How many organizations approach hybrid and remote work has been deeply influenced by the work of Nicholas Bloom, a Stanford University economics professor. Bloom has studied remote work for over 20 years and his WFH Research initiative and annual academic conference about remote work, both in collaboration with others, have crystallized key research. (Longtime Charter readers will be familiar with Bloom from previous interviews and events.) We reached out to him in the wake of his latest conference and as we near the six-year mark since the start of the pandemic for an update on the latest findings on how work is best organized. Here are excerpts from the conversation, edited for space and clarity:
What do you know now about work that you might not have a year ago?
The first thing is partly from the Atlanta Fed. I just was looking at preliminary evidence for a big survey run for 1,200 firms and I think we’re in ‘agree-to-disagree’ equilibrium. Levels of work from home have not changed much for the last two years. That seems very odd. Why is that? It turns out when we surveyed more than a thousand managers, there are just a range of views. If you survey firms where they’re all back in person, the CEOs claim work from home reduces productivity by 12%. If you survey firms where they’re all fully remote, the CEOs tell us working from home increases productivity by 9%.
One explanation is they’re talking about different activities. One is call centers, which typically is best done at home because it’s quiet. And the other is maybe marketing campaigns or investment banking or something where you’ve got to be in person. But a lot of it is just people don’t agree—as much as they don’t agree on what food is best, what’s the greatest movie ever, or which is the best town to start a business in. It’s not like in 10 years’ time we’re all going to agree that Italian is the best food. There’s going to permanently be disagreement. And that’s where we are with working from home. It’s 2025, there seems no convergence in views now for two years. By 2023 everyone had figured it out and come to their own conclusion and nobody’s really changing their mind.
And, actually, I think that’s fantastic. It’s like the modern economy. Think of food. It’s great that there’s Thai, Italian, French, Spanish and Mexican. You wouldn’t want to have them all in one flavor, in the same way that it’s because people have different preferences and they go through different life stages and we’re not all of the same views. It’s the same with work from home. You see folks in their twenties really want to go into the office four days a week. You’re in your thirties, you’ve got two young kids, you actually want to work from home maybe three, four days a week. You may be injured or have a back issue—you want to go fully remote. And maybe in your mid-fifties you want to go back in because you find it lonely at home, you’re an empty nester. I actually think we’re in a great place. It’s kind of like ‘celebrate capitalism.’ I also think it’s why the current administration trying to force everyone back five days a week is wrong. There’s a couple of reasons why it’s wrong, but one is why restrict choice? Just let the market provide choice. I mean it’s all good and there is no one size fits all.
Is that because all these things are good options? One of my takeaways from your research is that for most firms, the best arrangement is roughly three days a week in-office hybrid as a way to maximize retention, productivity, and all the things that people want to maximize. There may be lots of different kinds of restaurants but, in this case, is there an objectively better arrangement for white-collar companies?
For a typical set of professionals and managers, that is the best policy. But, you’ve got to be well managed. So it’s kind of like saying you can go faster flying a plane than you can go in a car, but if you can’t actually fly a plane, you ain’t going anywhere. You’re just going to crash off the side of the runway—so you’re faster in a car. It’s like that with three-two [three days in office, two days remote weekly.] It’s actually quite hard to get it right. Two things are critical. One is good performance reviews. When someone’s in the office you can see what they’re doing. And if you are my boss, you can see, am I at my desk? When you walk past, you can see what’s on my screen and you can hear my phone calls and you have a pretty good sense.
It’s not perfect, but I would call that four- or five-out-of-10 management. I could be typing really slowly, but it’s a lot better than nothing. If I’m at home, you have no idea what the heck I’m doing. So unless you have performance management systems around outputs, achievements, targets met, customer surveys… Well-run companies typically have six-month performance reviews and every six months have a range of things, talk about how you’ve done on new ideas and billing and producing training, etc. So you need good performance management systems.
The other one is you need to coordinate when you’re in the office and actually enforce it. I have talked to a bunch of companies that will say, ‘We ask for people to come in three days a week. We don’t tell them which days and we don’t honestly check up on it.’ And when you actually talk to the employees in these organizations, what you discover is it is every which way. What happens is employees are typically quite dissatisfied because when they come in, the office is half empty and they spend the day on Zoom and they’re like, ‘What is the point of coming in to spend the day on Zoom? I’m only doing it because my boss is checking up on me. I don’t think they really know. So I come in one or two days a week.’ So you get everyone coming in one or two days a week just to do enough to avoid getting in trouble. And then it’s a really awful one or two days a week. So the second thing is why not just coordinate on Tuesday, Thursday and force it? Make sure people come in and then it kind of works.
You may think enforcement’s terrible. But the reason to enforce it is there’s also a spillover effect. If there’s six of us in a team and you decide to stay at home and watch the Champions League, the other five of us when we want to talk to you have to get on Zoom. If we’re on a team meeting, we have to do it on Zoom and it’s very annoying. So three-two is probably roughly the best thing, but you need to have good performance evaluation and good coordination with enforcement. If you don’t have those two, that’s probably no longer the best thing to do. And probably you want to have people come in more days a week. Badly managed organizations often struggle and force people in five days a week, while well-managed organizations seem to make hybrid successful.
Where are you seeing the intersection of AI implementation and hybrid and remote work?
For your readers, I would say fully remote is really risky in the era of AI. Why? Because AI is great at a lot of fully remote activities, but it can’t empathize with an employee, give them a performance review, win a sales pitch—which all typically are done in person. Also I’ve heard execs say they’re nervous about AI because of authenticity. They really are not certain when someone emails them or something, whether it’s really them. So AI is actually the big enemy of fully remote. It doesn’t mean you’ve got to come in five days a week, but just coming in one day a week is enough to show people you’re not AI and that you’re actually doing your job.