In this episode of the McKinsey Global Institute’s Forward Thinking podcast, co-host Michael Chui talks with Nicholas Bloom. Bloom is the William Eberle Professor of Economics at Stanford University and a senior fellow at the Stanford Institute for Economic Policy Research. He covers topics including the following:
- The history of remote work
- How remote work can enhance productivity
- What practices make hybrid work most effective
- How the views of employers and employees compare on working remotely
An edited transcript of this episode follows. Subscribe to the series on Apple Podcasts, Amazon Music, Google Podcasts, Spotify, Stitcher, or wherever you get your podcasts.
Michael Chui (co-host): Janet, you and I have worked together for years, even before we started doing this podcast together.
Janet Bush (co-host): Yes, it’s true. We do this remotely, and actually we have only met each other in person a handful of times. So we have been remote workers for a long time.
Michael Chui: Remote and hybrid work is something that today’s guest has been studying, including what works and what doesn’t. I was surprised by how much remote work was going on even before the pandemic.
Janet Bush: Yes, and I realize that I have been working from home, remotely, for 20-odd years. I worked from a farmhouse and that’s longer than I have been working for McKinsey, so it’s very familiar to me. But I am fascinated to know what our guest says about what works and what doesn’t work.
Michael Chui: Nick, welcome to the podcast.
Nicholas Bloom: Thanks very much for having me on.
Michael Chui: Let’s start with your background. Where did you grow up? What did you study? How did you end up doing what you’re doing today?
Nicholas Bloom: It’s a bit of a random walk. There’s that old thing about a random walk through Wall Street; feels like a random walk through a career. You can probably hear I’m British. I grew up in London. I actually never planned to be an academic, really.
I worked in this place called the Institute for Fiscal Studies, which is a fantastic think tank in London, very policy focused. Ended up doing a PhD at UCL [University College London] because I was there. And then actually briefly worked in the UK Finance Ministry for a budget cycle, but worked at McKinsey for a bit.
And after a while at McKinsey—I enjoyed it but, you know, it’s stressful, long hours. I guess some of that hasn’t changed. But I ended up returning to academia and worked in the LSE [London School of Economics] and have been at Stanford for now almost 20 years.
Michael Chui: You’ve covered a lot of topics in the research that you’ve done. But one of the things that you’re very well known for is studying how the world of work has been changing, particularly various forms of remote work. So maybe we’ll talk a little bit about that, and hopefully we’ll have some time to cover some of the other things that you’ve worked on as well.
A lot has happened over the past few years, particularly during the pandemic. But you started looking at this even before the pandemic started, so maybe we’ll start there. What were you starting to find about how work was changing, even before 2020?
Nicholas Bloom: I did a couple of surveys and a large, randomized controlled trial. And we found that work from home makes people happier. They typically don’t want to work five days a week, but most people really want to work two, three days a week from home. That’s still true now, as true as it ever was.
And also, if it’s well organized and well managed, it can improve productivity. The challenge was, prepandemic, hardly anyone did it, so there wasn’t really an ecosystem, the technology. Certainly if you go back to when I first started looking on this, in the early 2000s, we didn’t have cloud, which enables things like Dropbox file sharing.
We also didn’t have Zoom or Teams or anything. It was telephone calls and emailing files and sending stuff through the mail, so it really wasn’t great. There was always the potential. I mean, Jack Nilles talked about this in the ’70s.
I recently put out data, you can measure it, from the American Community Survey and other census data going back to the ’60s. Even in 1965, 0.4 percent of people said they worked from home and didn’t commute, and that was growing. So that was 5 percent in 2019. That’s already about a 12-fold increase, running up to the eve of the pandemic.
But then of course the pandemic happened, and that was an explosion. It went from 5 percent to 60 percent in the space of a couple of months. And it’s settled back down to 30 percent of days now are worked from home in the US. And it looks like that’s kind of where it’s ending up now.
Michael Chui: But actually, quite remarkable. That 5 percent figure, prepandemic, what’s the denominator of that? Is that the entire workforce? Does that include all the manufacturing workers, agricultural workers, etcetera? Because that’s one in 20.
Nicholas Bloom: To give you a view of 2019 and now, in 2019, 5 percent of full pay days in America were completely from home.
Who was doing that? About half of those days came from people that were fully remote, a lot of people in call centers, data processing. Some people, like a friend of mine, she moved from Stanford to NYU in New York, and her husband kept his job at Cisco and just moved remote. And then the other half of it, people who did the odd day here and there at home, but not many people.
And then if you look at now, half the workforce, to be clear, can’t work from home. In fact, 55 percent in America do not work from home ever. They’re folks in McDonald’s, Chipotle, in hospitals, teaching, etcetera. The other half—probably most of your listeners will fall in that bucket—are mainly hybrid. That’s 30 percent. And then the remaining 15 percent are the fully remote.
Everyone thinks fully remote is, like, Airbnb. When I talk to execs, particularly, they’re like, “Oh, that’s elite coders.” And yes, there are some of them, but most of that 15 percent is back to call centers, payments processing, HR, benefits.
Michael Chui: So where does the productivity benefit come from? You saw increased productivity, even prepandemic. Why is that true? Because many of us who had the work-from-home experience know that there are a lot of things that can also get in the way of work when you’re working from home.
Nicholas Bloom: Absolutely. To be clear, the evidence—there’s a few randomized control tests I’m not going to go through, I’ll just summarize—are the, what I’ll call, well-organized hybrid.
That means, say you are doing what Zoom does, for example. You’re going to come in the office Tuesday and Thursday. Everyone comes in. You have meetings, social events, presentations, trainings. There’s a good reason to come in. All your face-to-face stuff is pushed into those two days.
On Monday, Wednesday, Friday, you go home. A typical firm is doing something like that with, say, three days in the office, normally Tuesday to Thursday, and two days at home.
The two benefits on productivity are, (a) if you look in the data on home days, Americans—and in fact Europeans as well—save about 70 minutes a day from less commute, less time getting ready. And about 30 minutes of that is spent working more.
So if you are an employer, and you have an employee that works from home two days a week, they work about one hour more a week for you. And in a 40-hour week, that’s about 2.5 percent. So that’s game one there. You could call that productivity or more hours, depends how you count commute, but they’re working more.
But they’re of course getting a benefit, too. The employees are getting an extra 40 minutes a day, on average, they spend on child care, leisure, whatever, other things. And then the other benefit for organized hybrid is the activities you do at home on the Monday and Friday, say, they’re quiet or deep work, so reading, writing, preparing presentations, writing documents, thinking about stuff, maybe one-on-one.
So, Michael, me and you having a one-on-one now, this actually works fine, maybe it’s even easier, where there’s peace and quiet. It turns out that deep work or quiet work is better done in a quiet environment. And again, from our randomized control trials, you see productivity uplifts of often 2 to 3 percent per hour from quiet.
Add those two things together, you’re getting something like 3 to 5 percent improvement in productivity. But again, to be very clear: (a) this is not talking about fully remote; I’m talking about hybrid. And (b) I’m talking about well-organized hybrid. So it’s not the nightmare of 2021 where you’re in the office, I’m at home; I’m in the office, you’re at home. It’s where everyone comes in on the same anchor days, and everyone stays at home on the same home days.
Michael Chui: Is there a productivity boost for fully remote or not?
Nicholas Bloom: Fully remote is very different. I’ve not seen studies—there isn’t so much of a direct study. So I’ll tell you the ups and downsides of fully remote. Well-organized hybrid seems like a win–win. So just to be clear, if you’re five days in the office and you go to three in the office, where everyone comes in, and two at home, that seems like almost all good.
The evidence is, if it’s well organized, there’s pretty much no major downsides. Fully remote has more major upsides but starts to incur some big costs. The two big positives for fully remote that I hear over and over again from execs is, one, you save on office space. You can basically get rid of it, and that’s 20 to 30 percent at cost. That’s a big benefit.
And secondly, you can hire at least nationally, or if not internationally—depends on the way you’re set up. Particularly think of your tech firm out in Silicon Valley. It’s very expensive to hire people here. You can hire certainly across the US, and I remember talking to one exec and they were saying, “We’re hiring in Nigeria and Bulgaria and Mexico.” Just globally.
So those are the two benefits, and you can put numbers on them. But those are big, big numbers. You may reduce your total cost, say, by 50 percent.
The two downsides I hear a lot are, one, about mentoring. Mentoring is something that’s very much better done in person. And the fact that it’s hard to mentor also will make it harder to recruit.
If you look at 20- to 29-year-olds, they have a very strong preference for having at least two, three days a week on site. Strategically, as a company, if you want to hire people like my students from Stanford, it’s not that appealing being fully remote. So the mentoring thing means you mentor less, but you’re also going to struggle to hire people under the age of 30 because typically they’re not the ones that want fully remote jobs.
The other cost is around creativity. There’s some pretty good research on that, showing that you certainly can be creative when you’re in a remote team, but it’s harder. There was a piece in Nature a few months ago, it was a somewhat staged experiment. They randomized whether teams are in person or remote to come up with ideas for some products they’d randomly given them, and then they assessed it with some separate panels. And found, look, people that are discussing things like that in person tend to have better outcomes than people remotely.
Again, I don’t think anyone’s going to be that surprised by that finding two years into the pandemic.
Michael Chui: But creativity, that is, for some reason, being in person sparks more ideas or better ideas?
Nicholas Bloom: So many different people from very junior to senior, across—it’s not just firms, like UN, the World Bank, the IMF, the British government, my local city council, my hospital. All of them say the same thing, that it’s just easier to have some of these problem-solving meetings or creative things in person, partly because some of this stuff actually happens at the beginning or the end of the meeting where people are walking out. Partly maybe it’s more relaxed, you take your time a bit more on it.
People are also paying attention. I notice it with students in particular. Anyone that wears glasses on a Zoom or a Teams call, I often see the colors on their glasses changing. And I’m like, “You’re watching Netflix,” you know? “You’re so not paying attention,” or “you’re doing an email.” Or you can see people’s eyes, they’re looking. So that’s the problem.
You have a problem-solving meeting with eight people, and three of them are doing email, and so, not surprisingly, their contribution’s not as good. So, yes, the evidence is there’s a few different pieces of research on that. Creativity seems to be better in person.
Personally, in research seminars and academic economists, they’re a lot better in person because everyone’s in the room, they’re focused. Having a discussion on Zoom, it feels like reduced bandwidth, you know? You have no idea. You have a bunch of faces that are there. Some of them have their screens off, the camera’s off. Creativity is the thing that a lot of people have bemoaned, actually, particularly in 2020 when you were fully remote—that it was hard to come up with new ideas.
Michael Chui: Are there other characteristics of the meetings that are better done in person?
Nicholas Bloom: If you look at really fully remote companies, so think of folks like Upwork, Automattic, Quora, these are companies I’ve spoken to a chunk about stuff. They do meet up in person every so often.
Normally, what they’ll do is meet up for, let’s say, a week every other month to try and build some connection. So fully remote is not normally completely you never meet the person. It generally means on a day to day, you’re meeting remotely.
In terms of meetings, I think one-on-one meetings—like, Michael, if you and I had been in the office yesterday and we’re on Zoom or Teams today, it probably doesn’t matter that much. We have this connection, particularly if we’re sharing documents, it may even be better to meet online because you can both look at a doc. Now, I’m kind of shortsighted, so I can peer at it very closely into the laptop without invading someone else’s space, and fiddle around with it.
On the other hand, big meetings where there’s—I have some survey evidence on this, but larger meetings, particularly ten plus, they’re not enormous. They’re ten to 50. Town halls with 2,000 people, obviously, it’s very hard to do this in person. But if it’s a ten-, 20-person meeting, they seem to be better in person, partly because everyone’s just forced to pay attention.
In a 20-person meeting, most of the time, people aren’t speaking, they’re listening. And it’s more of a shared experience. And people are focused. It does mean that you should keep them short. So, to be clear, online, you see people spend more time in meetings. On the other hand, they’re not really fully there. I mean, they’re doing email and other work. In person—it is costly to have a 20-person meeting in person. You want to limit these things. But it seems like everyone’s attention is kind of efficient in that sense.
Michael Chui: Although sometimes you see some meetings and clearly people are still doing email, despite being in person as well. So that happens, too.
Nicholas Bloom: It’s amazing, in tech, the norms vary a lot across industry. I’ve presented at conferences in tech world, and everyone’s on their laptop. And, you know, it is what it is. If you can’t grab their attention, you know they’re coding or doing email or something else.
Michael Chui: Indeed. I was once doing a presentation at a tech company, and I actually thought everyone was doing their email because it was early in the morning. It turns out that they were actually discussing the topic that we were talking about. It turns out it wasn’t as negative as I had thought at the time.
There’s an old saw about, remote is good for maintaining relationships. But in order to start a new relationship that’s trust based, you need to be in person. I’ve heard people say during the pandemic that’s been challenged a bit. That, in fact, people have been able to start trust-based relationships on a remote basis. Any reflections on that? Has that changed?
Nicholas Bloom: I agree with it. I haven’t heard hard data on it but, again, I’ve heard this over and over again. Just as another great example, Stanford has this very elite exec ed program for directors of companies. These are very senior people, a bunch of them are CEOs of companies you’ve even heard of, very, very senior.
I gave a session, and it was a fascinating discussion about board meetings. And folks saying, “Look, you really want remote.” Board meetings, they’re four times a year or six times a year. You can have one or two of them remote. But you certainly want enough, a regular cadence of in-person meetings, particularly the first time you join, because you really get that sense of connection and building trust, as you say.
You could say we’re evolved over two billion years if you go back, whatever, to the first single-celled organisms, but certainly humans as we currently exist, tens of thousands of years old. And I think to set up our hardware and software and the way you think about us is about in-person interactions.
I personally find it easier to certainly connect to people in person. I have a number of coauthors that I started working with on work-from-home topics during the pandemic, and a bunch of them, I met. We had a recent remote-work conference. And it was really nice to meet a bunch of my coauthors. And I do feel more connected. You know, you’ve met them for the first time and you have lunch with them and you chat about stuff, just different things.
Again, it’s why hybrid has become so dominant. I don’t think you need to be fully in person. But even if you’re remote, you still need some points of contact. And coming back to if you look at, just about every company I’ve spoken to that hires graduates—so professionals, managers, basically your listener group—even if they’re fully remote, they still try to meet up every other month.
They say, like, “We’re going to meet in Albuquerque,” or Barcelona, or something, for a four-day off-site retreat. “We’re going to use it mainly for work, but we’re also going to use it, to be honest, to bond people and build culture.”
Michael Chui: You’ve mentioned several times, for this type of hybrid work, that it’s important that it’s well organized. What are the characteristics of being well organized?
Nicholas Bloom: One, people are in on the same days. If you survey people—again, we’ve been surveying. I have something called the Survey of Working Arrangements and Attitudes, SWAA. We survey 10,000 people a month. We’ve been doing this for over two years. We’ve talked to 100,000, 200,000 people.
If you ask them, “Why do you want to come into the office? What are the benefits?,” numbers one and two by a long margin are, (a) coworking face to face with other people and, (b) socializing. One of the options is “to spend time with my manager”—and I can assure you, that doesn’t get many hits—or to use the equipment, or the free bagel, or the ping-pong table, whatever it is.
If people are coming in for three days a week and they’re spending one of those days sat in the office, quietly working, let them work from home. The function meets purpose here. Once you get it right, I know quite a lot of organizations have done that, generally the employees are quite happy. They see real purpose in coming in. They find it’s beneficial. It’s social, it helps them. And then they get to benefit two, three days a week from home.
Michael Chui: Do you get the real estate benefits, though? I guess you still need the same amount of space, it’s just utilized less.
Nicholas Bloom: Yeah. Before the pandemic, when I talked to folks about work from home, they always focused on real estate as the big benefit. That was really where the savings were. Postpandemic, it was actually the last of the four benefits.
If I go through the four benefits, number one, by far, is keeping employees happy. The numbers I can put it in are typically, if you let people work from home two days a week, they value it at about an 8 percent pay increase, which is an enormous amount of money.
Just to context that, in the US right now, the employment cost index is going up by about 7 percent a year. If you’re listening, as an employer, you have to have a pay increase of about 7 percent just to tread water. If you then want to say, “I’m going to force everyone back into the office full time,” and your rivals are letting them work hybrid, you need to throw on top of that an extra 8 percent to keep them whole.
That’s a 15 percent pay increase. Nobody is doing that. Therefore, that’s why hybrid has become dominant. It’s the most cost-effective way to retain and recruit staff.
Number two is productivity. We talked about that. These numbers aren’t huge, but they’re maybe 3 to 5 percent.
Number three is on diversity. I won’t spend too long on it, but if you look at who’s most keen to work from home at least one, two days a week, it tends to be minorities in the workplace. To be clear, it’s not just—people immediately think of race and gender, which is true. But interestingly, it’s also by age, by political preference, by religion.
We survey folks, and if you have less than 10 percent of your coworkers in that cell, you say, “I feel slightly less comfortable coming in.” Like, I’ll be 50 next year. And if my office was full of 20-year-olds talking about TikTok, I’d be like, “Maybe this isn’t—I don’t feel that comfortable coming in. I don’t want to come in every day of the week.”
And it’s true for minorities across multiple dimensions. And what it tells you is, if you get tough and force folks in five days a week, you can imagine who’s most likely to quit, and that’s costly for DEI. And then, finally, space.
Michael Chui: Just to pause on that one for a moment. We’ve heard from another podcast guest about this critical mass idea, that you need a certain amount of people with your identity, gender, what have you, in order to feel comfortable. You mentioned 10 percent. Does that number change when you’re remote? Is that what you’re suggesting?
Nicholas Bloom: The 10 percent, to be honest, I’m definitely not an expert on this. I picked the 10 percent because we literally put in a survey question, we asked people, “Are more or less than 10 percent of your coworkers the same gender, age, race, religion, and politics?” Then we asked people subsequent questions on, “How comfortable do you feel coming in?” “What would you do if your employer forced you to work from home?”
And what you see is people that are in minorities are less comfortable coming in and more keen on work from home. It also tells you, by the way, that you risk a bit of a downward spiral because what happens is, if you start to lose diverse employees, it makes those that are left feel less comfortable, because they’re more outliers. So there is a real stability thing.
When I talk to firms, it’s one of the clearly important factors that, particularly folks in tech and finance that are struggling with this, some business services. They’re like, “Look, if we’re very tough and force people back to the office, some of the senior managers want that, but it’s going to come with a massive cost in terms of loss of diversity.” And so that was factor three.
I was going to end on space. Space is the biggest disappointment. Every exec is kind of frustrated with their space folks, saying, “Why can’t you cut space?” And as you know, and everyone’s kind of figured out, is of course if people are coming in on Tuesday to Thursday, you can’t do it.
Now, there are solutions. There are kind of clever software solutions. And I think that’s where we’re going to be in ’23, ’24, onwards. But to be clear, what it needs is that you do what we do for, like, university lecture halls, which is we rotate.
So you need to tell Team A, “Look, you guys are going to come in Monday, Tuesday, Wednesday.” “Team B, you’re Tuesday, Wednesday, Thursday.” “Team C,” etcetera. And the cost of that, (a) people who have to come in on Monday and Fridays, they don’t like it, (b) they’re not in on the same day. So if Team A and B want to work together, it’s less overlap.
There’s also another issue about clean desks. You can’t leave stuff out on your desk. There’s a number of complexities. Now, I don’t think they’re insoluble by any means, but it is complicated. And so most managers I speak to are saying, “This hybrid, it’s hard enough. It’s so complicated getting this thing right, and there are so many different views. And some people want to be fully remote, and some fully in person. I’m just focusing on getting this thing nailed down, I just don’t want to deal with the space headache as well.”
And, in fact, a few companies I’ve spoken to cut space in 2020 and 2021, and are now telling me they are struggling to get people back. And in fact, it’s become a real pain for management because they’re like, “We want to get people back to a normal hybrid three/two plan, but we don’t have enough desks. And we’re thinking about taking on more space, or shrinking, or repurposing,” etcetera.
Michael Chui: You mentioned the SWAA data set, the survey that you’ve been conducting, which is an amazing data set. I’ve also seen that you’ve expanded that to a global survey as well. How are you seeing things vary by region, or is it the same all around the world?
Nicholas Bloom: Really interesting question. North America, Canada, the US, look pretty similar, pretty high levels. Northern Europe is the same. As you go into Southern Europe, work-from-home levels are lower. Australia, New Zealand, what I called developed Asia, so think of, you know, Japan, South Korea, Taiwan, Hong Kong, etcetera—lower.
India, China, all of these places are lower, but their proportionate increase has been pretty similar. So in the US, you’ve maybe gone from 5 percent to 30 percent of full paid days’ work from home. In Africa, it’s maybe gone from more like 2 to 10 percent. But it’s still a very large increase. Around the world, we see big jumps up.
The reason that the levels are particularly high in America is a combination of (a) industrial structure. So it tends to be professional services, tech, finance is more amenable to this. Manufacturing, agriculture’s harder. (b) Infrastructure. In a lot of Asian cities, like think of Tokyo or Shanghai, people don’t—it’s so expensive, they live in small apartments. And it’s not so appealing, particularly for a couple, to work from home.
And then, (c) in some countries, I remember talking to a multinational with folks in Indonesia. They said in some of the areas in the smaller towns and cities, there just isn’t a good enough connectivity to connect to support Zoom. Email’s fine. But to have video calls, it’s not just high average speed. You actually need it never drops below about five megabits per second, otherwise you disconnect. In some of these places, it’s fast and then flakes out and then fast. And it just makes it very hard to connect properly.
Michael Chui: That makes sense, in terms of how things are evolving in different places. Another piece of data that came out of your research with the survey is a little bit around maybe a mismatch in expectations. Some of our MGI colleagues have looked at what’s possible. Taking some of the Bureau of Labor Statistics breakdowns of different occupations and trying to assess which activities have the potential of being done remotely at roughly the same productivity. And it estimates that more than 20 percent of workers could work remotely at least three days a week.
But some of your survey research compares what workers would like, in terms of number of days remote, and what employers would like. And you mentioned, for instance, that employees actually value the potential. What were those findings? And what are your reflections on maybe a bit of that mismatch?
Nicholas Bloom: Let’s break employees down into two groups. There’s roughly half of the population in the US and most of Europe which can work from home. It’s professionals, managers. For that group, on average, they want to work from home two and a half days a week.
There’s a big spread, though. About 30 percent want to be fully remote. You should think of 30-, 40-somethings with young kids. There’s about 15 percent that want to be fully in person, actually. They tend to be in their 20s. They want to get mentoring, they want to go socialize. They have small apartments. They don’t want to spend much time in them.
There’s a big variation, but the average is two and a half days. If we ask their employers what they’re planning for this group, in 2020 it was about one day. At the beginning of the pandemic, there was a lot of discussion of employees saying, “We like this; it’s really great,” and employers saying, “Well, don’t get used to it. You know, you’re coming back. Once this thing is over, you’re back.”
By 2022, that gap’s gone. Month by month, what employers have been telling us has been going up and up and up. So for the half of the population that’s professionals and managers, both sides are on about two and a half days a week. Now, it still means there’s a lot of griping.
If you look at Apple, they’re letting employees work from home two and a half days a week. That’s kind of what the average person wants. But remember, 30 percent of Apple employees, based on our survey data, probably want fully remote. There’s 150,000 employees in Apple. That’s almost 50,000 people.
Not surprisingly, a bunch of them are signing petitions. So the fact that you’re offering the average doesn’t mean that everyone’s happy.
The bigger group that, in some ways, is harder is the other 50 percent that can’t work remotely. They want less, but still in our survey we asked them the question, “How many days would you want to work from home?” And it’s hard to interpret it slightly, if you work at McDonald’s and you say you want to work from home, quite how to interpret it. But they’re giving us numbers about one and a half days a week, and they’re getting basically zero. So in some senses, that’s the biggest gap because that group, it’s very hard for them to work from home at all, but they’d quite like to.
Subscribe to the Forward Thinking podcast
Michael Chui: You also characterize some of that 50 percent as being higher-wage workers, which is managers, professionals, etcetera. And I think that’s easy for us to conceive of. But earlier on, you also mentioned there are a number of lower-wage occupations which are also done, in some cases, fully remotely. Say, contact center work, for instance. How has that evolved, if any—as we think about not just the highest-wage workers in the economy, but some of the lower-wage occupations which end up being remote?
Nicholas Bloom: To be clear, there’s every which way. For example, some of the most paid people out there, think of traders in, I don’t know, Goldman Sachs. These folks basically have to be in the office because of the equipment and the speed of connectivity.
Or brain surgeons or heart surgeons or airline pilots—no one would be happy to be told as you board your plane that your pilot is working remotely from his living room in Ohio. You’d be like, “Uh, I’m sorry. I’m getting off right here and now.”
There are a number of elite jobs that have to be in person. But you’re correct that if you look at people that are fully remote, they don’t tend to be—apart from Airbnb and a few big tech or smaller startup firms— a lot of them are call centers, payroll processing, HR, benefits.
The thing that, if I were those employers, that makes it slightly nervous, in some senses, it’s great that they can do that. At Stanford, for example, some of these people are no longer in California. They’re moving to other parts of the US, and it’s good because we’re providing employment across the country and spreading it a bit out from the Bay Area.
But a number of companies I talk to are saying, in the longer run, we’re thinking about why they even have to be in the US. And currently, because we’re short of labor supply in the US, it’s a good thing if we can effectively use labor supply from other countries—if folks in, I don’t know, Mexico or Chile or whatever are working as payments processing for American companies.
Of course in the long run, as we know from the whole China world trade debate, there could be some political ramifications of it. But for businesses, there is an enormous opportunity in terms of outsourcing and offshoring. And I hear a lot of companies saying, “The pandemic made me realize there were parts of my business that don’t really need to be my employees.”
Why does IT support, if it’s frontline and it’s done entirely remotely and has been for two years, maybe there are better firms that could do this at scale. They don’t necessarily need to be in the US. There are various confidentiality and other positional issues, but I think in the next five years, we’re going to see a big increase in service-sector trade, service-sector globalization driven by this.
Michael Chui: One other thing you mentioned—we’re both based in the San Francisco Bay Area. There’s a lot of tech going around here. “Metaverse” is a term that comes up all the time. You’d mentioned that remote channels of bandwidth for communication are more constrained than in person. Is there a potential that we put on the goggles and we can see in 3-D? And maybe we have legs, maybe we don’t. Does that change what’s possible, or what do you think?
Nicholas Bloom: Yeah, absolutely. Another thing that’s really striking is the role of technology. Not so much in the short run, but definitely in the long run. I’ve been working on working from home for 20 years. When I first started working, it was doing stuff over FreeConferenceCall.com, phoning people up, emailing files. It was not great. You know, it was really honestly not that good.
About ten years ago, looking at 2012, 2013, the cloud and things like Zoom and Teams, or its forerunner, Skype for Business, really started to take off. And that meant that you could work much more effectively. So technology has really saved us now.
I have another project where we scrape the patents that come out of the US Patent and Trademark Office. There are about 20,000 a month, and we scrape them for words like “remote work,” “working from home,” etcetera. That was about half a percent running up to the pandemic; it’s started to rise pretty sharply, and it’s gone above 1 percent now and is heading rapidly up.
The reason I think it’s really positive is, like, every hardware and software firm out there has said, “Look, the number of people working from home in America, the number of days, has gone up five-, sixfold.” That’s that 5 percent to 30 percent number. That’s a much bigger market. We should pour a ton of R&D dollars into it. A lot of start-ups are entering.
The rate of technological progress is really going to pick up. What that technology is, is harder to call. To give you an anecdote, I heard an interview with the founder of Dropbox about a year ago. And he was saying, “Look, when I founded it”—maybe it was, I think, 2009—“only techies had more than one laptop.” But it’s turned out that that kind of technology has been invaluable for file sharing in Box and everything that’s come after it.
So I think currently now, there’s a lot of technology, as you say, and maybe metaverse, maybe holograms, virtual reality, better screens. For example, cameras all over the screen. Right now, all of these cameras are at the top of the laptop screen and you don’t really have eye contact. But you can imagine miniaturizing them and burying them into the screen.
This kind of stuff will make remote work more appealing. Three, certainly five, for sure ten years from now, the technology will be substantially better. It’s not that we’re all going to go fully remote. But at the margin, we may creep from two to three days a week. More professions may go remote.
Another example is that one of my neighbors is a doctor. And she said prepandemic, five days a week seeing patients in the hospital. Now she has one day a week of remote visitations for people that want repeat prescriptions or tests back. Some patients would rather that. And so that’s the sense in which I think technology’s facilitated that. She wouldn’t have done it over the phone but now, because of regulatory changes and because you can do it over a secure video link, that’s now become possible.
Michael Chui: Certainly changed the game in terms of telemedicine there.
One other thing I want to pull on, too, is we’ve been talking about hybrid with a particular lens, which is the organizational lens where sometimes people are remote, sometimes people come together in person. But there’s another way to use the term “hybrid,” which is, for instance, in a meeting where you have some people in person and some people remote. And I think a lot of organizations have found that to be particularly difficult.
Part of it has to do with hardware, as you just mentioned. Any reflections on some of those challenges? How do you make sure that people who aren’t in person can participate in the same kinds of ways, etcetera? And what the future might be for that type of hybrid.
Nicholas Bloom: A few points here. You’re right, particularly given the technology and hardware that exists, you have to be realistic about where we are. Not every company’s dramatically overhauling their meeting rooms. That’s not ideal.
So, yes, it’s best avoided if possible. One of the things—for example, I work with Lazard. It’s one of the big investment banks. And they try and be fully in person Tuesday to Thursday internationally, and remote Monday and Friday. That means if you’re going to have a meeting with five people across five different offices, it makes sense to do it on a Monday and Friday. Everyone’s at home anyway. You’re on an equal playing field. It’s easier to do it.
If, on the other hand, you want to have ten people from two different offices, it’s maybe hard to get around it. And you have to take a call whether you want 20 small screens or two people on a Tuesday where there’s ten in one room and a big screen, and ten in another. And it’s not ideal.
So partly planning, to think about the best days for these cross-branch international meetings. Partly, if you have a lot of these, trying to make sure that globally, as an organization, you have harmonized in-person days and harmonized at-home days. And partly, in the long run, I think technology’s going to play an important role at making this easier. Three to five years.
One example, it’s not just hardware, it’s software. If you watch a football match on TV, there’s like ten, 15 cameras in a stadium, and they rotate between them to get a great view. And you really feel like you’re there and you’re in the action. For a lot of meetings, there’s one camera. And that one camera’s at a weird angle. Person X is speaking and you can see his or her ear or part of their neck or their leg, and that’s it. And it feels so disconnected.
So one of the things is getting more cameras. Another is just using AI to try and make sure the camera’s focusing on the person that’s speaking using direction or sound. There’s a bunch of things that aren’t there now but I think in five, seven years will be. I mean, it’s starting.
Some of these devices try and focus, but they’re far from perfect. But I think, in a few years from now, because the market’s so big, every company’s spending on this. So if you get it right, you’re in a jackpot. And so I think this is going to improve a lot.
I think of work from home as kind of like a Nike swoosh. We’re kind of at a low point now. It may drop a little bit if we get hit by a recession. But in the long run, I’m thinking three, certainly five years out, it’s going to pick up and it’s going to be higher than it is now. If I was planning for buildings, hiring strategy five, ten years out, you should actually be thinking of higher levels of remote work then than they are now, not lower.
And that affects everything from office space to who you hire, because different demographics have different preferences, your products, your strategies, your locations, etcetera.
Michael Chui: A lot to think about. As much work as you’ve done on this idea of work and the workplace, you also do other research. You coauthored a paper entitled “Are ideas getting harder to find?” Actually, one of our other guests mentioned it. What’s going on there? It sounds a bit scary.
Nicholas Bloom: You can either be positive or negative on this. Just to explain, we show from the ’60s or ’50s till now, the rate of productivity growth and the development, or what I call new ideas, has been slowly declining. So productivity growth is falling, and at the same time, the amount we’re spending on science, R&D professionals, money by firms, by the government has actually been rising.
It looks like it’s getting harder and harder to come up with new ideas, and Moore’s law is a good example. For many years, the speed of these chips was doubling between every 18 and 24 months. And you think, “Well, look, progress is steady.” But if you looked at it, the amount of money spent on improving those chips, the size of the research teams, was itself rapidly increasing. In the ’70s, it was eight-, ten-, 12-people teams were doing this, and there were thousands and thousands of scientists doing this.
It seems pretty clear, and the paper goes through a lot of different areas, from cancer treatment to heart disease to agriculture, etcetera, whereby you see more resources going in, but progress is either flat or slowing.
In some ways, you could take this as very negative. I don’t think it’s so negative in the bigger sweep of history. To give you that long sweep of history, productivity growth, we have better data for the UK. But if you go back, there’s data going back, like, over 2,000 years, so it was very, very low. It was below 0.1 percent per year. People in England in 1700 were not dramatically richer than people in 0 AD. And that was an enormous sweep of time.
You then see the Industrial Revolution starts. Growth starts to pick up, starts to accelerate. It peaks in about 1950. So 1950’s kind of the heyday. Universities get smart. They start to focus on research. All the companies have R&D labs. America particularly, but to some extent Europe, becomes more focused on being progressive and trying to do research and focusing on science.
We’re now falling off that high, but we’re still growing at 1 percent per year. It’s just not the 3 percent in the ’50s. So you could be negative and saying, “Ideas are harder to find.” Yes, that’s true. But we’re benchmarking ourselves against the best period in human history. We’re still making progress, it’s just at a slower rate than in the ’80s, and a much slower rate than the ’50s.
You can read this—if you’re an exec out there, or a company—I think the last ten years, 15 years are probably a good prediction for what’s going to happen for the next ten, 15 years. I don’t think there’s going to be some tech-induced explosion in growth. We keep being promised that; it just has yet to happen.
On the other hand, growth does not collapse to zero. Productivity growth’s about 1 percent a year, and I guess that was what we’ll maintain for the next ten, 15 years. Forecasting beyond that is, I would say, in British English, what we’d call a mug’s game. I wouldn’t feel confident forecasting more than ten, 15 years out. Even that comes with some uncertainty. But given there’s only been two changes over the last, what, 200 years in rate of progress, yeah, I think it’s reasonably safe to say we’re making steady, slow, reasonable productivity growth.
It’s not zero, but it’s certainly harder. It’s like a mine. There’s still plenty of coal down there, or gold, or whatever you’re mining. But you’re having to go further and further into the mine to get the stuff out. But you never know. A rock fall, a war may break out and there’s an enormous scene. It could be the singularity point happens, or some exciting thing. Being an empiricist, looking at history, it predicts we’ll have slow but steady growth for the next few years.
Michael Chui: Some people are more hopeful, but I understand where you’re coming from. All right, if you don’t mind, let’s do a quick lightning round of some quick questions, quick answers?
Nicholas Bloom: Sure. I’ll try my best to be fast.
Michael Chui: All right. What’s your favorite source of information about workers?
Nicholas Bloom: The media. I read the BBC, Economist, Wall Street Journal, New York Times, FT. I try and read pretty much all of them, but I think there’s five. I love that, pretty much that informs me every day.
Michael Chui: What’s the one thing about workers you’d most like to measure but haven’t figured out how to yet?
Nicholas Bloom: Expectations. There’s great data on what people are doing, what they did in the past, what they get, what they’re earning. It’s much harder to get people’s expectations, maybe even their prejudices. We’ve been trying to collect that in SWAA, in the survey asking people, “What do you think? What are your feelings on X? How would you change your behavior if this happened?” That’s much harder to get, and you have to come up with survey questions that elicit a good answer but are easy to understand.
Michael Chui: What’s the most surprising finding you’ve discovered about how work has evolved over the past few years?
Nicholas Bloom: Maybe that work from home works. I didn’t even believe fully my own research. I had a paper called “How working from home works out.” I had some papers going back to the ten years prepandemic on how positive it was.
It was very positive. I just thought, “It worked in this one incidence. It’s not going to work in aggregate.” As we’ve seen, post-2020, it’s turned out to be much better than pretty much anyone thought, including me. I mean, if I wasn’t positive on it, it’s hard to think who was positive enough. But very few people—a few insightful companies like Automattic or Upwork etcetera that were fully remote even prepandemic. But it’s hard to find folks beyond that.
Michael Chui: What’s the most surprising occupation that is amenable toward remote work?
Nicholas Bloom: I was told a story about street cleaning, and I don’t know whether this is an apocryphal story or not. Somebody told me, and I’m not sure this is—I’ll tell you the story, and you can see how it could work in the long run.
A company has developed these robots, and there are little cameras inside the robots. And they go around and they clean streets, but there’s an exception. They go back to some kind of HQ and they check out what’s going on. And so one former street cleaner can now be an office-based street cleaner, in charge of five, six robots.
So you think that their productivity’s way up. Probably each of those robots isn’t quite as good as a human, but certainly they’re better than a fifth of a human, and there’s five of them.
So if you can do that—I had a call earlier today, actually, with Steve Davis, who’s talking about deep-sea oil exploration. There used to be these folks that would go down and had to be there and looking at the camera feeds and the drilling stuff and actually measuring it. Soon as the pandemic happened, they couldn’t leave, and they’re actually doing it from their living rooms. And, again, it wasn’t as good as in person. But look, if you’re not an exploration person, it takes you a week to travel somewhere and you’re only 20 percent more productive, maybe it makes sense to do this stuff remotely.
Michael Chui: What does your perfect office setup look like, physically?
Nicholas Bloom: Comfortable. Probably not too big. You know, a meeting room that you can meet one or two students or visitors. I spend a lot of time talking to students, advising them.
The other thing is you’ve got to have a—we’re now all hybrid even in our office, in the sense that I have a lot of Zoom calls, Teams, whatever—and so you want a nice camera, and a good microphone, and a background.
Actually, my office is pretty much a mess at Stanford. I’m taping this from home, but my home is much tidier. I could only find one wall that looked OK. It has a picture of Glasgow in it, where my wife is from and I got married. So often calls start off with “why is there a picture of Glasgow behind you?” And then I explain.
But that works well. You just need one of your walls to look reasonable. And of course it has to be facing the light. It can’t be the window because you’re, like, this silhouette from one of those good cop–bad cop investigations.
Michael Chui: Where in the world would you most like to do remote work from?
Nicholas Bloom: Maybe Greece. My father-in-law lives there. My father- and my stepmother-in-law live there, and it’s fantastic. They live on the coast in a quiet part of Greece that’s really lovely. We go there every summer. It’s very nice, very civilized, great food, lovely climate.
It’s not that dissimilar to California. To be honest, I’m aware that I’m very lucky living in California. It’s a lovely place to be because you’ve got some outside space.
Actually, one of my students, she just sent me something today on what’s called digital nomads, people who have gone fully remote. She’s interviewing a lot of them in Mexico, and there’s fantastic quotes from people saying stuff like, “In the pandemic, I decided I wanted to learn to surf.” They’re coders. They basically moved to Mexican surf towns and just did everything remote for a year and learned to surf or run a DJ business. “I know I’m never going to make enough money to live off it, so I’m still coding in the day on Java.” So, you know, there are other places, like Thailand, Mexico, etcetera.
Michael Chui: If you weren’t doing what you’re doing today professionally, what would you be doing?
Nicholas Bloom: I enjoyed my time at McKinsey. I also worked at the UK Treasury. It was good. I worked at the IFS. I’ve had some related but not the same careers before, and I enjoyed them. Maybe one of them. I had, as I mentioned at the beginning, a very random reason I ended up in academia. They also are very appealing careers, and I enjoyed them a lot. The UK Treasury, you can imagine, right now would be pretty scary but also a fascinating place to be.
Michael Chui: And if you had one piece of advice for listeners of this podcast, what would it be?
Nicholas Bloom: I guess everyone that’s listened this far is focused on remote work. And my piece of advice would be to make this work. There’s a lot of scare stories in the media saying hybrid doesn’t work. Mainly it’s because either they get confused and are talking about fully remote, or they’re talking about disorganized hybrid.
My one piece of advice is just be intentional, be organized. Basically, get folks in on two, three days a week of what I call anchor days. And be really tough about having them come in. And then let them work from home on the other days.
And then on their home days, give them what I call performance management. So say, Michael, if you were my boss, “Nick, I want you to get this stuff done.” So you’re measuring my output rather than micromanaging what I do minute by minute.
Because in another study, we see that people on work-from-home days, they work about an hour or two less on those days. And when you ask them what’s going on, they say, “Look, I’m going to the gym. Picking my kids up. Going to the dentist.” But they make up for it on other days.
So interestingly enough, hybrid work-from-home workers work about the same hours, if not slightly more, per week. But on the Wednesdays and Fridays they’re at home, or Mondays and Fridays, these are great days to go to the dentist, have a round of golf. So, be smart about hybrid work from home. I think it can work really well, and lots of data shows it can, but it has to be well managed. It doesn’t just naturally, randomly happen that way.
Michael Chui: Nicholas Bloom, thanks for joining us.
Nicholas Bloom: Michael, thanks so much for having me on. It was a lot of fun.