Certain aspects of scientific life do not lend themselves to working from home. Archaeologist Adrià Breu, who studies neolithic pottery at the Autonomous University of Barcelona, Spain, can’t dig for artefacts in his kitchen, and Claudia Sala’s experiments in molecular microbiology at the Toscana Life Sciences Foundation in Siena, Italy, oblige her to commute to her laboratory most days. But both these researchers also get to work from home — when they write up papers, for example, or analyse data.
It’s a familiar story. The COVID-19 pandemic accelerated a move towards hybrid working in science, as in many other professions, with many universities and institutes formally allowing staff to divide their time between working in the office or lab and working at home. Millions of people altered their work patterns almost overnight, and the changes have stuck.
But the impact of this sweeping shift is less clear. Remote workers claim that they are happier and more productive. But some studies suggest that teams that work in close proximity, including academic research groups, produce higher-grade, more innovative results.
On supporting science journalism
If you're enjoying this article, consider supporting our award-winning journalism by subscribing. By purchasing a subscription you are helping to ensure the future of impactful stories about the discoveries and ideas shaping our world today.
As hybrid working becomes established, researchers are racing to understand the full implications — for science and for everything else. Drawing on economics, psychology and communication theory, they are investigating many aspects of hybrid work, from the way people respond to e-mails and video calls, to how teams that are working remotely collaborate and transfer knowledge.
They are also exploring what science can offer to bridge the divide between office-based and remote teams, and thus make hybrid work a success.
Remote possibilities
Working remotely was an option for some people before the COVID-19 pandemic, but not for many. In 2016, just 4% of full paid days in the United States were worked from home. That proportion rose to as high as 60% in May 2020, and has since levelled out at about 25%. It’s a similar story in other countries. In UK government figures from 2022–23, almost half of workers reported spending some time working from home.
Researchers across the sciences have been ahead of the trend when it comes to working in geographically distant teams. As technology and policies have encouraged the exchange of ideas, data and materials, and as expertise has become more specialized, so the geographical spread of collaborating research teams has increased. A 2011 analysis looked at the addresses of some 39 million authors of research papers, and found that the average collaboration distance had increased more or less linearly from 334 kilometres in 1980 to 1,553 km in 2009. This indicates that remote collaboration was well established by this point and that teams were becoming more international.
Members of these remote research teams were generally not working from home. But the challenges of collaboration at a distance, and its reliance on technology rather than in-person communication, have much in common with how organizations and companies in all sectors are trying to build successful hybrid structures, says Ágnes Horvát, a communication and computer-science researcher who studies the impact of remote-working practices at Northwestern University in Evanston, Illinois.
In terms of the ways in which scientists work, says Horvát, “the problems we are facing are quite general.” This suggests that researchers can look to studies of remote or hybrid work at insurance firms and in other workplaces and apply the lessons to science, she adds.
There were plenty of studies to draw on even before the pandemic. Firms, researchers and business scholars have been tracking and predicting the consequences of remote working for decades.
In the 1980s, the US banking corporation American Express ran a successful pilot called Project Homebound, which was trialling a home-based alternative office system for people with disabilities. The project was hailed as a success, and the firm boasted of cost savings and increased productivity. But union officials were worried about exploitation, and called for a ban on “electronic home workstations.”
More recently, a series of small studies on specific groups such as call-centre workers and IT professionals have shown that fully remote workers tend to be less productive — by about 10–20%. They handle fewer calls, enter less data and take longer to perform the same tasks. This runs counter to claims in the early days of the pandemic that people who are based at home do more work than do those who are in the office.
In theory, hybrid work balances workers’ desire to be flexible with concerns from bosses about output. And a 2022 study of 1,612 engineers and marketing and finance employees at the global travel agent Trip.com seemed to back that up. The company assigned people to work from the office either full time or for two days a week. Staff working the hybrid pattern were happier and less likely to leave the company than were those who worked from the office full time. The results, posted as a working paper and not yet peer reviewed, suggested that, although the team members who were assigned to the hybrid group worked different hours and patterns from those who were office-based, the overall productivity of the groups was the same. Workers with longer commutes were more likely to report the benefits of being remote.
Permanently remote
Although such post-pandemic analyses are providing useful data, say researchers, they need longer-term studies to fully assess the rise of remote work.
“The pandemic showed us the effect of working from home in a rather short run, but we need much more evidence of what’s going to happen if we really keep on working remotely for years,” says Marina Schröder, an innovation economist at Leibniz University Hannover in Germany. She studies the effects of remote working on creativity, and has shown, for example, that communicating through chat software results in less innovation when compared with face-to-face conversation.
Late last year, one such long-term study led by Carl Frey, an economist at the University of Oxford, UK, produced the strongest evidence yet that remote work can alter the nature and quality of what researchers collectively produce.
Those based at the same site make more breakthrough discoveries, the team found. Although remote collaborators benefit from greater collective knowledge, such teams are less likely to be creative, and are better suited to making incremental progress.
“We showed in the paper that remote teams are more likely to collaborate in technical tasks,” says Frey, “whereas on-site teams are much more likely to collaborate actually in the conceptualization of new ideas.”
The study analysed 20 million research articles published between 1960 and 2020, and 4 million patent applications submitted between 1976 and 2020, around the globe. The researchers looked at the affiliations and geographical range of the contributors, and used citation analysis to assess how ‘disruptive’ publications were.
When the collaboration distance increased from 0 kilometres to more than 600 kilometres, the probability of disruption fell by about 20%. “Remote teams are less likely to create breakthrough findings,” Frey says.
Horvát says that the study provides a valuable note of caution amid the rush to embrace remote working. “This is not the way we want science to evolve. So, I think we absolutely need to take this very seriously.”
Innovation decline
What could be contributing to this trend? “Somehow the ideation process is more difficult when it’s mediated with technology. I think that’s as close to a mechanism as we have,” says Horvát. “That’s an important lack of knowledge on our part, because how are we going to fix it if we don’t know what’s causing it?”
Frey says that there could be several explanations for the decline in innovation. One is the value of sporadic encounters, which are much more likely when people work in the same place.
Meeting in person also exposes people to more knowledge. “If you go for lunch together and things like that, you get more ideas that are sort of filtered down to you because other people have read a lot of stuff themselves.”
A third possibility is what Frey calls collaboration intensity, which drives innovation by bringing together existing ideas from different fields.
“Fusing ideas takes time and effort,” he says. “It can click sometimes, but usually it’s a process. And it’s harder if you’re not in the same place and if you’re not communicating on a very regular basis.”
The nature of online communication, with appointments and priorities, is quite structured and hierarchical, adds Lingfei Wu, an information scientist at the University of Pittsburgh in Pennsylvania, who worked with Frey on the study. This can impede informal conversations and the casual generation of ideas, and might make it harder, for example, for early-career scientists to communicate with more senior colleagues.
“Those who went through junior stage, we all understand how hard it is to get a senior professor to respond to an e-mail,” says Wu. “But if you actually run into a senior professor in the hallway, then it’s easier to say you have a couple of ideas.”
He points to this effect in the data that were collected for the remote-collaboration study. By assessing the relative status (based on numbers of citations) of co-authors on published papers, the analysis showed that collaborations between researchers of markedly different status were much more common when the two individuals in question shared an office or building than when they worked remotely from each other.
A lack of collaboration could have negative consequences for scientists of any status: in a preprint that has not yet been peer reviewed, Wu’s team shows that younger scientists can help older scientists to produce more innovative work.
The group performed an analysis of 241 million articles published by more than 244 million researchers over the past two centuries, and examined the related patterns of citations. It found that the longer that scientists work in a field, the less often their research is classified as disruptive. That trend has become more pronounced in recent decades. In the 1960s, researchers with 20 years of experience produced more than 2% of the most disruptive work. By the 1990s, that had slipped to less than 0.5%.
In a finding that will surprise few early-career researchers, the analysis of publications and how they were cited showed that older scientists were much more likely to criticize emerging work than they were to produce innovative research themselves.
Remote collaboration and the lack of sporadic in-person encounters could be reinforcing hierarchies and exacerbating the trend, Wu says.
Water-cooler effect
The value of spontaneous in-person encounters for generating ideas — known as the water-cooler effect – is especially associated with creativity. And a 2022 study from two US social scientists demonstrated that communicating through screens can’t replicate this personal touch.
Melanie Brucks at Columbia University in New York City and Jonathan Levav at Stanford University in California asked pairs of volunteers to think of alternative uses for objects such as a frisbee disc and bubble wrap. Half of the creating couples worked in the same room, whereas the other half communicated by video call using laptops. The researchers also set up a similar study among pairs of engineers working on product design in five office locations around the world.
The remote collaborations created fewer ideas than the in-person teams did. But, in follow-up tests once the ideas had been generated, the remote pairs were just as effective as were the in-person pairs — or more so — at analysing the options and deciding which they should pursue.
How did the screens limit their creativity? Eye-tracking technology showed that the virtual couples paid more attention to each other — and the screens did not seem to stop the pairs from generating feelings of connection and trust, or to prevent them from mimicking each other’s language or facial expressions. Instead, the researchers argue that concentrating on a relatively small screen narrows cognitive focus. In turn, this switches off the mental ability to associate and combine concepts, which underlies ideation.
Face-to-face meetings could also boost creativity because they enable teams to fully exploit collective knowledge, in a way that remote collaboration does not.
“If my teammate is really good and I’m seeing what they’re producing, that’s kind of impactful to me,” says Glenn Dutcher, an economist at Ohio University in Athens, who has studied the effect.
Zoom fatigue
Similarly to other industries, some labs have seen the value of in-person meetings and have moved to restore them. “We met on site for the first time after nearly two years last December and were all surprised by how good it felt to be all back in the same room,” says Viktor von Wyl, an epidemiologist at the University of Zurich, Switzerland, who runs a lab of ten people. “We have now decided to go back to at least one team meeting per month in person.”
Although videoconferencing might not be as effective as meeting in person, it still conveys higher-grade communication than do tools such as e-mail and instant messaging. That’s because psychologists see phone and video calls as ‘synchronous’ media, in which communicating in real time helps participants converge on the meaning of complex information. E-mails and messaging, by contrast, are asynchronous channels that are better suited to simply conveying that information. And when people work remotely, they tend to send e-mails.
That effect was shown by the computer giant Microsoft, which used the enforced shift to remote working as a natural experiment to assess how the company’s 61,000 staff members in the United States responded in the first half of 2020. The analysis showed that remote working actually decreased the number of video or phone calls across the company, as staff switched to e-mail and messaging.
Something similar showed up in the data from the Trip.com analysis. Hybrid workers, the study found, were more likely to message colleagues than they were to use the phone or speak to them in person, even when they were all in the office.
Horvát argues that ongoing improvements in technology could fix some – but not all – of the issues with remote work, including its impact on creativity. Experiments with virtual reality, for example, have shown that participants can use and pick up on gestures and body language, which is a crucial part of in-person communication. And file and data sharing through the cloud have streamlined the way in which remote teams carry out joint projects.
“Technology looks very different now, especially post-COVID,” she says.
There are certainly reasons to be cheerful about the future of at least some remote collaborations.
In a 2022 working paper (not yet peer reviewed and published), Frey and his colleagues at the University of Oxford looked at remote collaboration and scientific innovation from 1961 to 2020, and found a surprising twist. After 2010, scientific papers written by remote collaborators were more likely to contain breakthroughs than were papers written by single-location teams.
Unlike their 2023 study, which found fewer breakthroughs over time, this analysis looks only at the output of existing teams that start on site and switch to remote working; it does not capture the impact of teams that have always been remote.
The switch after 2010 makes sense, says Nick Bloom, an economist at Stanford University, because that’s when file-sharing technologies such as Dropbox emerged. (Bloom studies remote working and has co-authored two papers on the subject.) Frey adds that the trend after 2010 could be due to what economists call knowledge spillovers — each collaborator exposes others in their home institution to the ideas.
Researchers who study work patterns say that there’s no single solution that optimizes everything about jobs, especially in science. Although breakthroughs are important in research, says Dutcher, they often require major investments, such as getting people together. “We need the big discoveries, and for those maybe we need face-to-face meetings,” he says. “But we also need the small advancements.”
This article is reproduced with permission and was first published on March 4, 2024.