Work From Home And Productivity – A Conversation With Nick Bloom

>> Steven Davis: Welcome to Economics Applied. I'm Steven Davis at the Hoover Institution. Today I'm here with Nick Bloom to talk about work from home and productivity. Nick is a Stanford economics professor, prolific researcher, noted authority on management practices, economic uncertainty, and work from home, the topic of today's discussion.

His household spans major parts of the Anglosphere. Nick hails from England, as you might notice from his accent. His wife is Scottish, his kids are American, and he lives here on Stanford's campus, welcome, Nick.

>> Nick Bloom: Thanks, Steve, very much for having me on. And I should say, we are long term friends and co-authors as well, to add into the mix.

>> Steven Davis: We are, and I was just gonna put the cards on the table in that respect. We worked extensively with Scott Baker on economic policy uncertainty and stock market jumps, and with others. And in recent years, since the pandemic struck, we've been studying the rise of work from home and its implications.

And I tried to count up the number of collaborators we have in the various projects, and it's more than a dozen. And I couldn't figure out exactly how many we had. So, also on that same front, along with Jose Maria Barrero and others, we conduct the US Survey of Working Arrangements and Attitudes, that's monthly.

And the global survey of working arrangements with a larger team, that's roughly at an annual frequency. So audience members can find survey results and sign up for our monthly newsletter at wfhresearch.com, if they're interested. So that's the advert. Let me just set the stage for the discussion today, and I'll do this briefly.

But I think you and I both come from a similar viewpoint on where we are and what's happened and why. And it's gonna be a premise, I think, for the rest of the discussion. So, full days work from home, as of mid 2023, or somewhere around a quarter of all paid working days among Americans 20 to 64 years old.

And that's screening out people who only earn like 10 or 20,000 a year. So we're screening out low wage, part-time jobs in that number. That number comes from the Survey of Working Arrangements and Attitudes that we run with Jose. It's about four times the rate in 2019. Okay, so that's the big shift to work from home.

The pandemic has also triggered an innovation speed up in technologies that support remote collaboration and video conferencing and like, and that reinforces the shift to work from home. You and I, along with folks at the Federal Reserve Bank of Atlanta, also survey business executives regularly. And in July, we asked them, what do you think is gonna happen to work from home at your own company?

Not in general, the macroeconomy, but your own company. And then we tallied up the responses and averaged them. And business executives think, on average, that work from home rates are actually gonna go up a little bit over the next five years. So that's kinda, I think, some factual background.

You and I have written at length about how the big shift to work from home came about, why it will stick. For today's conversation, let's just take it for granted that work from home is here to stay. There are profound implications for the economy and society, and I wanna focus today's discussion on the productivity implications.

So, Nick, you are one of the world's leading experts on this topic. In fact, you wrote a major study on the topic years before the pandemic, titled, Does Working From Home Work? Evidence from a Chinese Experiment, and that paper appeared in the Quarterly Journal of Economics. So tell us, Nick, how this study came about.

Not many people were thinking about work from home before the pandemic, why you wrote this study and what you learned from it.

>> Nick Bloom: Great, so, the study came about, in all honesty, from a piece of good fortune. So, James Liang, who was the founding CEO, and at that point, Chairman of Ctrip, happened to be sitting in my class.

This is very Stanford, I had somebody who is worth hundreds of millions of dollars quietly sitting in the back of the class. I had actually done some work on work from home back in 2005, so I'd long been interested. One of four kids, both my parents worked full time for the British government, they used to work from home sporadically when there were childcare disasters in the 80s.

So I'm 50, so I was like, I remember in the 80s, I mean, it was horrible back then. It was shuffling pieces of papers before the era of the personal computer. And James said his company was in Shanghai, real estate was really expensive, and they were trying to figure out how to expand the call center, which is pretty critical.

This is a call center operation. And they thought, well, look, maybe we have folks work from home. And their fear was people just goof off. There was the joke about the three great enemies of work from home was, the bed, the fridge, and the television. And someone would fall victim to one or many.

And so they set up this experiment under the view of, look, they're obviously gonna goof off, how much will they goof off? How much less productive will they be? Let's compare that to the space savings. So what the experiment did is it took 250 people and it randomized them by even and odd birthday.

And if you, inverted commas, the winner, the treated people, you got to work from home four days a week. And if you're the even birthdays, you had to come in as normal, all five days.

>> Steven Davis: Nick, everyone have to participate in this randomization, or could you say, no, I know that I won't be able to resist the fridge and so I'm just gonna opt out.

>> Nick Bloom: Great question, so, only about half of people volunteered, so a typical team's got 15 people. In a typical team, that means maybe seven volunteered and eight didn't. And then within the volunteers, we randomize. So in a typical team, you're looking at three folks work from home. And so we tracked them, we have amazing data, cuz they're in a call center.

So first thing's they have headsets on and the calls drop in. They randomly listen into 1% of calls to give you a quality metric. They're typically working shifts, let's think Monday to Friday, 9 to 5. So what did they find? Three findings. One is, amazingly, they were stunned, performance went up 13%.

So they were expecting a negative number, turns out quality was totally the same and they were processing 13% more calls. And like, how on earth did this happen? And it turns out, 9% of it was they're just working more minutes. And if you look in the data, turns out if you're working from home, you're not late for work, so your shift is 9 to 5, you really start at 9.

In the office, people are coming in 9:20 cuz the motorbike broke down or whatever. They're taking much shorter lunch breaks, shorter toilet breaks, less breaks, less sick leave, put it all together, they're working 9% more. And then the other 4% is they're just processing calls faster per minute.

When you interviewed them, they said that's obvious, it's quiet at home, it's really noisy in the office. There's a cake in the breakout room or somebody crying, or there's someone that told me in the cubicle next door to her, someone would clip their toenails under the desk.

>> Nick Bloom: She said, they think I don't know, but I tell you, I know, I totally know, and it's disgusting.

So that was finding, I'll pause there, but that was finding one that's amazement. They were 13% more productive if they worked from home.

>> Steven Davis: So, I want to go back to people who opted out. And the reason I ask about this is the experiment that you're describing has already got a form of selection.

People who know that they're not going to be effective working remotely, either because they can't concentrate or they don't have enough space to work at home are presumably disproportionately among those who opted out. And that's quite different than the experience in the midst of the pandemic, when everybody was compelled in many jobs to work from home.

So did you ask these people who opted out, why? Or do you have any insight to share on that?

>> Nick Bloom: No, we didn't, but a really interesting thing exactly on point is what happened at the end of the experiment. So I'll very briefly give you results two and three, then go exactly the point.

So result 1, productivity is up 13%, result 2, quit rates dropped by a half, on average these folks really like it. I mean, so much that the quit rates fall by half, which is an enormous difference. Fact 3, unfortunately, promotion rates also fall by half, and so that's the sting in the tail.

You're not in the office, you're not seeing what's going on, you're missing out and stuff. Then in terms of opting out or selection effects, at the end of the nine month experiment, the company looked at it and said, this is great. We're saving on space, people are 13% more productive, they're much less likely to quit.

Promotion rates, something we can really work on, but it's not so directly impacting the company's profit, so it rolled it out. And what you saw is, of the people that originally opted not to take part in the experiment, one third of them changed their mind. So these are folks that said, you know what, maybe I do wanna work from home.

Interesting enough, 50% of the people that did initially volunteer also changed their mind and came back into the office. And that group we did interview, and we asked them, why are you now voluntarily commuting five times a week rather than once? 50% of their pay is performance pay, so they're implicitly taking a pay cut if they're more productive, why are you doing that?

And they basically said, it's incredibly lonely at home, it's lonely, depressing, isolating. On your point, also, what was stunning, we wrote it up in the papers. If you look at the people that volunteered, so the hardcore that stuck with it, their productivity increase was not 13%, it was like 22%.

So it's classic, exactly as you said. Some folks, and look at students, for some students that work in their bedrooms because they're very focused, can do it. Other people need to go to the library, and others, as we know, don't work at all. So there's a spread, and it seemed to be the same thing.

People know they have pets, or somebody told me about a dog, that they had their dog at home, and the dog would sit at their feet looking up for snacks all day, and they said, I can't deal with it.

>> Steven Davis: Two follow-up questions, Nick, these are almost questions for James, maybe we've talked to him about it.

If people are really lonely at home and that's why they don't wanna work from home four or five days a week, well, maybe what if they want to work from home just two days a week? And is that an option? The second follow-up question is on this promotion point.

So again, I'm putting myself in the CEO's shoes. And I'm learning from this experiment that most people are more productive when they work from home, at least the ones who are willing to do so. And yet they might reasonably be concerned about their promotion prospects. So is there some way to thread that needle and make it so people can be rewarded for their higher productivity when they're working from home without hurting their promotion prospects?

Did the company put in any changes after your study to address either of those concerns?

>> Nick Bloom: Great, so I take it in reverse, on the second part, on promotion, I think there are two things going on. One is what I call discrimination, you're literally not in the office, people forget about you.

They don't treat you, and it's nothing to do with real performance, that you can try and train against, and there are some ways to deal with that. It's not perfect, but you can help to reduce the impact. The other factor is how you quite measure productivity. I was actually giving a big exec ed dinner speech last night, and this came up a lot, is we are measuring calls processed per minute.

But there's another side of productivity, who's learning the company culture, knowing what's going on, setting yourself up for management. What you heard is the folks in the office said, well, look, we are 13% less efficient, but part of that time is spent over lunch and over coffees, chatting to people, and it means we know what's going on, and we're actually better suited to be managers.

So from the firm's perspective, the discrimination part, you can train against. The training thing, you do need some in, it's a short and long run trade off. When I talk to firms now, the big thing really is everyone in the team should be on the same schedule. If all of them, all 15 have been working from home four days a week, we wouldn't have seen this effect.

The tricky thing is you have a team of 15, of which 3 are staying at home and the other 12 are in the office. So it's pretty obvious who's gonna get the promotion. So that's really, this issue with a lot of companies, you've gotta have some parity amongst pools of people competing to get promoted because otherwise you have this nasty leave out issue.

>> Steven Davis: So the last question on this study that comes to mind, did the people who participated in work from home, what happened to their productivity over time when they were working from home? Did it immediately jump? Did it gradually increase how to work remotely? What was it?

>> Nick Bloom: Yeah, it was pretty flat, I think it went up.

I have to go back, I did this study now 13 years ago actually, we're now late 2013, and I did it back in 2010 to 2012. It was pretty flat over time, I mean, this matches what the story said. It's about saving time and about qua, there wasn't so much about, I mean, there were two or three days of drop, so there's some chaos initially where people's computers weren't working, etc.

>> Steven Davis: Okay.

>> Nick Bloom: Then other question you asked about why not do two days a week, that company was called Ctrip. It actually bought another company called Trip, and they, effectively, merged, and it took the Trip name, so it's now called Trip.com, same company, basically. It's one of the global big three travel agents with Expedia and Booking.

And they in 2021, 22 did another experiment, which is more like what you asked about. So they randomized work from home for two days a week, same, even-odd birthdays. But this time it's on managers and professionals, so probably much more like people that are listening. So they're all grads, one third were post grads with CS, MBAs, etc, masters.

There they found absolutely no effect on productivity. So this is people writing code, doing marketing campaigns, doing accounts. They looked at promotions, lines of code written, performance reviews, all kinds of stuff, just a flat zero everywhere. They again found a dramatic reduction in quit rates, fell by a third, they didn't find any effect on promotions.

That was a much more positive story, in a sense. Cuz when the firm looked at it and rolled the whole thing out again. And they said, look for professionals, for senior folks, having them work from home two days a week, at least the option to, they're only taking up typically one and a half days a week, doesn't seem to affect performance.

We're organized, they come in, we get our stuff done on the three or four days in the office, but they love it. And you talk to the employees, and I'd say, look, if I wanna go to the dentist and now at home on a Friday, I can do that.

If I wanna go see my parents, I can leave at 3 PM on the Friday and make up a bit of time on the train. So that is why hybrid has become so dominant in the US, there doesn't appear to be much impact on productivity of any, actually but employees really like it.

>> Steven Davis: Right, so let me just make one broad point, which I know you're well aware of, but it's worth making for the audience. You've described two experimental studies where, by and large, there were positive effects. And the first one on productivity and retention, on the second one, employee happiness and retention, if not productivity.

But I think it's really important to understand that there is no uniform answer to the question of how work from home affects productivity. It's gonna vary across people, as we've already discussed, tasks, jobs, workers, managers, and so on. It's also gonna depend on the organization's capabilities, whether they've optimized their management practices, their performance evaluation and so on for work from home.

So maybe you can say something about settings where work from home doesn't work very well.

>> Nick Bloom: Totally, so there's two different broad things. One is fully remote. So to be clear, fully remote is you're working from home five days a week, week in, week out. The other is what I'm going to call hybrid, where you're typically working from home, let's say Monday and Friday.

So the evidence on hybrid, I discussed a bit of it, but generally it's flat. So typically the evidence on hybrid is it's not massively positive or negative. There's a number of studies that will honestly come out close to zero. It's super popular cuz employees like it and it reduces quit rates.

So let's come back to fully remote. Fully remote, the numbers are all over the place. So there are figures of -30 up to the one we discussed earlier is +13. So youve got a 43% span. So how on earth is this going on? If you look at the big negatives, the things that you find drives these big negatives is basically no planning and bad management.

So the big negatives tend to be firms that rushed to work from home early on in the pandemic. And we have some kind of treatment and control rough setup. But you can imagine if you suddenly go fully remote in April 2020, there's no equipment, managers aren't trained, you don't have performance reviews.

There's a lot of people using horrible surveillance screenshots, I mean, it doesn't really work. If you look at the positives, take the Ctrip experiment I mentioned. They had remote quality reviews cuz they're listening in. There's a lot of performance data. They're well managed. They saw this thing coming.

They spent three months training and getting up and ready. So you're totally right. It's very variable. And I think fully remote can work quite well for the types of tasks that are relatively repetitive, in particular, easy to monitor and evaluate for people that wanna do it. So take Stanford, our employer.

They have about 1000 people working fully remote. These are all volunteers, no one's forced into it. They all have pretty rigorous performance reviews. They're often doing a lot of things like IT support, payroll, call centers, where there's a large flow of relatively similar stuff. And so it's easy to manage and motivate them.

The other extreme, marketing. Creative marketing or maybe federal government employees. That's much harder. That's more creative, harder to monitor, harder to evaluate. That's a better fit for hybrid.

>> Steven Davis: Right, so let me say a few things in response. First, the way I think about the point you made, if you do it with no planning, it doesn't work very well.

And of course, when the pandemic hit, it was a shock. So there was a lot of massive shift to work from home, including fully remote work, without much preparation. And that was part of the reason why a lot of those early experiments show negative effects. So I think there's a broader lesson there, which is, if you're going to make remote work, whether it's fully remote or even hybrid work effectively in your organization, it does take careful planning.

It takes complimentary investments in IT support and performance evaluation. It may take a very different set of skills among managers. So it's not an easy transition. Even when work from home, in settings where it will work well, it's not an easy transition. The second thing, another thing you mentioned is, well, the workers loved it, even when it didn't appear to be productivity enhancing.

And here I wanna introduce a key conceptual distinction that often is running around in these discussions of what workers think versus managers think, but as a source of confusion, because it's not stated explicitly. So when you go to managers and you ask them about productivity, they're naturally thinking about how much work gets done by my staff per unit of paid time.

If it's an hourly work or how much per hour, or if somebody gets paid weekly or monthly, then it's how much work do they do per week or per month. That's a perfectly natural way to think about productivity. In fact, it's what the statistical agencies typically try to measure, some notion of output per hour.

But if you think about it from the workers' perspective, they're saying, how much do I accomplish for the time I have to devote to my job? And that time includes commuting time and also making yourself presentable in the morning before you rush off to work, so you can have a big time savings on the worker side, that, from their perspective, means I'm more productive.

I get more done for the same amount of time devoted to my job. But from the manager side, doesn't look like much. So, just to put some numbers around this, suppose you work an eight hour day and you've got a 30 minutes commute each way, and you're equally productive as conventionally measured, whether you're at the office or at home.

Well, then the worker says, wow, I get the same amount of work done and I spend an hour less per day. That's more than a 10% productivity boost. So I think that distinction actually is running around many of the contrasting narratives you hear coming from workers and managers.

And the other thing to note on this score is that from society's perspective, it's that less conventional approach to productivity that's probably a more appropriate one. Because from society's perspective as a whole, what we care about is how much market output gets produced for the time it takes to produce that output.

And commuting is part of that time. And so, back to these negative effects of fully remote work that several studies find, that's using the conventional measure of productivity that doesn't factor in the time savings associated with commuting.

>> Nick Bloom: Totally agree. Yeah, I mean, I think this is one of the most positive things to happen to US society in a while.

If I go through it, it's profitable for firms. It's why firms are mass adopting hybrid. I'm not trying to push fully remote. Mostly people are not fully remote. But hybrid is make saves firms money by reducing retention and recruitment costs. Is great for employees. We know from our work.

So Steven and I with Jose, for example, be collecting data showing employees value that at about the same as an 8% pay increase. And it's good for society. It appears to reduce pollution because there's a big reduction in commuting. And there's some amazing studies looking at time use showing that all this time that people save from commuting, about 10% is spent on time with their kids.

So you can imagine in the long run, we have this huge educational shock from the pandemic. Some of it is thankfully being undone because parents have more time to spend with children.

>> Steven Davis: Yeah, exactly. Let me elaborate on this a bit. I'm fully on board with the idea that, on balance, the shift of work from home is really a huge social benefit for the reasons you described, doesn't mean there aren't losers in the process.

And I guess what I would add to your remark is part of the change that's happened, that I see is a good thing, is before the pandemic, for 95% of the population, if they wanted what they regarded as a good job, they were kind of stuck going to their employer's work site.

If they were full time, five days a week, you didn't have much choice in that regard. Now there's this explosion in the variety of working arrangements on offer, concentrated among better educated people. It's less true for people who work with their hands or do face to face work with customers or have to work in a factory.

But among the college educated class, for many people, they can now choose among jobs that are traditional in the office five days a week, fully remote or hybrid. Hybrid is the biggest expansion. That expansion in choice is just tremendously beneficial because it allows people, if they want to, to save on commute, to spend more time with their kids.

But if they don't want to, they don't have to, if you are the kind of person who thrives on working in the office five days a week, you can go work for Elon Musk or some other company that really wants to do business that way. So there's this explosion in the variety of working arrangements that I think is tremendously beneficial, and you and I have discussed, we kind of like to do research on this, but I don't really have a lot to report yet.

But I'm of the view, and you can tell me if you disagree, that the whole shift of work from home has expanded the opportunities to participate effectively in the labor market by groups that were somewhat disenfranchised or on the margins before. People who live in remote, out of the way places, people who might live in urban, poor urban areas where there aren't many good jobs nearby, people with young kids who really wanna be at home with their kids, maybe they still wanna work, but they wanna be at home around their kids.

And people have physical mobility impairments, for some people, the commute is just a huge physical challenge. So that's another reason why I think this shift is actually got positive social consequences.

>> Nick Bloom: No, I mean massive effects. One of the exciting things I think about the long run effect is the impact on growth so, as we know, inflation is high, why is inflation high?

Because the economy can't produce enough goods and services, we need more people, labor markets are tight. Turns out, if you look in the data, work from home may generate a big increase in Americans that can work, so two groups, we have some kind of preliminary figures on, one is Americans with a disability.

So if you look at labor force participation rates for those without a disability, which is so like 85% of the workforce, that's flat pre- versus post-pandemic. If you look at participation rates for Americans with a disability, which, to be clear, is quite a broad terms, that would include people that can't walk long distances or have difficulty carrying large weights or serious back pain.

That's gone up by about 5%, and that's equivalent to about 2 million extra Americans that are now in the labor force. So I have a friend of mine, and she works from home now four days a week, and she says, look, I have real back problems, makes commuting unpleasant, also makes sitting in a desk chair really horrible.

When I'm at home, I can lie on the bed and she sets something up where the laptop's, like, affixed to this frame, and she can work lying on her back, and that's a complete game changer for us. So theres 2 million more Americans it looks like are working now with disabilities, and then another group we've seen some improvement on in labour force participation is also female employees.

So that's gone up by about 1% versus pre pandemic, and it looks like that's maybe an extra million employees, a lot of folks that would have heavy childcare duties and can now go to work. Think about my parents when they were close to retirement. They kind of didn't want to go into the office five days a week, but they'd maybe go in three or two or so I think it's hugely positive.

I think we're going to see many millions, as you said, many millions more Americans working, there may not be entirely full time, but they're definitely working. I think it's really positive, and actually it benefits all of us because it helps combat inflation.

>> Steven Davis: Yeah we're gonna come back to this topic on this podcast with a future guest.

So I agree with you, it's really important, so, Nick, last question. What are the big open questions about the connection between work from home and productivity?

>> Nick Bloom: I tell you three questions that come up a lot, they're kind of practical/research, so one of them is, how many days a week?

Everyones gone to hybrid, but even last night someone said, look how many days? Im like, to be honest, Raj Chaudhary has a bit of research in this at Harvard Business School, but theres not much, that isn't a huge sample. We honestly don't know, should you come in three, four days?

I'm involved, I'm an advisor to a few startups, I'm like some of these have invested in a sense, my companies have a personal interest even for them. I'm like, I don't know, so how many days? I think, as you said earlier, it's going to depend. So if you're a coder, mostly working on your own and occasionally catching up with team members, it may be one day a week or one day or retreats every three weeks will work.

On the other hand, if you're designing some maybe complicated marketing campaign that you need to work with six other people, you may need to be in four days a week. So one is how many days, a second is how to save on space costs, so it's been an enormous frustration amongst employees and CEO's about we havent managed to reduce space.

So I ve heard this so many times from heads of real estate saying this CEO is beating me up because the CEO is saying footfall is down in the office by 50%, but space is only down 10%. I'm failing, I'm missing that other 40%, and the problem is everyone's coming in on Tuesday, Wednesday, Thursday.

So research or a practical question is, can we somehow reschedule things? So, Steve, you and I teach, and universities know this, with teaching rooms, we don't all give class at the same time, we have a schedule, so the rooms are pretty full. But we're teaching at different times, so will that be the future of firms or will different teams?

And then the third question that comes up a lot is on coordination. So there's this big battle on hybrid coming back to how you motivate it, should you let each person choose? We like choice, Americans like everyone likes choice, should I get to choose if I say when I come in Monday, Friday, because my spouse comes in Tuesday, Wednesday, Thursday?

Or should we coordinate? And should each team or company coordinate? And the benefits of coordination is the reason Americans want to come in or anyone wants to come into the office mainly is to work face to face with co workers. So I actually see coordination slowly winning out, but its a tough thing, managers are struggling to persuade their employees, and I think in econ terms, the spillovers, positive spillover.

So if you and I work together, if you come in, it benefits me too, and that's really why coordination seems to be trapped. But again, the research on this, this is all anecdote and practical evidence, much more than research.

>> Steven Davis: Yeah, youve reminded me that even though work from home is here to stay, exactly the form it will take is still very much a work in progress.

And companies are grappling with exactly this issue for reasons of coordination and what it means for productivity, but also for the cost savings reason you talked about earlier. So, Nick, I'm gonna wrap up here, and one more advertisement, this time for a three day conference on remote work that you and I with others co-organized.

That was held, I think, last month at Stanford, sponsored by the Hoover Institution at Stanford, and the Stanford Institute for Economic Policy Research. There are a lot of fabulous papers presented at that conference, and some of the presenters will be featured on future episodes of this podcast. I'm lining that up now, if you're interested, our audience members can find links to the papers, the slides, and videotapes of the presentation at my website.

That's stephenjdavis.com, it's the website. Nick, I wanna thank you for a truly fabulous conversation and also just thank you for being a collaborator with me in this really exciting area of research. There's a lot going on and everybody's interested in it, so thanks so much.

>> Nick Bloom: Fantastic, Steve, it's great to do this, I should say.

We are hybrid, you're in the office, I'm at home, and we're actually doing this over Zoom. So I don't know if people can imagine this, but five years ago, you'd always do podcasts in a recording studio. But now, this is the Brave New World.

>> Steven Davis: Exactly, all right, well, for more podcasts on the economic and societal implications of remote work, stay tuned to this channel.

Bye for now.