As the world gradually begins to reopen for in-person experiences after more than a year without, businesses are evaluating which aspects of the temporary switch to remote work should remain permanent.

From increased productivity and freedom from commutes, to feelings of isolation and a lack of cohesion among teams, working remotely elicits mixed reviews. That variation in worker preferences is exactly what employers are now tasked with incorporating into workplace policies.

Choice, above all, is what workers want, says Stanford University Professor of Economics Nicholas A. Bloom. While businesses scrambled to adjust to new work-from-home practices and technologies last year, it was familiar territory for Bloom, who has spent decades researching remote work and management practices. He has surveyed thousands of firms and individuals at corporations all over the world.

Discussions around how the workplace should function are rapidly shifting as businesses begin to think about the future. “If you look at the survey data, you see very clearly there’s a huge variety of preferences over how many days people want to work from home,” says Bloom. “For example, around 27% of people basically say they never want to work from home ever again, post-pandemic.”

Bloom joined Greylock general partner Sarah Guo on Greymatter’s Work From Anywhere podcast series to talk about the evolving workplace, what the data shows on both employee and management wellbeing, the allure of the hybrid-work model, and more.

As Bloom notes in the conversation, around two-thirds of firms right now are opting for a hybrid workplace, with employees working a few days per week from the office and the rest from home. This has its perks, as Bloom and Guo discuss, but brings its own challenges around diversity, management styles and promotion opportunities, along with the economic impact from changing office real estate needs.

“I think more and more, leaders are recognizing that the connectedness and creativity and growth problems are actually the ones that they’re going to have to deal with, not the productivity problems,” says Bloom, who also serves as Co-Director of the Productivity, Innovation and Entrepreneurship program at the National Bureau of Economic Research.

You can listen to the full podcast here, or by hitting the Play button on the embedded SoundCloud player below.

Episode Transcript

Sarah Guo:
Nick, super excited to have you on the podcast. Would you mind introducing, for our listeners, yourself and a little bit about what your research areas are?

Nicholas Bloom:
Sure, I’m Nick Bloom. I’m Professor of Economics at Stanford University in California. You can probably hear I’m British, hopefully. I’ve been in the U.S. 15 years. I’ve long worked on management practices in parts by working at McKinsey a long time ago.

For a long time, I’ve actually been interested in this weird niche called working from home, which was like a complete boredom to anyone until suddenly, you can imagine March 2020, it became the hottest topic around. I’m now spending most of my time working on that, talking to execs, collecting data, etc.

Yeah, it’s been incredibly interesting as we keep an ear to the ground with our portfolio and the leaders in the startups in our ecosystem. But what are you hearing as people make these decisions about how to come back into the office or not? What are you predicting firms will do?

Well, I’ll just give you background. I’ve been serving thousands of firms and individuals, also talking to a lot of… I’ve been doing a lot of consulting. Most firms have decided — I would say something like two-thirds for now — have opted for hybrid.

Within tech, that’s an even higher share, so tech has gone heavily for hybrid. Here the setup is you say, We’re going to come in to work, let’s say, Monday, Tuesday, Thursday. Work from home Wednesday, Friday.

For example, Google just announced exactly this, the 3-2 plan. It’s three days a week in the office for most people, two at home. That’s by far the most common. [In other industries], more in investment banking, for example, people have said, We hate hybrid, you’ve got to come back in. There’s good reasons for some of these firms to say that.

For example, if you are trading, you need a high-speed connection. You’re dealing in confidential data. There are other firms that have said, We love remote. [Quora] was basically going to stay remote almost permanently. But I would say the vast majority are saying hybrid.

Then there’s a whole range of debate about how you make that work. Most firms are all remote now, but they are going to get people back into the office for the majority of days, but definitely not all of them.

OK. That’s also similar to what we’re hearing; that most workplaces, especially in technology — we should talk about the industry variation as well — are going to be hybrid. What do you think is going to be hard about that? What are leaders worried about, or what are the issues people are asking you about?

I should first say why hybrid. So hybrid, I’ve been advocating this since literally the lockdown. I wrote a blog piece in May 2020. Tech has been really advanced on this, for example Mark Zuckerberg made a statement last May as well.

To me, hybrid seems pretty obviously the way forward and the reason is it’s a nice trade-off between two challenges. You get a big advantage in-person on A) innovation, and B) cultures. Just about every manager I speak to says you really want face-to-face meetings to be creative and innovative, and it builds bonds and builds culture, so you want a few days a week on that.

But there’s also a big upside to working from home which is: A) you save the commute, because a lot of the time the average American spends an hour a day commuting. And B) for work that’s actually individual, or maybe one-on-one like we’re talking now, it’s fine. It’s maybe even better remote. Certainly, if I’m reading a report or working, it’s quieter at home than in the office.

Hybrid is kind of the best of both worlds if you organize it properly. What that means is within a team, around people you work with, you make sure you’re all in on the same day. The team manager says, “Everyone’s going to come in Monday, Tuesday, Thursday.”

That way all the team lunches, training events, leaving do’s, client things, big meetings — [you get through them in] those three days. The other two days, Wednesday, Friday, then you think, Those are my days for my quiet reading, etc.

At that level, that works pretty well. There’s a big debate around choice, but that’s basically why hybrid is the overwhelming majority, because it trades off two tensions around, “We need some face-to-face, we need some home time.” It’s kind of the sweet spot. Three days a week may not be right for all firms. It may be four, it may be two, but it’s somewhere in the middle. It’s roughly 50/50.

I’ve done a few surveys. I’ve been doing exec ed at Stanford on this and talking to a lot of firms. I’ll tell you the choice issue. I’ve even changed my mind since the beginning of the pandemic.

Here’s the choice issue: If you look in the survey data, you see very clearly there’s a huge variety of preferences over how many days people want to work from home. For example, I think it’s 27% of people basically say they never want to work from home ever again, post-pandemic. They’re fed up with it, it’s boring, it’s isolating, it’s terrible. They tend to be younger, single people or empty nesters.

Then there’s a bit over 30% that actually say the exact reverse. They say, “We love working from home, we want to carry on working from home five days a week.” They tend to be kind of late middle-aged or middle-aged and married with kids. I guess like me, and living in a house.

Then the majority, which I should say would include myself, want to come in something like one, two, three, four days a week. You think on the one hand, We should let people choose. This is America. We like freedom of views and it’s going to make the employees happy. I’ve heard many execs, particularly in tech, actually say that.

There are two problems, one obvious, one less obvious.

The obvious one is what I’ll call the curse of mixed mode. The mixed mode is when there’s five of us on a call, three in the office, two at home. You have the three in the office who are all in some little conference room, two at home on separate screens.

It’s not an equal thing, you can see three tiny heads. When you talk to managers they say, “Ah-ha! We’ve got a solution to that.” We’re going to do what I call the American and Germany example, which is, as an American, when you go to Germany, if you’re out for dinner with four Germans they all speak English. They just humor you and speak English.

It’s similar that if there’s just one person at home, everyone in the office has to connect in on their laptop so that you’re all on equal footing. Fine, but there’s a problem with that as you can see — which is, of course, once the meeting’s over, everyone in the office stands up, grabs a coffee and the meeting effectively continues. So mixed mode is still somewhat problematic.

The much worse thing is what I call diversity. Which is, if you look in the survey data, what we see for example amongst college-educated respondents to our surveys, which is about 60% of the labor force.

If you look at people with kids 12 or under, almost a 50% higher share of [women report that they] want to work from home five days a week [more often] than men. We also see people living further from the office, disabled people. There are certain groups that clearly have a stronger preference for working from home more days than others. So that’s fact one, and that kind of is what it is. Then there is another fact.

From years of research, we know that if you’re in a team whereby some people are coming in every day and others are working from home, those working from home face a big promotion penalty.

For example, I did a large randomized control trial out in China several years ago where we randomized volunteers into working from home. Those that were randomized into working from home for four days a week, their promotion rates were half those in the office after 21 months.

That’s an enormous difference. Now, otherwise identical, they’ve been literally randomized. If you put that together you can see there’s a huge risk with the choice plan, that 5 or 10 years from now you have all the single young men coming in five days a week, rocketing up the company.

Let’s say people who are married with young kids, particularly women, choose not to and then just don’t get promoted. I think that’s pretty difficult to deal with. I’ll pause, but that is a hugely contentious issue over where firms are deciding on choice individually, choice at the team level or no choice at all.

Yeah. We’re thinking a lot about this. The sort of second-class citizen problem is how we’re referring to it internally at Greylock.

I think one thing that you and I were discussing was, is it too ambitious to think that we can actually solve some of these problems if we to some degree construct the choices to put people on an equal playing field? And then improve the tools and processes for hybrid teams.

Yeah. In China years ago, I had a graduate student called James Liang, who is now the CEO of We did a big randomized control trial on his firm and we randomized hundreds of people. It was a scientific experiment. You only get that if the CEO is a PhD student.

It was an amazing setup. We did lots of interviews, it was really great and James was a fantastic guy. I know probably a bunch of venture firms have invested in James’ firm and I think done very well over the years. We actually interviewed and tried to find out, Why do people that work from home not get promoted as fast? I’ve talked to a lot of friends and colleagues and firms.

I think there are two reasons. One is fixable, one is harder to deal with. On the fixable one, you can use data to try and make sure that people working from home aren’t just overlooked. One of the two reasons we found why they didn’t get promoted as fast is they’re just forgotten about and so you can use data to try and think more carefully about who’s in your promotion pool and consider them, whether or not they’re physically in the office that day with you.

The other thing was much harder, which is, it looked like from Ctrip, and I’ve heard this anecdotally as well, that you do develop managerial skills [from] being there in person. It’s kind of like the flip side of why people are more productive at home because they’re not having all those lunches and coffee chats and what seems, in the short run, maybe to be gossip but in the long run is actually really important for culture building in the firm.

If I’m in with you everyday, I have a good sense of what your outside situation is and your thoughts and your abilities etc., and I probably will be a better manager for you.

That’s hard to address because fundamentally if you’re working from home, you may be great at your current job but you may be even better, actually, at your current job. You may perform better than people coming in because you’re more focused, but you may lose out on some implicit kind of development that would work well as a manager.

Yeah. One of the companies I work with is called Remotion which is explicitly trying to enable people to focus on this cultural health and sort of social fabric of organization’s problems. It’s interesting because they’ve also been running and looking at survey data and it’s just been a fascinating year in terms of what leaders are worried about.

Cycling very quickly because now I think more and more, leaders are recognizing the connectedness and creativity and growth problems are actually the ones that they’re going to have to deal with, not the productivity problems. We’ll see if the investments work but I think a lot of people are taking the risk now.

Yeah. When I’ve been advising firms, I’ve generally said this is an enormous revolution so it’s easier to go wrong. I would be slightly risk- averse and kind of go slowly and learn.

It’s the exact reverse of “go fast and break things.” Go slowly and learn from other people’s mistakes. I would probably do it centrally, and the reason to do it centrally is at the center you say, “Corporate team, you guys are Monday, Tuesday, Thursday and [other team] you guys are Monday, Wednesday, Friday. Residential, you guys will do X, etc.”

Because then you can make a better use of office space. You can also think which teams should overlap because I think corporate, industrial, I want them in on the same day so may give them exactly the same three days.

Then within the team, everyone sticks to those three days. That sounds like a very centralized plan. But it means you can A) design your overlap and use office space effectively, and B) you don’t have the diversity crisis because the opposite is like it’s a complete free-for-all.
Everyone can choose. Then you can’t really shrink the office and then there are all kinds of issues that come up with it.

I’m not saying that’s where we end up, but I would honestly centralize in the short run and then relax out from there because the reverse is really hard. You’re opening Pandora’s box if you say, “People can choose. If you want to work five days a week and you want to move to Hawaii or Alaska, go ahead.” Because if you discover that’s a problem, it’s very hard to unwind from.

Yeah. One of the things I think is so interesting right now is in my professional career, I think in everyone’s, there’s never been a time where there are such divergent attitudes about the level of risk people are willing to take on workforce strategy to take advantage of the opportunity. The window of what was normal or what was accepted in terms of workforce strategy was rather narrow, right?


You were often your own little corner with remote work and everybody was like, Oh, that niche thing.

I will say, we have very large companies in the portfolio. The entire office is like, It’s a WeWork. Hotel in, hotel out. We’re basically remote through our tools and people can choose to do what they wish and so I’d say there is a real range.

I think people are taking some of the risks you are advising against, but I think those people are really going to need to lean into practices that are… If you’re going to get very decentralized that way, you’re going to need to be technology-first, right?

Totally agree.

And we’ll see how well that works.

It may work out. I mean, look, it’s like taking a gamble.

The other thing to point out that’s changed in the last, for me in the last six months, and I mean the market’s even aware of it so then the New York Stock Exchange and NASDAQ, is it’s very obvious now that the labor market’s become extremely tight.

You’ve seen as of today, actually, the data coming out is showing there’s the highest number of open job vacancies there’s ever been. This points towards firms having to offer some working from home day perks.

In our survey, it turned out people value the ability to work from home two days a week at about 8% of earnings. If you’re trying to fish employees out of a tight labor market and you do not offer working from home two days a week, you decide everyone has to come in five days, you’re going to have to pay at least 8% more to get… It has just become prohibitively expensive so I suspect tight labor markets are just going to force working from home a couple of days a week on the firms.

Otherwise, they’re going to have mass quitting and they’re just going to find it impossible to fill positions.

Yeah. One of the interesting things we’ve seen is, as you said, I think some leaders recognize choice is going to be challenging to deal with, but they also look at their workforces of highly skilled people who are… They’ve been given that choice by the pandemic and they say, “What do we do? Look, we need to keep these people.”

To some degree, a piece of it is being forced upon us.

Going back to one thing I’d love your point of view on, or any learnings you have from the data, is how you think this will impact different levels of the organization.

I mean, a couple of points have come up. Again, going all the way back to the beginning, I remember Mark Zuckerberg talking about this in May 2020. It’s the new hires. This is a slightly different world, so I would advise new hires coming in more days.

Maybe not full-time, imagine you’re on the 3-2 plan. You may say for new hires, This can be one day extra a week when there’s just you guys and we’re going to do training events where you’re going to work, just basically socializing and getting you up to speed. That’s one group.

Then maybe there are people that are very technically expertised, probably aren’t aiming for further promotion within management, could probably easily spend three days a week at home if they’re in some technical job.

The other thing that’s come up a lot in terms of firm organization is what’s it like to generate a huge increase in outsourcing and offshoring. The number of firms I’ve talked to have said, You know what? It’s been kind of amazing having these teams working from home for five days a week for a year. It’s been really great. I wouldn’t say it’s perfect, but they work really surprisingly well. I’ve been thinking, if they work really surprisingly well remote, why don’t we move them to insert country name, India, Mexico, whatever. Outsource them.

In some way, it’s like China in the 2000s … I think we’re likely to see a huge uptick now in services offshoring, in part also because of the restrictions on travel and immigration.

Another sector that I assume, from what I’m hearing anecdotally, is kind of exploding is service sector firms that are offering basically outsourcing services either within the U.S. or abroad.

Yeah. It’s interesting. One framework we have for sort of the change in workforce strategy internally is the future of work is more distributed. You can describe it as hybrid or remote, but more distributed than it used to be in different structures.

That’s very interrelated to it being more global because if you’re going to be somewhere else, then you might as well get the cost advantage which I think is something that people do not like to talk about in these sort of discussions of the soft and fuzzies of workforce strategy.

But there are extraordinary developers and support people and salespeople and many functions elsewhere and I think it’s also flexible, right?

Yeah. Absolutely.

Because people say, If we have to intermediate the work we’re doing by scoping it more clearly or communicating through digital tools, does it need to be a full-time employee, actually, to do this thing?

I think the set of changes is actually quite related and so when we talk to CEOs about how this is happening, they’re like, Yeah, well, we’re going to do contractors. We’re going to do global and we’re going to let people work 3-2. Right? It’s kind of all together.

Exactly right. Yeah, precisely. Offshoring, outsourcing, both of them are just exploding.

The other issue that has come up is it’s a lot easier to be multinational because suddenly now we’re all much more familiar and at ease with Zoom and video calling. Imagine the firm had a policy of being at home mostly on Wednesday or Friday. That’s the perfect day to have multinational conversations with the Europe and the U.S. office because it’s very straightforward to do it, because you’re all on a level playing field.

I’ve heard there are many multinational firms [that] have said, It’s harder to communicate within my office in the U.S. I find it easier to communicate with Portugal now, oddly enough, than I do with my teammates in the office.

That’s funny. Nick, you’re a Brit, you’re running studies in China, you’re talking to American and global companies. What geographic variation do you think we will see around these shifts?

I’ve seen some data. The best of us see it as partly anecdotal and partly as a paper in the World Bank. There’s two main things driving this.

One is just the level of development. It seems kind of obvious, but more developed countries have better infrastructure on broadband. People are richer, they’re more likely to have a house so you can work and have space and more likely to have a computer at home.

They’re also more likely to do the types of jobs that enable working from home. Once you think about it, it becomes pretty apparent, but the U.S., for example, has a lot of graduate jobs that work on computers that can be done at home.

If you go to a company that has a big subsidiary in Indonesia, there’s a lot of this more physical, manual stuff. You basically can’t do it at home.

One huge angle is development, and as you go from Northern Europe and the U.S. down to Southern Europe into the developing world, working from home shares go down because it’s just less applicable and harder.

The other second angle — you hear it more anecdotally, it’s slightly tough to get data on [but] I’m trying to collect it — is culturally, some Asian countries, particularly Japan, you hear stories about there’s a huge importance of face time.

If that’s true, I believe that… Have you seen it from your portfolio companies at all, that in Asia the uptake’s been lower? Now, one reason, of course, they live in small apartments. The Americans tend to have a lot of space and big houses but I hear, anecdotally, cultural factors is another reason why it hasn’t taken off as much given what you’d expect for the level of development.

Yeah. I do think one thing that is a little bit sort of understated right now is the degree of cultural change for leaders who have operated in a very different way over a long period of time.

Because even in our portfolio, people who think of themselves as extreme innovators, and we do too, there are a lot of people who said, Well, I’m an office culture guy. I’m an office culture gal. I like to see butts in seats. There’s a piece of it, as we were talking about, that is… The data would show that culture, creativity, progression; there are things that we need to enable more differently if it’s going to be remote that are not automatically solved by people using Zoom and Slack and such.

But some of it, I think, is going to be just like attitudes in different companies that are going to come from leadership. Some of them may also be geographic attitude.

Yeah. The other thing, I had a fantastic and fascinating conversation with Marissa Mayer about a year ago now. I kind of reached out to her to ask her about what happened in Yahoo! in 2013 and what we could learn from it.

It was really interesting and it’s a great parable for the importance of management. Marissa said when she took over at Yahoo!, they’d been through tremendous upheaval and chud and they didn’t appear to have the best performance management system in place, so she starts putting this stuff in place, like good performance tracking of individuals.

Halfway through this process, she discovers there’s group of people working from home, and some of these people full-time working from home, as in they’re permanently at home, but apparently not logged in to their computers.

There’s not just low-level slacking, it’s this complete permanent goofing off. She basically canceled it and then at some point relaxed it back a bit. But the parable of this from talking is there’s really two ways you can manage people. There’s what’s called input-based management I sit in the office and I physically watch you and I check, Are you typing away furiously? You appear to be productive. Are you kind of asleep or talking to friends?

Or there’s output management whereby I say, “Sarah, I don’t care what you do and where you are. You get your job done, that’s great and if not, you’re in trouble.”

Now, input-based management’s horrible at home because you can’t watch people and for firms that relied on it… You can think of it, it’s called basically bad management but if you’re input/bad management, working from home is a real challenge.

Whereas if you have proper performance-based output management, it’s not that difficult. You were never breathing down people’s necks in the first place. From talking to companies, it seems that management, particularly performance management, is very complementary for making this a success.

A lot of firms are kind of beefing up data collection and HR because they’re doing the reverse. They’re saying, OK, it wasn’t great before the pandemic. It definitely needs to be good now. Let’s beef that up.

There’s different ways to do data collection that I think will be effective and not. I’ll give you an example of not. I’m sure you’ve heard of this, too.

Especially early on in the pandemic, we talked to a number of leaders that basically wanted to view whether or not people’s computers were open all the time or if people were typing. Perhaps because I have spent too much time in engineering culture, but it’s kind of like engineering metrics, you shouldn’t measure lines of code, right?

Because anything that is even a bad output metric or an input metric that can be gamed by engineers, it will be, right?

No, totally. I do see that surveillance software as remote input management. I agree, it doesn’t work at all so it’s like, I can’t see that you’re working so I’m going to try and do it remotely.

But it was a bad idea from the outset and it seems creepy now to do it remote. I totally agree and I think it’s kind of died. I interviewed a surveillance software company back in July last year and they said, “Business has never been better.” It was pretty worrying, but I think it’s dropped back off.

I’m sure Marissa’s question back at Yahoo! during that period of time is like, How did we not notice that these people had no output? Or that should have been the very first question, right?

She basically put in an output performance evaluation system and once it’s up and running, then you can let people work from home, say, three days a week or two days a week.

I, personally, even though I was the working-from-home person before the pandemic, I was never advocating five days a week. Most people don’t even want that. Again, coming back to the earlier discussion. Most people want to see their colleagues. It’s depressing and isolating.

But I think if you’re going to work from home two days a week, you just need to be assessed on output. Then at that point, it works well.

What other advice would you have for just managers about how to adjust to this?

I mean, another thought is on office space. That comes up a lot. There’s the very obvious things, obviously sign short-run leases, the things are extremely uncertain. I have data on lease length and that’s fallen, on average, from nine to seven years so everyone’s doing it. I mean, it’s pretty obvious. Don’t sign long leases now or if you do, get a big discount.

The second thing is, if you think about the way offices are going to work with hybrid work, the office should now be a place that’s really going to have two activities. Most of the time, it’s collaborative work so [you] want a lot of open spaces, meeting rooms and a bit of the time you may be off in a little cubicle basically taking a Zoom call, in which case you want tiny cubicles.

What you don’t want is the Mad Men-style era of executive offices because people aren’t really going to be sitting on their own in the office doing individual work because that’s going to be at their home base.

In terms of office design, it actually goes completely contrary to what was there during the pandemic to stop infection spread. Actually, you want open plan spaces and collaborative spaces, and you don’t want Perspex screens and people in cubicles.

Then the other comment is, in terms of offices, the net demand for office space in the U.S. is slightly down. Not massively. You’d think it would be hugely down because you think, Well, look. Everyone’s going towards working from home. We know an aggregate. We’re going to go from 5 to 20% of days of working from… Full days worked at home. That’s a 15% drop in footfall into offices.

But you have to remember, that’s been offset by an increase in demand for space, as in reduced density because of fears of contagion. So net, it’s about equal. What most firms I’m talking to, they’re taking having working-from-home days, they’re using that to de-densify the office.

It’s not just elevators. It’s reception areas and kitchens, etc. Turns out, the one piece of office space that’s kind of cursed is high rises.

You think of skyscrapers, those things, they’re still basically empty because, say, Salesforce Tower or some huge building in New York. A) To get to the front door, you got to pay for the subway. You can’t do that in the morning and evening rush hour without being packed in like sardines, [like you are in] the Tube in London. Then B) by the time you get to the front door, how do you get up to floor 40 without taking a packed elevator?

Elevators tend to be jammed on the way up in the morning and down in the evening, so those buildings are in real trouble. If I owned space in that and I was going to renew a lease, I’d be looking for a massive discount. I mean, I don’t know, 30, 40, 50%.

Because in reverse, in fact, what tech firms often have, these kinds of campuses that you can drive or bike to and you can walk upstairs if you want are actually quite popular. That’s the other twist. It’s not that the demand for the space has cratered, it’s that it’s cratered for high-rise buildings and city centers.

Yeah. One thing we’ve been talking about at Greylock for the last year is also if we make these predictions now, what are the second order effects and the third order effects and can we invest them?

We were talking about the donut effect and the impact on residential real estate as well. Can you explain that?

Yeah. It’s very, very clear in the data. I’m working, in fact, with an undergrad here, quite amazingly, Arjun Ramani. We’ve been pulling two different sources of data.

One is Zillow, so ZIP code prices by month, and the other is we do the Freedom of Information Act from the U.S. Postal Service. We got their change-of-address data by the businesses and households by month, by ZIP code. You see really clearly in both data sets the donut effect.

Just to explain it, what’s happening is a central business district, a real core downtown, large cities — think of New York, Chicago, L.A., D.C., San Francisco — are seeing massive outflows of households and of businesses and price drops. Rental, very obviously, but even resale prices.

The suburbs are seeing huge increases. Then places far away like, I don’t know, Topeka, Kansas, or Tulsa, Oklahoma, are going up a bit. What’s happening is the story that kind of ties it all together is people are not leaving U.S. cities. People are not really permanently fleeing New York and San Francisco.

What they are doing is moving out to the suburbs. It makes total sense because if you’re going to work from home two days a week, you can base your commute a bit further and you also need space for a home office and so the whole world, and business included, is moving out.

Now, the economic impact is rental rates are down in city centers and also things like retail traffic. We estimate New York will probably see, and San Francisco, a drop of about 10% of retail sales in city centers.

It’s not cataclysmic, it’s not like we are going to see back to the 1980s whereby centers are dangerous, ruled by a crack gang. Maybe you’re going back 10, 15 years. I mean, city centers have been on an upswing since 1980 and maybe their relative price is going to go back to where they were in 2005. They’re still expensive, they’re just less so.

I think the market has already [shown] this but if you could go back in time, it’s pretty obvious you’d want to short city centers and go long on the suburbs.

Already, the property market’s doing that. I suspect it may continue a little bit, actually.

What’s your prediction for, or maybe you’re seeing it in the data, for what we’d think of as secondary cities and if there’s a lot of movement there?

They’re up a bit, not much. I mean, quite surprising. We broke into the top 12 cities, which kind of goes from New York to San Francisco — that’s one and 12. Then the next up to out of 50, so a place like Kansas City is like number 30. Then the remaining, whatever, 300 MSAs and outside them, they’re kind of small towns.

What you see is this donut effect is very heavily driven by the top 12 cities. There’s a bit in [the] middle and really not much in the smaller ones. This is what’s kind of interesting in terms of movement of individual people and businesses, they are slightly moving out from big cities. Not massively, a bit. You see a slight exodus but in terms of prices, the reverse is going on.

You may think that’s surprising, but I think what’s happening in the data is the industries that are heavy in big cities like tech and finance have done really well. The people that work in tech and finance have become dramatically wealthier over the pandemic and can afford to buy much nicer houses and apartments and that’s why they’re driving up prices in San Francisco and New York.

But they tend to have lower density, so there’s probably a lot of lower-income people moving out, and middle-income into Tulsa, Oklahoma, etc. If you want to have the hottest property in the entirety of the U.S, it will be the suburbs of New York and San Francisco. Those things have gone up like 50, 100%. It’s now become effectively a suburb of San Francisco, that 100%.

The worst property markets, oddly, are the central business districts of the very same city. The reverse was happening for the last 40 years, that people were moving into the centers and have now moved out at speed.

Since it’s about purchase prices and businesses are probably moving, this thing is not going to entirely snap back. Some of it will. They’re not going to permanently stay like this. Some of it will snap back, but certainly 10 years from now city centers will be less expensive and lower density than they would’ve been without the pandemic.

So what you are not seeing yet, which was a prediction by many earlier in the pandemic, is that the sort of second-tier of cities massively rising. You know, companies opening offices in Kansas City and such.

Not really. Probably like you, I talked to a lot of journalists and they love anecdotal stories about digital nomads, but just literally in the data, given this is tech, in the data from the U.S. Postal Service change of address and Zillow pricing data, we do not see that.

In fact, prices are going up fastest overall in the large, big cities and there’s some very shallow movement out of them, but honestly not a lot. I think the reason is until big companies leave large cities, nothing’s going to leave.

If you already know the Goliath National Bank, to take that fictional company from “How I Met Your Mother,” or let’s think of a company like Google or something. You’re probably not very tempted to open an office in Tulsa. You’re probably sticking with Silicon Valley and the Bay Area and New York and a few other large metros because that’s where a lot of the grads want to be.

They’re just now tending to live a bit more out in the suburbs and a bit less in the center. I think you’ll see what L.A. is like, maybe the development, the kind of multipolar city. You’ll see cities where there’s a kind of cool area where the young, single 20-somethings live and an area where 30s go with really nice restaurants and maybe 40s and 50s, I don’t know, art galleries.

You can imagine that you don’t all need to cram into the center, you can spread out. L.A. has a bit of that, the different parts of the city, and that may be what the suburbs turn into because if a lot of people move out with money, the suburbs become more fun. There’ll be more restaurants. Think of weekday dining. It’d be easier to go to downtown Palo Alto on a Wednesday because there’s a lot of people working at home and so there’d be good restaurants.

Interesting. I do think that perhaps there is a virtuous cycle or a secular virtuous cycle that hasn’t shown up in the data yet, considering the compression of how quickly this has happened in that as companies, as we were discussing, enforce or move to a workforce strategy that is aggressively more distributed. And then invest in the tooling that enables that, then they may become more open to new locations, as you said, that are fully remote or just secondary offices.

Possibly. I mean, you could argue that look, if I’m Google, say. I do care that on the three days a week you’re in, you actually are networking with other people. If I open an office in Tulsa, with 15 people in it and they just network with each other, that may end up being completely insular and really unproductive.

In some ways, hybrid puts more emphasis on, “You really do need to be social and connected on those three days a week you’re in the office.” They will be exhaustingly social.

By the end of each one of those three days, you’ll come home and want to put a towel on your head, [like] I don’t want to talk to anyone for half an hour. I’ve spent all day in meetings. Meeting after meeting. That’s the trade-off in return for having the two days of quiet, serene thinking on your working-from-home days. But that probably won’t work if you’re disconnected.

Another thing I’ve heard, by the way, is people are talking about, You’re going to work remotely for four weeks and you’re going to be allowed to live in Alaska, then come in a week where you do… The problem with that plan is most of these meetings, you need big groups. You can’t wait for a month.

[If you’ve] got a new product coming out, or there’s a disaster that needs to be fixed, we all need to get together, huddle as a team. You can’t say, Let’s do it next month. I think hybrid has to be more or less every other day in the office, every other day at home and then it’s very easy to shuffle tasks around. Most stuff isn’t so urgent it can’t wait a day.

Right. For that future I painted, that actually involves more economic opportunity or more jobs in these secondary cities, companies are really going to have to figure out the networking and the connectivity across offices and with the rest of the organization, and that’s a tall order.

Yeah. I mean, I had the same thought as you. I thought early on this would happen. We just are not seeing it in the data and we have data up to March 2021, so it could still happen. I don’t want to rule it out, but so far we’re a year into the pandemic.

A final thought that’s kind of interesting, another long run trend is the technology around working from home is dramatically improving. I have data from the U.S. Patent Trademark Office that I’ve been looking at with two other co-authors, scraping the number of times the words “working from home” comes up or “remote work” or “video calls.”

In fact, you see about 1% before the pandemic and then it suddenly just starts to explode from March 2020 onwards. There’s clearly a massive charge of people into innovating around working from home technology.

Which, I mean, what we’re experiencing now will clearly get better. If you look back 20 years, things like Skype in 2002 was a huge step forward so you had video calls over the internet, and then Dropbox and cloud around 2010 so you could file-share.

Those kind of massive innovations, I think we’re going to see more of them. Who knows what it will be. If I knew I’d be investing, as I guess all of us [would].

But five, 10 years from now, almost certainly some fantastic new working from home tech is going to come out, there’s companies I know [where that is] their forte. I mean, the market for working from home tech has gone up 4-5x so of course the R&D dollar is pouring into that, and the startup funding has exploded.

That’s exciting, but it also means this thing is only really likely to go in one direction. I think we’re going to get better at it and so that’s another force pushing us towards continued working from home.

You know Nick, we don’t know and we are investing in it. But it’s certainly something that we believe in just because I think people want the choice and the potential economic benefits and social benefits are very high, but it will not be trivial.

You’ve obviously been thinking about this area for a long time, but you also have such a good nose for data as to how we can actually understand what’s happening and be factful in our understanding of it. What research of yours should we look out for, or what else are you interested in understanding in the coming months?

Oh. You kind of miss the data nerds. I know there’s an entire field called data science and it’s kind of econ and data science are more or less one and the same. I’m just continuing to collect data on how things unwind, actually.

Not long after the beginning of the pandemic with Jose Barrero and Steve Davis, we started up a survey of initially 2,500, now 5,000 Americans per month, just to collect data. And we have another survey of firms with Atlanta Fed. They’re just really interested in seeing how you can see people charging home, you can see them slowly unwinding.

A lot of the stuff from that, [which] is fascinating, is views on people’s… say, on social distancing. For that one question when I give talks, I often use the polling feature in Zoom. I think it’s almost a new innovation. I ask people the same question we ask people in massive surveys. We always get the same answer to [the question], Post-COVID, after you’ve had the vaccine, what’s your behavior going to be?

There’s a kind of scale where the highest score is, “Completely return to pre-activities,” and the lowest is, “Continue social distancing.” You can reverse it. It doesn’t matter what’s highest or lowest, but there’s kind of two ends of the spectrum. What you find is over three quarters of people say they won’t fully return and they say they’re going to avoid subways and crowded elevators because they’ve seen too many sneeze videos.

I mean, honestly, they’re too terrified and so that kind of stuff is striking. Now, that’s true. That has huge implications for New York, London. I used to commute every day working back in London on the Tube, and according to that data most people will be too nervous to get on a crowded Tube train.

And if you’ve traveled on the Tube in rush hour, you know your face is like, pushed against the glass half the time, it’s so packed and it’s clear that one person sneezing, everyone catches it. Whatever, and you used to accept it but there are some pretty interesting things to watch.

I don’t know how much that will carry out but that could be another massive impact of the pandemic, is the distaste for density that makes certain activities really hard to carry out. I think that has big business implications as well.

Awesome. Thank you so much for joining us, Nick. Super, super fun and we’ll just keep watching the world as this unfolds.

Great. Hey, lovely to talk as always, Sarah.