Microsoft Research Podcast - 132 - New Future of Work: How remote and hybrid work will shape workplaces and society with Jaime Teevan and Sid Suri

Episode Date: August 12, 2021

For Microsoft researchers, COVID-19 was a call to action. The reimagining of work practices had long been an area of study, but existing and new questions that needed immediate answers surfaced as companies and their employees quickly adjusted to significantly different working conditions. Teams from across the Microsoft organizational chart pooled their unique expertise together under The New Future of Work initiative. The results have informed product features designed to better support remote work and are now being used to help companies, including Microsoft, usher their workforces into a future of hybrid work. In this episode of The New Future of Work series, Chief Scientist Jaime Teevan and Senior Principal Researcher Siddharth Suri explore the many ways people were impacted by work shifts during the COVID-19 pandemic. They talk about how race, gender, income, and other factors are indicative of how people have fared and what this means for the future of work. The researchers discuss the importance of examining potential hidden consequences—and patience—when using short-term data to make long-term decisions, emphasizing aspects of burnout and innovation. Topics covered in this wide-ranging conversation include benefits of commutes and a silver lining in the shift to remote and hybrid work—the movement of more innovative jobs out of large metro areas, creating momentum for greater opportunity in diverse locations. The research that Siddharth Suri describes in this podcast was jointly done with Hana Wolf of LinkedIn. https://www.microsoft.com/research

The people who can't work from home are more likely to be white and more likely to be in the top half of the income distribution and more likely to have employer-provided health insurance, which is something you really need when there's a pandemic raging. Conversely, who can't work from home? They are more likely to be people of color, people in the bottom half of distribution, and people less likely to have employer-provided health insurance. Think of people like bartender, waiter, mechanic, nurse, these kinds of jobs. And so right then, you can see that working from home just sort of clothed society into it, exacerbated already present structural inequalities in society today. Welcome to the Microsoft Research Podcast,

where you get a front row seat to conversations on cutting edge technology. I'm Jamie T. Vann, and I'll be your host as we investigate how work practices have changed because of COVID-19 and what it means for creating a new and better future of work. In this session, we turn to the last chapter of the new future of work technical reports titled Societal Implications. And we're fortunate to have one of the authors of that chapter, Sid Suri, joining us. Sid is a senior principal researcher at Microsoft Research and a computational social scientist. His early work analyzed the relationship between network topology and human behavior. Since then, he's

become one of the world's leaders in designing, building, and conducting experiments on crowd platforms. And in 2019, he published a book on gig work with Mary Gray titled Ghostwork. Welcome, Sid. Thank you, Jamie. It's a pleasure to be here. So, Sid, in this chapter, you and your co-author, Hannah Wolf, cover the research done at Microsoft and elsewhere that tries to tease out some of the long-term implications of the rapid shift to remote work that we've observed this past year. And I mean, that must be really hard. We're trying to see these long-term trends using just a year's worth of data. What are some of the approaches that you use to have confidence in your ability to do this? So first of all, yeah, it was pretty hard.

But I also want to say that it was made a lot easier by my extremely talented co-author and collaborator, Hannah Wolf of LinkedIn. She was an amazing person to work with. And she and I are just looking for excuses to keep working together. You guys brought us together and I couldn't be happier that you did that a lot of what we had to do unlike a lot of the other tracks is we had to get data from the outside world outside of Microsoft so for example shortly after the pandemic hit I literally found myself downloading and analyzing BLS Bureau of Labor Statistics data to understand the disproportionate impact working from home had on African Americans, Latinx people, etc. And the numbers, they kept me up at night, Jamie. The African American population is overrepresented in the education and health sectors. They got

decimated by work from home. The Latinx communities overrepresented in the hospitality sector. They got decimated by work from home. And it's just one of the many examples we have of how this pandemic disproportionately hurt people of color as opposed to those who are not of color. It was sobering. Yeah. Well, and so you bring up an interesting point there. Most of the research in the New Future of Work report focuses on information workers. And what you're talking about right now are different populations of workers. What can you tell me about sort of how work breaks down between information workers and essential workers and others? Yeah. So basically what we found pretty early on, and this is relying on a couple

different economic studies that we can put the references in the show notes if you want, was that who can work from home? The people who can work from home are more likely to be white and more likely to be in the top half of the income distribution and more likely to have employer-provided health insurance, which is something you really need when there's a pandemic raging. Conversely, who can't work from home? They are more likely to be people of color, people in the bottom half of distribution, and people less likely to have employer-provided health insurance. Think of people like bartender, waiter, mechanic, nurse, these kinds of jobs. And so right then, you can see that working from home just sort of clothed society into it, exacerbated already present structural inequalities in society today. And I just described how working from home split society across race, also income. But Jamie, every single dimension we looked at, whether it was

LGBTQ status, disability status, caregiving status, you name it, there was a disproportional impact of working from home along that dimension. You know, so one of the things that I noticed is that some of the kinds of work that we don't historically think of as information work actually did sort of out of necessity become information work during this period? You know, for example, folks are attending exercise classes remotely or musicians are hosting concerts remotely. We're visiting doctors from home. You know, was that something that was observed in the data? Are there implications of this for us to understand? There are. Actually, if you go back to some of those economic studies I was pointing you to, if you look at some of the assumptions in those studies about what jobs can and cannot be done from home, it's really illuminating. So, for example, one of the studies assumed that teaching could not be done from home.

But sure enough, necessity was the mother of all invention. Right. And all school became done remotely and virtually and from home. And, you know, and you mentioned telemedicine. And all school became done remotely and virtually and from home. And, you know, and you mentioned telemedicine. Basically, people have been talking about telemedicine since we were in grad school, you know, decades ago. And it basically got nowhere until COVID hit. And now all of a sudden, telemedicine is an everyday part of our life that we just accept.

So it really did ratchet up the digital transformation. Now, another theme of our work is the unintended consequences of all of this. That sounds great, right? At first, oh, I can see my doctor online. You know, I don't have to go in. I don't have to risk getting COVID, et cetera, et cetera. Conversely, people are putting off a lot of preventive medicine. I myself, I didn't go to the dentist for a year and a half. That's going to make the dentist fun. Oh, yeah. It totally did not make it fun at all. I don't want to talk about that. And so what are the long-term consequences of that? It's yet to be determined. And this goes back to your first question. One part of your first question I want to highlight is a lot of people are making long-term decisions based on short-term data, based on what happened after COVID hit. And I think that is a very, very dangerous place to be. You know, we have to be making decisions now, though, and we don't have long-term data. What

should people be doing? For sure. So, you know, Microsoft, for example, Satya Nadella, after COVID hit, I couldn't have been prouder to work for him. He came out with this quote. It was brilliant. It was like, you know, yeah, okay, productivity hasn't fallen off a cliff. We're doing okay. But what are the effects of social capital? Are we headed for a burnout situation? It's just that he had this long-term vision in mind that we don't quite know what this picture is. And I want a little bit more information before I take these drastic actions of permanent work from home, policies and things like that. And I think that was a very insightful decision on his part. What can you tell us about what's happening with social capital? Yeah. So with Long-Chi Yang, Sonia Jaffe, David Holtz, myself, and a whole team of E&D, including yourself and Brent, Brent Hecht, that is, we spent a lot of time looking at the social and collaboration network of Microsoft employees. And through some clever statistical methods, what we managed to do

is separate out the effects of working from home from the effects of COVID. And the idea here is, God willing, COVID won't last much longer, and we might move to a situation where there's more remote work. So what can we say about that situation? And what we found in a nutshell is that working from home causes the network of information workers to be more siloed. There's less intergroup communication. So one business group, another business group, there's less bridging. So if I'm connected to you and you're connected to Brent, but I'm not connected to Brent, you act as a bridge between Brent and I. There's less of that. So that causes less information flow around the network. So these kinds of things, if you sort of turn that knob and let it go to infinity, you get

a very sort of siloed network where people aren't communicating between teams, between groups. They're not getting that fresh sources of information from other teams and other groups. And it really scares me in terms of innovation and how we're going to innovate. So it's a worrying trend. It's another example of how there's a short-term phenomenon here, but the long-term consequences of which we have not yet seen. So this all seems pretty dire. Were there trends that you saw that made you hopeful, or there were positive things to look forward to in the new future of work?

Yeah, that's a good point. I'm not going to sugarcoat this. I would say the vast majority was pretty dire. There were a few silver linings. Number one would be something around geography. So in the United States today and Scott Counts and his team and MSR are really tackling this issue, a lot of the innovation you see is around big cities. That's where a lot of the high-paying jobs are around information work and things like that. And one silver lining is companies are starting to say, hey, wait a minute, this remote work thing, it's kind of working. So I can pay $160,000 for somebody in Silicon Valley, or I can pay $100,000 for someone who's even more experienced, but lives in Omaha, Nebraska. Why don't I just do that?

And so it's creating a little bit more opportunity for people in diverse geographies than before. Are cities going to look different as a result of that? Like, is that going to change the way that our geography is set up? This is one of the things that I personally am most interested in, and I'm definitely keeping my eyes on as time goes forward. On the one hand, you know, there's the argument I just made. You know, you can hire people in geographies with lower wage costs. Great. On the other hand, people have been predicting the demise of New York City ever since the Spanish flu hit in 1918. So it seems a little premature to do that. So I don't know how it's going to shake out,

but I definitely see a world with more remote work. The question is how big of an effect will it be? Yeah, you know, you've talked a lot about the differential impact of the pandemic on different populations as well. And I know rural Americans are 12% less likely to have home broadband. Do you think rural infrastructure is going to be an issue? Or is this crisis going to force our rural infrastructure to get better? Okay, yeah. And I want to give a special thanks to Bill Buxton to alerting us to this issue, both in the US and in Canada. Yeah, this is a real issue. Broadband penetration in rural

United States is nowhere near what it is in cities. I forget the statistic, but it is actually intersocietal implications chapter. I looked it up. What we see, especially in the education context, because the broadband is so much poorer in rural networks, kids were having a hard time staying in school. I mean, they just literally just can't get online to see their classes. And this has caused almost kind of a wasted year for many kids across rural America. And this is also super, super problematic. And you just made a good point that also in terms of work going forward, if you live in a place without broadband, you're not going to get that high paying information work job. So absolutely rural versus cities is yet another dimension that has cloven society into. Also related in some ways to geography and in particular, the fact that we're not all commuting right now.

I remember like there was this early meme post pandemic where everybody was posting like, you know, polar bears wandering the streets of cities and like nature is healing. I know on the Microsoft campus, we've had bobcats hanging out and like outdoor cafe areas. You know, what can you tell me about the environmental impact of COVID-19? First, I'll tell you about the environmental impact, but then I want to talk a little bit more about commute on humans in a minute. So the environmental impact, I had the pleasure of being on a panel with a woman from UNDP not too long ago, and her name's escaping me. And I asked her point blank about this. Is there any hope that will make a dent in climate change due to not having to commute so much? And what she said is a small dent, but what would actually make it a much,

much bigger dent is if companies would be able to shut off their climate control in their offices because everyone's working from home. That would actually magnify the dent much more. That was pretty interesting to me. Before we get to the commute thing, I mean, another thing that I've heard that's sort of interesting, if you think about people moving remote and then needing to come together still at times. Maybe you have more air traffic, maybe the fact that people are living in rural areas, they have to travel further for groceries. Like it feels like there's these potential long-term externalities to that as well. For sure. That seems, and it can be counter, what you're saying is this can be counterintuitive.

You might just think, oh, I'm not commuting to work every morning. There's got to be a savings here, right? Duh, right? But then what you're saying is everyone moves an hour away, two hours away. Then for that one meeting where I got to drive that hour or two hours, I just made up like two or three weeks worth of commute in one shot. So it's a little less clear as you might think. But there's an added cost that I also want to bring up, which is like, if we're all apart, then are we innovating? Yes, I'm saving that commute time, but are we innovating? Am I contributing to my company in the same way? Am I contributing to society in the same way? So yeah, I saved on commute time, but maybe my output is also down. So how does that net out? This is really, really, really complicated accounting.

Yeah. So then you said you were going to say a little more about commute. I didn't mean to cut that off. I just got excited about the environmental stuff. I'm excited about both and about the commute. So one of the key themes that came out of the, it was the software development track, Chandra Madilla and Ginger Hudson, one of the phrases that came out was productivity is personal. And this is a key scientific insight that came out of our whole initiative, but especially that track. And what they mean by that is the same phenomenon can help one person's productivity and hurt another's. So let's take commute, for example. So they surveyed people, Denae Ford and her team, they surveyed people. And a lot of people were saying, look, I don't have to commute. I get that time back. Boom, I'm more productive. But then other people were saying, look, I used my commute to work to plan out my day, to think about how I was going to allocate my time to what things on my to-do list. And then when I'm on the way home, I would use my commute to decompress,

to switch out of work mode and switch into home mode. It just never crossed my mind that commute could possibly have a positive benefit to anyone ever. And it turns out it does. And then, you know, Shamsie Ekbal, your longtime collaborator, and you and, you know, your team built this thing called virtual commute to get the good bits of a commute back, but leave out the bad parts, which I thought was a brilliant piece of innovation. So like I said, the same phenomena can boost one person's productivity and can take from another. Yeah. Can we look at the scale of the shift to remote work? Do you have a sense for how many people were working remote pre-pandemic and then how many people were working remote post-pandemic? So I recently had to give a talk about this around MSR, and this is

the first slide. So if you rewind in your minds, everyone on the audience, just rewind in your mind back to February of 2020. At this time, less than 5% of working Americans worked from home three or more days a week. Nobody worked from home, really. Like that was not many. Yeah. Exactly. Exactly. Now you fast forward to March or April. 37% of Americans were working from home five days a week.

So in the course of a few weeks, a third of working Americans went from basically never working from home to working from home five days a week. I would argue that is the most significant disruption Microsoft has ever seen. It's an entire history. Never has there been a disruption that affects a third of information workers in a matter of weeks. Yeah, it's a third of all workers, right? And most of those are information workers.

Yeah. Yeah. Are things going to go back to normal? Normal, like in quotes. So things have been really different this last year, but like often these huge disruptions end up like in retrospect, just being little blips. So Abby Sellen made a wonderful point. Another one of our illustrious collaborators, you know, before COVID, basically, like I just said, less than 5% of people were working from home three or more days a week. So basically, almost all work was being done in the

office. During COVID, like I said, almost all work is being done in the office. Now, the future is going to be almost, by definition, somewhere in the middle. And prominent economists like Nick Bloom from Stanford, he predicts two days a week at home and the rest in the office. I actually kind of believe that. I don't believe we're just going to go back to February 2020. I think we're going to go to some new kind of equilibrium with more remote work than this sort of hybrid model where some people are full-time at home, some people are part-time at home. Do we know what people want? Like, do people want to be working remotely? Do people want to be back in the office? So our Asia lab is ahead of us in the sense that they've already returned to the office,

so to speak. And if you look at their usage of how many days a week on average people come into the office, it's pretty much uniformly distributed between one and five days a week. Economists would call that a revealed preference. Secondly, there's been surveys done of Microsoft employees. They predominantly are preferring, again, some kind of situation in the middle, not full remote and not full in the office, but something like two days a week at home, three days a week in the office. Do you have a sense for how business leaders are starting to prepare for hybrid work or sort of looking forward to the future for what they should do? Or do you have advice for business leaders as they move forward? Well, in terms of what business leaders are actually doing, I'd actually push that question

back to you, Jamie, because you got more of the ear of the business leaders around Microsoft than I do. So I'd push that one back to you, Jamie, because you got more of the ear of the business leaders around Microsoft than I do. So I'd push that one back to you. One piece of advice I would give would be, like I said, don't fall into that trap. Don't make long-term decisions on short-term data. Think about the data you're staring at. Think about how it was gathered. And think about, is this the right thing for me to be extrapolating from? And I think thinking like that is going to save you a lot of heartache going forward. Do you have any observations from what the business leaders around Microsoft are thinking about? Yeah, certainly you see across the tech industry, a range of things that people are doing, right? You have some companies moving to remote first

models. Microsoft has taken, I think, a good thoughtful measured approach from what I can tell. So we're looking to support folks working from home up to 50% of their time, which seems roughly consistent with the literature I've seen that suggests you can work, you know, two and a half days from home without deteriorating your social networks. It's interesting to look at how folks are revisiting space and rethinking the workplace and what space looks like and how we're using our in-person, in-office time. Right. Just before we get to the space, there's the one bit I want to highlight is Microsoft's more measured approach, which we brought up a few minutes ago. And one of the reasons I love

that approach is because, and Shamsi and Mary Cherwinsky and their team, they did a survey about this. And they found out that women were struggling more than men. And moreover, people with caregiving responsibilities were struggling more than those without. And when you put that together, women caregivers were struggling the most. And this more measured approach, I think, is the right way to go. Right now, women caregivers are really, really struggling. So if we can take this more measured approach, maybe we can figure out ways to relieve some of the pressure off of them and get them, you know, sort of back to equal.

What I don't want to see is the wage gap, which women have worked tirelessly for the past few decades, at least to get it to 88 cents on the dollar. I would hate to see that work come undone. So that's one thing I really like about this measured approach that Microsoft's taking. Now, I certainly know as a mother to four kids that it's pretty hard to focus on work with schools closed. You know, I want to be helping my kids manage through this crisis. I think that's a great example to your point about making a short-term, it's actually almost the opposite. It's like making a short-term decision to address a short-term issue, but it will have long term implications to the extent that women scale back the work

they're doing or leave the workforce. Exactly. Exactly. I'm actually afraid that wage gap is going to close because the women will start leaving the workforce and it would actually close for the wrong reason. Yeah, that's interesting. That speaks to your point of really looking carefully at the data as well. Well, is there anything that I forgot to ask or an important point that you want to make sure we hit on before we sign off? I guess the important point I would like to make before we sign off, like I said, first and foremost, I want to give a ton of credit to my collaborator and co-author, Hannah Wolf. It was a pleasure working with her. She was absolutely brilliant. Secondly, I want to give a thanks to you, Jamie and Brent, you know, all the leadership work you guys did. It was just a, honestly, it was kind of fun and it was a ton of work, but it was just super interesting. There was a time I'll never forget it. Like Hannah and I were on a call. We're like, you know, neck deep in BLS data or something. And she just told me, she just chuckled at myself. She's like, every time I opened the New York Times,

she was like, Sid and I figured that out like a month or two ago. And it was because of your leadership, it's because of Brent, because of all the people you guys brought together that, you know, we had all this data in our hands. We were swimming around and we could draw these insights and we were sort of ahead of the game. And it was just a ton of fun to work on. Well, great, Sid. Thank you for your time today. And thanks to our listeners for tuning in. You can learn a lot more about the research that we discussed today at aka.ms slash new future of work. Also, be sure to subscribe for new episodes wherever you listen to your favorite shows.