clock.
But America’s scope for expanding its labor force is limited: The population is increasing slowly, the baby-boom generation is retiring, and Trump has promised to heavily restrict immigration and deport millions of immigrant workers who are already in the U.S. Stronger productivity would help bolster the economy and support an aging population.
Productivity also helps keep inflation in check: A more efficient business can be more profitable and pay its workers more without raising prices.
In November, there were a seasonally adjusted 157,678 “high-propensity” new-business applications, those with a high likelihood of turning into businesses with payroll, according to the Census Bureau—nearly 50% above the monthly levels that prevailed before the pandemic.
That is a positive sign for productivity, for two reasons, according to University of Maryland economist John Haltiwanger.
First, when there are new opportunities for innovation, as with cars a hundred years ago or computers in the 1980s and 1990s, new businesses proliferate. Second, new businesses are quicker to adopt new technologies. That can allow them to hire fewer workers to get things done.
“They’re more likely to do radical things,” Haltiwanger said. “They don’t have anything to lose, so to speak.”
Hybrid-work arrangements might have also helped productivity for white-collar workers by creating a balance between the quiet of home and face-to-face interactions of the office. Hybrid work also appears to improve employee retention, said Stanford University economist Nick Bloom, meaning businesses don’t lose time training new workers.
No emails, no problems
Viktorov, the gym owner, had a sudden heart attack at the age of 40, after working in management consulting for two decades. That fueled his decision in 2022 to open a gym focused on athletic-style training.
He employs two salespeople and eight coaches but manages the rest—marketing, human resources, information technology, facilities and other functions—on his own.
Viktorov uses generative AI, which can create content from large pools of data, for more than marketing tasks. It also helps with complex projects such as figuring out financing options or drafting an employee handbook, he said.
“If I can save an hour, two hours a day by speeding up these tasks, it makes me much more efficient,” said Viktorov, whose gym is a Boston-area franchise of Tennessee-based D1 Training.
It takes time, though, for a successful technology to be used widely enough and effectively enough for it to show up. So while ChatGPT and other GenAI tools are garnering lots of attention, and some businesses are using them, they are probably too new to move the needle on productivity across the economy yet, said Harvard University economist David Deming.
“They haven’t been around long enough, and there hasn’t been enough embedding of them in organizations in ways that change practices,” he said.
But older types of AI technologies could already be making businesses more efficient, Deming said. For example, some AI that can help companies manage inventory predates the pandemic.
Raj Karanam took over Architectural Surfaces, which distributes stone and other materials for homes, five months ago. In that short time, he has reduced product shortages 95%, largely by using advanced analytics and AI to manage inventory.
That leads to as much as $2.5 million in added revenue a month from sales the company would previously have lost because materials were out of stock.
In the past, he said, a showroom in Denver might need a slab of quartzite that is in stock in Austin, Texas. Dozens of emails would go back and forth to approve and initiate a transfer. Now, he said, all of that happens automatically. “Emails don’t even get triggered so you eliminate that waste, and we’re getting inventory to the right location as quickly as possible.”