Cafeterias, Bars and Innovation
This is the fourth in a series of articles on the transition of the workplace from in-office to hybrid. You can see the previous posts about the impact on office space (here), how the rest of the world is handling back to work (here) and on how hybrid won (here).
Post-pandemic, most American workers would prefer to work from home at least part of the time. Many would prefer to work from home all of the time. And a whole bunch of managers (mostly those over 50 years old) would prefer that they work from home none of the time.
There’s been an energetic debate about all that -- and it appears hybrid work has won. In many ways that is good. We’re not pumping as much CO2 into the air commuting to work and workers are enjoying the flexibility. It also appears workers in many settings are getting slightly more total work done when working from home, mostly because they are using some of their saved commuting time to work longer hours.
What’s not as clear is the impact of work from home on innovation. There are two ways researchers have been looking into this question.
Question 1: What happens to innovation within companies/organizations in a hybrid/remote environment? The “cafeteria effect.”
Beginning 50 years ago, researchers began highlighting the importance of “weak ties” – encounters of people who don’t know each other very well – in spurring job advancement and in-office innovation. When people aren’t physically together in the workplace, the theory goes, they lose “propinquity,” the benefits of proximity. The catchier name for this is the “cafeteria effect”; the positive benefits of bumping into colleagues in the lunchroom or at the copier or by the coffee machine. By contrast, when we work remotely, we are more likely to restrict our interactions to electronic encounters with a much-smaller group of people that we already know well and less likely to have fortuitous encounters with those outside of our core circle.
About a year ago, MIT published a study that quantified the decline in weak tie interactions during the first 18 months of the pandemic. Analyzing email interactions between all researchers in the university, the study concluded that during that period there was a 39% decline in weak ties connections being formed – there were 5100 fewer weak ties connections made when researchers couldn’t run into each other. When researchers came back together, they discovered that weak ties bounced back.
What’s the connection between weak ties and innovation? Mark Granovetter, who originated the concept, argues that “individuals with many weak ties are…best placed to diffuse difficulty innovation.” More weak ties = more innovation.
The iconic story of weak ties in action comes from Google. Larry Page, one of the co-founders of the company, apparently got outraged one day when he did a search and found that the ads that played during the search didn’t seem closely related to the topic of the search. He took a screen shot, printed it out with the words “these ads suck” on it. He stuck the paper on a bulletin board in a break room and went home for the day. Five different people from five different departments, none of them on the ad team, saw the note on the bulletin board, talked about it, and by the next day had designed a whole new algorithm to solve the problem. A gazillion dollars later, Google is still reaping the benefits of that chance interaction. That’s the power, the argument goes, of propinquity, or weak ties, or something.
Question 2: What happens to innovation between companies in a hybrid/remote environment? The “nearby bar” effect.
For years, one of my favorite researchers, Enrico Moretti, has been writing about the power of cities to engender innovation. It’s not so much that people come together in the same downtown offices and create new ideas. It’s that their offices are close to other offices, and, when they go out for coffee, or lunch, or to a bar after work, they bump into each other and talk shop. “The chance encounters facilitated by cities,” wrote economist Edward Glaeser, “are the stuff of human progress.”
Tom Wolfe described the phenomenon forty years ago, in a piece about Robert Noyce and the Silicon Valley semiconductor industry: “Every year,” he wrote, “There was a place…where.. the young men and women of the semiconductor industry would head after work to have a drink and gossip and brag and trade war stories about phase jitters, phantom circuits, pulse trains…” And, Wolfe found, innovation would ensue. Face-to-face collisions lead to ideas; ideas lead to innovation; innovation results in human progress.
How much innovation happens in these chance encounters between people from different companies, what economists call “knowledge spillovers”? A study last year from the National Bureau of Economic Research figured out a way to quantify it, using unique access researchers had to cell phone data.
The NBER analysis showed that as more tech employees worked from home, they were less likely to show up in the same bars and coffee shops as others in the same industry. The researchers then looked at the knock-on effects of fewer chance encounters, and were able to calculate that if one half of employees of a given company worked from home, they would have 35% fewer chance encounters with employees from other companies and that patent citations between companies would decline by 12%.
Over the past decade, private companies have increasingly tried to promote more chance encounters between employees of different companies. The Research Triangle Park in North Carolina has moved from encouraging sprawling private company campuses to trying to create a shared downtown “commons.” In San Francisco real estate brokers have watched as more and more artificial intelligence companies cluster next to one another, in what is being called “Area AI.” A recent New York Times article notes that one of the co-founders of the artificial intelligence company Imbue is hosting regular “Thursday Nights in AI” events to try to create collisions.
Freedom and productivity vs. innovation and teamwork
So where are we now. If we want more freedom and productivity, we embrace more work from home. If we want more innovation (and teamwork), we need to find more ways to interact face-to-face with other people.
There’s no easy way forward. Electronic collaboration platforms, which initially positioned themselves as innovation facilitators, have recognized they can’t engineer serendipity. Zoom has basically admitted innovation doesn’t work very well over, uh, Zoom, and has ordered employees back into the office (see this earlier post).
The answer, such as it is, appears to be that we need to make the best of the face-to-face time that remains. MIT researchers argue that going forward companies should organize their floor plans to create “more open, dynamic spaces that encourage the so-called ‘cafeteria effect’” when people are in the office, essentially creating intentional spaces for colleagues with different specialties to interact, as a way of encouraging more innovation.
Well, maybe. But when I hear ideas like this, it’s hard not to be reminded of the government decree that had just been issued when I arrived in China for my first trip in 2008. Having achieved undeniable supremacy in global manufacturing of commodity products, the government declared it was moving on, and henceforth China would be the world’s innovation leader. Fifteen years later and counting, it still hasn’t happened. Innovation doesn’t happen by dictate.
Similarly, you can’t throw a bunch of people in a room every Wednesday for two hours and tell them to bump into each other and innovate. You can’t flip on a switch and be creative. Serendipity is, by definition, serendipitous.
Maybe there’s a new killer app that can facilitate that. Or maybe there’s a new office design that will draw out innovation like a siphon. But until that comes along, I worry about the tradeoffs – can we really have freedom, productivity, innovation and teamwork all at once?
References:
MIT study on decline of “weak ties” when colleagues aren’t together: https://www.nature.com/natcomputsci/volumes/2/issues/8
NYT on the “serendipity” of cities: https://www.nytimes.com/2023/10/20/upshot/remote-work-tech-innovation.html
Harvard Business School article on remote work and innovation: https://hbswk.hbs.edu/item/how-does-remote-work-affect-innovation
Tom Wolfe on Robert Noyce and the semiconductor industry: https://www.esquire.com/news-politics/a19737/greatest-esquire-stories/
NBER study on decline of patents in work from home environment:https://www.anderson.ucla.edu/faculty/keith.chen/papers/Final_NBER22.pdf