Nicholas Bloom, a Stanford economics professor, studied remote and hybrid work arrangements for years before the pandemic—but never at the societal-wide scale that we’ve seen for the past two-and-a-half years.

We’ve interviewed Bloom several times since the start of 2020 and his recommendations for approaching hybrid have been widely followed by companies. Research Bloom conducts with colleagues shows that as of now, roughly 55% of US workers are fully in person, 30% are in hybrid arrangements, and the rest are fully remote.

We went back to Bloom earlier this week for his takeaways on how things have actually played out in practice, his prediction for how executives’ efforts to bring people back into the office more often will resolve, what common mistakes companies are making, and how a softening economy could alter workplaces’ trajectories. Here is a transcript of our conversation, edited for clarity:

You've said that hybrid is the best approach for most organizations, where the nature of the work allows it to be done remotely part of the time. We've now had a large-scale experience of hybrid for a few years. Based on studying that are there any learnings or nuances that you'd add to your earlier view?

Generally I think hybrid has worked really well. And the fact that most firms are adopting it tells you that. As an economist, there's something called 'revealed preference.' If people are always choosing, product A over product B, product A has to be, for the price, just a better deal. You don't need surveys if everyone's voting with their feet— you know that's really the best outcome. So the fact that hybrid has become dominant tells you it's generally working. The biggest challenge that I hear is basically of the 'choice versus coordination' type, and if pre-pandemic we'd switched directly into hybrid, I don't think it would be a challenge. Because people would be so happy to get two days off a week or three days off a week from going to the office that they'd accept being told by their employer that they can work from home Monday and Friday.

They'd see it as a godsend. But what's been tricky is there has been this transitionary phase whereby employers said, 'We're going to move to hybrid, but we want a bit of social distancing in the workplace. We're going to let people choose, to deliberately space them out. And everyone you speak to complains about the result of that. Most people, apart from introverts, have complaints about coming in and it being quiet and dead and people shouting into laptops—and what's the point of coming in to be on Zoom all day? This makes things harder now. I'm talking to a number of companies that say it's a bit tricky now to transition into, 'OK, you're coming in for the same number of days, but we're going choose which days or at least at the team level you need to coordinate.' If you look at the surveys, the key reason by far that people come into work is to see colleagues and work with colleagues.

When you ask people, when you come in to work, would you like your colleagues to be there, 80% say yes. So there's 80% of people that at least want their colleagues to be there. (The other 20%, maybe they're in the wrong firm. If you hate your coworkers, it's quite possible you're in the wrong job.).

But the biggest challenge—and maybe the one unexpected bit—is how much employees have got comfortable with choice of days in the office and like to flip that around each week, for whatever reason, childcare reasons, work, personal reasons, etc. I think it's going to settle down into typically team by team when it makes sense to come in together.

It's not always the case, by the way, that it makes sense for teams to come in on the same day. I spoke to one office of a general counsel who said 'We really need coverage of the full week.' And in fact, I was speaking to the administrator in our economics department yesterday who said the same thing: We need people in on every day. But for most organizations, being part of a team means you're working together. You are doing stuff together and therefore you want to be in together.

Why do you say that hybrid is a dominant choice?

There are two different ways to look at it. One is I've talked to so many companies. When you look in the news, pretty much all of them have gone to hybrid. The only examples are a few tech firms—but only tech firms—that have gone fully remote. If I were to ask you to name firms that have gone fully remote, everyone just lists a whole list of tech firms. I say, 'Name a non-tech firm that's gone fully remote' and there's normally silence. Airbnb, Yelp, etc. Some were already fully remote, like Automattic, Upwork. Quora went fully remote. They're all tech firms basically, or some media firms that publish entirely online, but they're basically bigger tech.

Then in terms of full return for managers and professionals, everyone just mentions typically JP Morgan, Goldman Sachs, and Tesla. It's not clear whether they are full return, even for those companies, which are in the media a lot for pushing it. And it is really just those three. I talked to actually the head of one of those three. And they were like, 'We're mostly in person, but less than pre-pandemic.' So they are mostly allowing some work from home. For example, taking Elon Musk's comment at face value, he said 40 hours a week in office. Well, these guys work 60, 70 hours a week. Who knows how to interpret that, he's a bit cryptic. But apart from those two extremes, basically everyone else has gone to hybrid.

And the other way is just to look at surveys of the labor force in the US from our monthly survey of 10,000 people. Some 55% are fully in person. That's kind of obvious who they are. That number is pretty surprisingly big for most people. Because your readers, like me are grads and used to hybrid and are surprised that most people can't work remote. Of the other 45%, 30% are hybrid. They're all basically graduates, well paid people. Then there's a remaining 15% that are fully remote. Initially when I talk to people, they think these are tech workers. They think Airbnb. But the US labor force is 150 million working Americans. And 15% of that is roughly 25 million, of which may be at most half a million are techies. The other 24 and a half million are basically IT support, payroll, HR, call centers. So the fully remote people turn out to also be mostly relatively low paid actually by number. There are a bunch of misconceptions that people are either fully remote or fully in person. Actually most senior people, most of your readership is going to be hybrid, I'd guess.

What can we say at this point about the business impact of this shift to hybrid?

I'll give you four reasons why people like hybrid and then see if I can actually put a profit number on it. So of the four reasons people like hybrid one is it keeps employees happier. Quit rates are down. In the randomized control trial we are just wrapping up, quit rates are down 35%. People report repeatedly in survey after survey that they value it at somewhere like 7% or 8% of a pay increase. There are various ways to get that number—I feel pretty confident because I've seen multiple, including experiments that pulled that number out. So number one biggest benefit is employees think of it as a 7% or 8% pay increase perk. A free pension plan is about the same value to employees.

Benefit two is it improves productivity. The numbers here are low, but the central estimate of over a number of studies and particularly placing most weight on the most rigorous kind of randomized control trial ones is 3%, 4%. Not huge but positive. Where do those numbers come from? They come from two places. Firstly is time. So if you are working from home two days a week, on average, you are saving 70 minutes a day commuting. You work from home, in survey data it looks like 30 minutes of that day is spent working more, 40 minutes is spent on other things. But as an employer, if you have people working from home two days a week, they work about an hour more for you over a 40-hour week. That's 2% more hours. So that's 2%. And then the other 1%, 2% comes from on the days you're at home if it's organized well, it's quieter so people are typically more efficient.

I have some extremely high-accuracy data on minute-by-minute activity and you can see people at home just take less breaks, less coffee breaks. They go to the toilet less. They're quicker. They take way shorter lunches. If you're working from home, your lunch break is typically 20, 30 minutes. If you're in the office it can easily be over an hour. So those two add up to about 3%, 4% benefit.

Three is harder to quantify, but it's definitely important, which is support for diversity, equity, and inclusion. Slack's Future Forum has some stuff. We have surveys showing that everyone, all demographics, ages, races, etc. like working from home. But there's a slightly stronger preference for people with kids, for women, for minorities. And what that means is if you are tough and force a full return to the office, you're going to see a higher attrition of diversity, diverse employees. That's a real DEI cost.

The final benefit, which oddly enough is last. but most people pre-pandemic put first is saving on space. It turns out that it's really hard to save on space because of hybrid. Typically people are coming in Tuesday, Wednesday, Thursday, because I want to work from home Monday, Friday. So you have this problem where you can't sublet office space on Monday, Friday. No one really wants it and it's hard with security and if you share office space, how are you going to lock up computers and filing cabinets and things?

So if you had to put a number on it to give a very, very rough figure for a typical business, with two thirds of costs from payroll, you basically reduce payroll costs by 7% or 8% because all else being equal, if people count hybrid as a 7%, 8% pay increase, you can basically not push their pay up by as much and have them not quit on you. So that probably is like a 5%, you know, net net addition to the bottom line. Productivity is up 3% to 4%. That's 3%, 4%, maybe more revenue, holding costs constant. It's harder to cost DEI and space, but you could easily see pushing up profits by 10%, 20% versus a full return. That's why as an economist I'd point out that firms don't do things that lose them money. They do things that make them money. That's why every firm just about out there is doing hybrid, because it's such a no brainer to increase profit. The bigger question is how to execute it. It's kind of like, how every firm that's in retail prices to market. It's really obvious you charge more at the airport than you're going to do in a low-income part of town. The harder question is exactly what prices to charge. That's the execution—it's a very similar concept.

In terms of execution, what are the biggest mistakes you're seeing organizations make and what should they be doing instead?

The biggest mistake is ceding full control over choice of days to employees over which days and how many days. It feels a bit like having kids honestly. You know, having teenagers where you want to clearly give them some choice and flexibility. If you give them too much, they go off the rails and it's just not good for them. They don't even appreciate it. Employees, when they come in want to see their coworkers. So if you let people fully choose, you'll find that in a team of 10 people, you never get a day when everyone's there. So every single meeting that the full team has, has to have people on Zoom. It's uncomfortable. It leaves people out. There's cliques that form. It's frustrating. The meeting ends, people on Zoom disconnect and of course the meeting continues in the corridor. So that's an issue.

The other thing is trying to get people back for too many days because of sunk-cost fallacies. Like 'We have this space, we should use it.' The fact you have this space doesn't mean you've got to force people in unnecessarily. There's plenty of evidence for the great resistance. And we find the great resistance is from two forms. One is we survey employers and only 80% are coming in for as many days as their managers or their firm wants them to. So 20% are not, which is a high number actually. If you surveyed people pre pandemic and said how many people are not coming in every day they're supposed to, it would have been close to zero.

The other issue is what's your manager doing about it for those that are not coming in? The largest response for about 40% is nothing. The managers are resisting too. The reason is that middle managers are basically saying, it makes no sense to be forcing employees back four days a week. So if my employees are coming in two or three, I'm just turning a blind eye. And it's kind of like that old saying about laws: a bad law makes an unenforceable law. Well, a bad management rule makes an unenforceable management rule because middle management just don't agree with it. They resist as much as the employees at the bottom.

The return after Labor Day seems to be a bit of a showdown at some companies where the executives want people back in and employees are reluctant, as you just said. How do you see that playing out?

I love this. I've been laughing at this Labor Day thing in the media because this is the third Labor Day showdown. I can tell you it's 2-0 to employees. So I don't have a lot of confidence. It wasn't even close. It's like Manchester City taking on Grimsby Town—I know what's gonna happen. It's going to be 3-0. They're just not going to come back. I'm not saying they're not coming back for zero days, but any firm trying to get people back five days a week, that's the fastest way to exit a whole bunch of employees.

I would support getting employees back two, three days a week, but mostly they're doing that. So I don't think you'll see a shift in the aggregate numbers. I'm going to make this prediction now, it's the very end of August. We're going to run the survey again in September and October and you can look at the data, but having looked at a series that's basically flat now in terms of the share of days working from home for the last six months, I really see Labor Day as a kind of a make-believe deadline that is going to come and go as it has in the past. There's been not just Labor Day, there's been multiple RTO dates that have come and gone.

To what extent do you see the uncertainty around economic growth affecting remote and hybrid plans, especially in terms of employee leverage for flexibility?

It's certainly true that the Fed is trying to engineer a soft landing. Even if it's a soft landing and not a hard landing, it's going to make labor markets less tight. So it's going to be harder to change jobs. Employers are going to have more power. The correct question is whether that will affect the uptake of work from home. I don't think so. Going back to the four reasons, keeping employees happy, increasing productivity, supporting diversity, saving on office space—collectively, I don't think those are more important in a boom than a recession. For some of them you could argue actually the reverse. In the short run, it will fluctuate a little bit, but I don't see it changing more than a few percent as it goes up and down.

The big issue in the longer run is technology. If you look back (this is the thing I tweeted out a few days ago), over time working from home has been continuously improving. Increasing its share of days, and the reason is technology's got better and better. When I first started working on it in 2000, it was like telephone calls and conference calls and emailing files was not great. My parents both worked from home when I was a kid. And for them, it was them picking up piles of paper and was really not very pleasant.

The pandemic has led to a sixfold increase in work-from-home days, which has led to an enormous increase in the rate of technological progress in hardware and software to support it because firms see it as a much bigger market. There's literature going back to Schumpeter in economics, almost a hundred years on market size effects. If markets get big, firms innovate to support it, like in America, there's lots of old people and pharmaceutical firms are developing drugs to support older patients.

The same thing precisely is happening with working from home technology. Things like virtual reality, augmented reality, holograms, better AI to support audiovisual. So that's going to mean certainly in three plus years—maybe even before—the trend is going to continue to go back upwards again. So actually if I was making any long-run decisions over organizational structure, strategy, office space, growth rates, I would actually predict if anything work from home is rising, it's not falling. Certainly five years now, I think it would be above where it is now. I'm pretty confident to make that prediction.

In the past, you've said that it's really hard to have both hybrid and remote workers because the hybrid workers come in to the office and have to sit on Zoom calls with the fully remote workers. Has that changed at all?

It's tough. For some folks, it's a real issue. I spoke to a friend yesterday actually who has hired a bunch of people into his team over the pandemic on fully remote. Stanford has done the same. Basically the problem is, imagine you have some people coming in three days a week and others fully remote, you know that the people coming in three days a week were forming in a clique. They get promoted faster. The people in fully remote will get angry, upset. You may even see lawsuits because being fully remote is also correlated to demographics. So there's a real problem. You find a group of people of certain demographics are not getting promoted as much and left out. That is a toxic recipe. So generally if you have fully remote employees, I would try and move them to a different team.

You could even have two teams, one that's fully remote, one that's hybrid. So that at least everyone they'd be compared against and working with is on a similar schedule. A lot of organizations have ultimately said, 'This is now a hybrid job. We'll give you six months to adjust. If it doesn't work after them, we suggest you move on.' So my advice generally to employees is now is still a pretty tight labor market. Now is a good time to move. For example, if your firm moves from New York to Cleveland and you don't want to move to Cleveland with them, just leave and find another job. Location is one type of attribute of work, as is hybrid versus fully remote. And if it doesn't match people's personal needs, they probably should just change jobs.

The handbook for this new era of business doesn’t exist. We’re all drafting our own as we go along—and now we’d like to start doing so together. You can sign up here to receive this briefing by email.