In 2017, Nicholas Bloom, an economics professor at Stanford University, was preparing to speak at a local TEDx event. His talk was about the then rare practice of working from home, which most people at the time associated with slackers and the occasional 20-something freelancing from a beach café in Thailand. He had led a study several years back that showed the surprising productivity benefits of the arrangement.
After a dry run, a conference organizer gave Bloom the typical TED advice. “You’ve got to be exciting for TEDx,” Bloom remembered her saying. “You’re too much of an academic. You’ve got to make it so it’s pathbreaking. Compare it to something really important.”
Bloom actually doesn’t fit the boring-professor stereotype — he’s an energetic speaker, the type who talks so quickly he’s difficult to interrupt — but he was willing to step it up. He channeled his former neighbor of many years Sebastian Thrun, who founded Google X and often kept one of the research lab’s self-driving cars in front of his home. Thrun’s own TED talk on the lifesaving technology of autonomous vehicles had attracted millions of views.
When it came time to step onto the stage, Bloom did his best to convince everyone that working from home — something only 8% of US workers did on a weekly basis — had the capacity to change the world.
“I actually think it has enormous potential — as much potential as something like, you know, the driverless car,” he told the audience.
The line drew a few unintended chuckles, and Bloom stumbled for a moment before moving on. “You’re trying to give a talk, and there’s people laughing because what you’re saying seems so ridiculous,” he said. His speech got 100,000 or so views on YouTube — maybe not a flop but pretty modest in the world of TED talks. The internet wasn’t all that interested in what was still an uncommon way of working. Neither were economists: The study Bloom summarized in his talk was his least cited paper published in a leading economics journal.
Four years later, self-driving cars are still years away from saving us from ourselves. But remote work has surged in a way that even its most ardent proponents couldn’t have predicted. Tens of millions of Americans are working from home, often for the first time, discovering the arrangement’s surprising benefits, even amid all the misery of the pandemic. And they’ve been hounding their managers for answers so they can start planning their post-COVID-19 lives. Should they re-sign the lease on their $3,000 studio in San Francisco? Can they move to Hawaii? Which school should they enroll their kids in this fall?
Desperate for advice, many companies have turned to Bloom. He was not only one of the few economists who had ever studied remote work before 2020 but also already in Silicon Valley’s backyard teaching at a university that’s basically synonymous with the tech industry.
Since last spring, Bloom, 47, has presented his research on remote work to about 100 managers at companies including Microsoft, Facebook, Google, Apple, Twitter, Intel, and ServiceNow. To better advise them, he has also conducted a number of surveys during the pandemic. One in three American workers, he has found, are preparing to keep working from home in some capacity post-pandemic. That would quadruple the time employees spend at home to 22% of all workdays — a massive shift away from the office that Bloom said would have taken 20 years without the pandemic.
From reduced commute times to better work-life balance, Bloom sees plenty to celebrate about the shift to remote work. But over the course of three video calls with me this year, he also said the transition would be rocky. He said many employees would be unhappy with the arrangements their companies offer, some would leave for competitors, innovation could slow, and promotion disparities would emerge as single men got more face time with their managers than women with young kids, which would culminate in all sorts of discrimination lawsuits.
Bloom is urging corporate America to navigate the turbulence by resisting offering the unlimited freedom that many employees have come to expect. His prescription lies in creating a uniform schedule for employees — one designed to balance the benefits of working from home with the need for collaboration and equality. Unless employers established a clear and level playing field, he said, the years ahead could remain filled with uncertainty and upheaval, for companies as well as employees.
“Revolutions are chaotic, and as we know from many revolutions in history, they’re often followed by further turmoil,” Bloom told me. “This has been a massive revolution — and we’re only halfway through.”
Born and raised in London, Bloom grew up thinking he’d find a job in business. As a teenager, he displayed an early entrepreneurial streak — he and his friends made a small fortune renting out an empty nightclub and charging their classmates to get in. As an undergrad at Cambridge, and then as a graduate student at Oxford, he studied economics to pursue his interest in business. He then worked at a think tank, while pursuing his doctorate at University College London on the side. After a stint at the UK treasury department, he landed a job as a management consultant at McKinsey.
The McKinsey gig didn’t last long; two years in, his Ph.D. advisor offered him a position as a research fellow at the London School of Economics. Still, his brief years in the corporate world proved formative for the career he was about to begin as an academic. Many economists at the time took it as a given that managers ran their companies well; otherwise, they said, the free market would drive them out of business.
“Having been at McKinsey, I realized that wasn’t true,” Bloom said. “At McKinsey, we were helping some clients that were really badly run.”
As he started measuring companies’ management practices and their outcomes, he developed an interest in work-life-balance policies, or what he likes to call “nice-to-people stuff.” His inspiration was partly his wife: She was offered very different maternity-leave packages from her employers over the course of giving birth to their first few kids. Bloom said he thought, “Surely there’s a best practice, if we could figure out what’s good and what’s bad.”
Bloom said he knew there was another way employers could promote work-life balance: letting people work from home. Both his mother, a civil servant, and his father, a doctor, fought hard to work from home for part of the week when Bloom and his three siblings were growing up. Following his parents’ example, Bloom had often worked from home too, going all the way back to his think-tank days in the ’90s.
In 2010, shortly after he received tenure at Stanford, a series of coincidences got Bloom thinking about remote work a lot more. He said he didn’t remember why, but one day he Googled the name of one of his Ph.D. students, a quiet man who sat in the back row of the class. The man’s name came up as one of China’s 500 richest people, Bloom said, “and he definitely wasn’t number 500 either.” He remembered thinking, “Ah, this is unbelievable, this person sitting in my class. This is so Stanford.” He ran to his colleague’s office to share the news. “You can’t believe who this guy is,” Bloom said.
The guy was James Liang, the cofounder and chairman of Ctrip, an online travel agency in Shanghai. At the time, the company was worth about $5 billion on the Nasdaq exchange. Today, under a new name, Trip.com, its market capitalization is about $21 billion. Apparently unsatisfied with just starting and growing his own internet giant, Liang was pursuing a doctorate in economics at Stanford. Bloom soon started talking to Liang, who told him about a dilemma that Ctrip had been facing. The company was doing well, but with Shanghai’s sky-high commercial-real-estate prices, executives weren’t sure they could afford to expand. So they were thinking about asking some employees to work from home.
Bloom jumped at the opening. This was a chance to do things scientifically. What if they randomly assigned employees to work from home and measure its effects? A randomized controlled trial is the social-science version of a drug trial; it’s the field’s gold standard and incredibly rare to run one at a large company.
So Bloom, Liang, and their co-authors went about designing their experiment, narrowing their subjects down to 249 call-center agents who handled airfare and hotel bookings in Ctrip’s Shanghai office. They assigned one group to work from home four days a week for the next nine months and the other to keep commuting into the office every day. Executives didn’t have high expectations, according to Bloom.
“The assumption was people would goof off. They’d watch the Chinese version of Oprah Winfrey or Jerry Springer, or play computer games,” he said. “This was the era when the newspapers were full of stories of people in online-gaming cafes playing for eight days solid and dropping dead.” But the hope was that, just maybe, the savings on office costs would offset the loss in productivity.
When the researchers ran the numbers, they didn’t find a productivity drop: Remote workers actually outperformed their office-bound peers by 13%. Most of that was because working from home allowed employees to spend more of their shifts working (they took fewer sick days and fewer breaks). But the remote employees also performed slightly better on a per-minute basis too. That, combined with how much Ctrip would save on office costs, worked out to a net benefit of $2,000 for the company for every employee who worked from home.
Ctrip’s managers “were astounded, literally astounded,” Bloom said, so much so that they subsequently rolled out a work-from-home option across the company. It looked like a massive success.
But over the years, Ctrip found that surprisingly few employees took advantage of the option, according to Bloom. People were worried they wouldn’t get promoted without regular face time with their managers. And they were right to worry: For all of the benefits to productivity, the study found that, controlling for performance, remote employees got promoted about 50% less than those who were in the office.
Bloom and Liang, together with the Stanford economist John Roberts and another Ph.D. student, Zhichun Jenny Ying, wrote up their findings in the paper “Does Working From Home Work?” The paper didn’t generate much interest in the following years. Journalists occasionally called Bloom about remote work when the topic was in the news, but Bloom himself didn’t pursue the subject again. He turned to an area of research on what happens when there’s uncertainty over government policymakers’ future actions, while he also did more studies on management practices.
Meanwhile, most Americans continued to commute to their workplaces. From 2017 to 2018, only 8% of employed workers worked one day or more from home per week, according to the US Bureau of Labor Statistics.
In January 2020, as a mysterious virus first found in China was starting to dominate the headlines, Bloom flew to Osaka, Japan, for a conference. He was startled to see the airport packed with Asian travelers who were all wearing masks. A month later, he flew to another conference in Mexico. Global stocks were starting to tumble as investors worried about the widening reach of the epidemic. It was the last conference he would travel to in 2020.
In early March, Microsoft became one of the first large companies to send its employees home. Other organizations quickly followed, including Stanford, which canceled all in-person classes with two weeks left in the quarter. Still, Bloom, like most in the country, thought this would prove a temporary blip. He told his students that they’d reconvene in person once this was all over in a few weeks.
It didn’t take long for the world to discover Bloom’s dormant Ctrip study. Journalists started to fill up his inbox, as did business executives. Working from home had suddenly morphed into a lifesaving policy intervention, and Bloom was one of the few economists who had studied it. He knew he had to seize the moment. “What’s the point of doing research if you have a national emergency, and you have research that’s connected to one of the ways that you can respond to the pandemic, and you don’t do anything?” he said.
But there was only so much he could say based on his Ctrip experiment, which was a decade old and conducted on only 249 call-center agents in Shanghai. So he started collecting more data. He fished out $5,000 from his research fund and set up a survey of workers in the US.
The results he analyzed in May confirmed that the shift to remote work was, in fact, huge. Out of 2,500 respondents, an astonishing 42% said they were working from home, while 26% said they were working on their business premises (another 33% said they weren’t working at all). “In a matter of weeks, we have transformed into a working-from-home economy,” Bloom wrote in a June report for the Stanford Institute for Economic Policy Research (SIEPR), where he’s a senior fellow. He teamed up with his former student Jose Maria Barrero, who had started teaching at the Instituto Tecnológico Autónomo de México, and Steven Davis, a professor of international business and economics at the University of Chicago’s Booth School of Business, to run more surveys through the summer, fall, winter, and spring, eventually amassing responses from about 30,000 Americans. The three of them coauthored a paper based on those results. They cowrote other studies based on employer surveys too.
Using these findings, Bloom has been giving two to three presentations a week. Among the businesses he’s spoken to is Microsoft, which deployed its own researchers at the onset of the pandemic to better understand working from home.
Bloom’s research “has been a valuable tool to help inform research at Microsoft related to the effects of remote work on peoples’ productivity and well-being,” Sonia Jaffe, a senior research economist at Microsoft, said in a statement, adding: “Researchers at Microsoft regularly share findings with internal teams to help inform decision-making, including those related to the company’s COVID-19 remote- and hybrid-work policies.”
Bloom has also spoken to Intel’s product innovation and marketing-insights group, which has been working with the company’s engineers to design a next-generation laptop that’s better suited for working from home. Bloom’s presentation “opened the mind,” Antony Barton, the director of the group at Intel, said.
From the beginning, Bloom has recommended that companies adopt a partial work-from-home policy: one to three days at home and the rest of the week in the office. Based on what he saw play out at Ctrip, as well as the early findings from his pandemic surveys, he’s said that both employees and their managers could benefit from incorporating remote work permanently. But he also said he knew that there would be drawbacks, like the difficulties around collaborating virtually and the risks of loneliness. By combining some days at home and some days in the office, “we can probably get the best of both worlds,” he said.
Within that hybrid framework, Bloom at first emphasized the importance of choice. “Let employees pick their schedules and let them change as their views evolve,” he wrote in his SIEPR brief in June. Here again, he had in mind the Ctrip experiment. After the original nine-month trial, Ctrip allowed workers to choose their arrangements, and productivity for those working from home rose even more, likely because those who realized they weren’t good at it went back to the office. There was also this to consider from his pandemic surveys: Workers had different preferences on how often they wanted to work from home. So why not let them choose? Everyone’s happy that way.
But after he talked to more executives, he realized that he missed a danger in his choose-your-own-adventure approach. A few of them told him that their female employees with kids seemed to prefer working from home more than their single male employees. Would that lead to problems down the road? Bloom went back to his surveys, and he saw the gap too: Among college-educated employees with young children, women were far more likely to want to work a fully remote schedule than men. Combine that with the fact that, in his Ctrip study, remote employees were promoted less than their office counterparts. It dawned on Bloom that this could have the makings of a “diversity time bomb,” followed by “a whole plethora of lawsuits,” he said.
Another challenge with implementing a hybrid workweek, he realized, was that people wouldn’t be guaranteed time together with their collaborators if they were free to choose which days they’d like to come in. With teams split between the home and the office on any given day, meetings would be awkward, some huddled in a conference room and others chiming in from little boxes on a screen. That would defeat the point of having them trek to the office in the first place.
So Bloom’s new advice is this: “Set one policy for everyone. Don’t let people choose.” That means that companies should require the same office-to-home ratio for everyone and for whole teams to come in on the same days of the week.
For a lot of professionals fed the promise of “flexibility” and “options” this past year, that’s probably not what they want to hear. Everyone now has their own ideal post-COVID-19 workweek, informed by their unique mix of circumstances and personalities and experiences. And even if we didn’t care all that much before about working from home, many of us likely hold strong beliefs about it now, now that we’ve tried it for a full year.
Bloom understands that’s fraught terrain, but he still thinks the alternative is worse — especially since it’s harder for a business to reverse course once, say, it lets an employee move to Alaska. “My advice is, in the short run, play it safe because you’re opening Pandora’s box,” he told me. “There are some adventurous firms out there. I would advise to go slowly and learn from their mistakes than from your own mistakes.”
Some companies have come pretty close to Bloom’s recommendation to require three days in the office and two days at home, like Microsoft, which is making a hybrid workweek of “less than 50%” at home the norm (employees will still have the option of going fully remote if their managers approve). On the other hand, Spotify, Twitter, and VMware have said that employees can pick their own schedules.
If Bloom is right, we should brace for some turbulence ahead. At the pick-your-own-schedule companies, women with young kids will probably be grateful to get to work remotely, only to find themselves out-promoted by their more office-bound male colleagues years later. The youngest employees could flail without the guaranteed face time with their managers and more experienced colleagues, as they already have in the pandemic. With fewer impromptu conversations in the hallways, innovation might slow down the road.
Even the companies that do listen to Bloom, and implement a uniform hybrid policy, shouldn’t expect the transition to be smooth. Employees who want to work from home either more often or less often will likely complain. Some may even leave for competitors that have arrangements better-suited to their preferences, as they’ve indicated in various surveys. And for all the trouble of implementing a new policy, businesses won’t save all that much in real-estate costs if they’re still having employees come in three days a week.
Or a mandatory hybrid workweek could turn out terribly for a whole different reason that we haven’t foreseen. After all, the vast majority of these companies that are considering such an arrangement have never tried it. What they tried over these past 12 months was working from home all the time. This is going to be a whole new experiment.
“It’s going to take awhile for businesses to figure this out,” Bloom said. “A year from now, there will still be a tremendous amount of uncertainty.”
Granted, these are likely concerns only the most elite corners of the US workforce will encounter. Even in the most terrifying stages of the pandemic early on, only about 40% of workers were working remotely, Bloom’s survey showed. According to a different study of each occupation in the country, that’s roughly the ceiling of how many jobs are even possible to do remotely.
Bloom’s survey suggests that one in three Americans will soon work from home partially or fully. That’s a whole lot more than before, but it’s still a minority of the workforce. The benefits of these new arrangements will overwhelmingly flow to those who already have the best working conditions: college-educated workers who already have high-paying jobs. That disparity will likely have big consequences for our economy and for policymakers too. “I see it as basically one more of many developments over the past quarter-century, or 30, 35 years now, that have favored the better-educated and the more highly skilled,” said Davis, the University of Chicago economist who has coauthored many of Bloom’s remote-work studies in the pandemic. “It just reinforces the need to improve our educational system in the United States, which serves many people quite poorly.”
For Bloom, some semblance of normality is just around the corner. At the beginning of April, when we had our third video call, he had just gotten his second vaccine. Stanford was slowly gearing up to reopen. Soon, all four of his kids would be back in school again. I thought he’d be more excited to get his kids out of the house, but he said he’d loved having them around. He told me he’d desperately avoided scheduling meetings at noon because that was when he ate lunch with his family every day.
He is, however, excited to start attending conferences again. Like his survey respondents, he’s been efficient working from home in the pandemic, executing on the ideas he’s already had. But sitting in his home’s guest bedroom, new ideas haven’t come to him like they used to. For that, he needs the buzz of a crowd, the serendipity of striking up a conversation in the coffee line with someone he otherwise never would have met. After all, working from home won’t always be the hot topic that it is today, once all the messiness gets worked out. “At that point onward,” Bloom said, “change will be much more gradual and boring.”
Andy Kiersz and Joe Ciolli contributed reporting.