Working From Home Doesn't Work

Zoom recently announced they are calling some workers back to the office. The company’s new policy requires close workers to return to the office at least two days per week. This announcement is only a continuation of a longer trend of bringing workers back. Many tech companies including Google and Amazon are calling reluctant workers back in after a years-long experiment caused by COVID-19 policies.

These bigger cases are representative of what’s happening economy-wide. According to Pew Research, 43% of workers who have jobs that can be worked remotely said they worked from home all of the time in January 2022. By February 2023, that number fell to 35%.

Why is this happening? Remote work isn’t working for companies. A recent working paper from the National Bureau of Economic Research corroborates this. The authors study the data entry sector in India. They find that working from home causes an immediate average 12% loss in productivity and accuracy, and the loss increases to 18% over time.

It’s important to note that this is, again, the data entry industry. It’s hard to imagine many industries more suited to make a permanent leap from in-office to at-home post-pandemic.

To be clear, remote work is still more common than it was pre-pandemic. Also, many workers are still retaining some hybrid remote work, according to Pew. But the trend is away from remote work. This flies in the face of a lot of speculation that COVID would push us to a zero office future.

So what’s behind the waning status of the remote work revolution? To understand why remote work isn’t working, we should first consider why there was so much optimism about the change to remote work.

Switching to Save
I was always extremely skeptical that remote work would permanently replace office work on a large scale. To understand why, think about the cost of office buildings. Office space in a busy place like Manhattan can run tens of thousands of dollars per month. And these are individual suites and floors. Imagine the cost, implicit or explicit, of an entire office building.

Offices are costly. This may sound like an argument in favor of remote work—at first. But ask yourself, why would businesses spend so much money to have a physical office? Why not get rid of the office building and have employees telework?

There are two possibilities. Either lots of businesses are making millions of dollars worth of mistakes over the years, or having a physical office space provides significant financial benefit to companies. In order for a very expensive office to make sense, it must provide very large productivity benefits.

I tend to believe the second reason is true. I think there are substantial benefits (which I’ll discuss more later) to sharing physical space with coworkers (and other business partners). Anyone who thinks these offices are a mistake should consider starting their own business—you’ll surely be more cost-effective than your competition if you cut out tens of thousands of dollars of mistakes each month.

Before the pandemic, however, fans of remote work had a good theoretical response. Costly office buildings demonstrate there must be a big upside to buildings. But it’s also the case that switching your company from a physical-meeting-space company to a remote work-based company is expensive.

So it could have been the case that switching was so costly that it made the office space expense worth it. In other words, maybe physical offices only made sense because it’s too hard to change a whole company to remote work.

It seemed unlikely to me that the cost of teaching employees how to effectively use email and conference calls exceeded the cost of an office building, but it was at least theoretically possible.

But COVID changed this. Due to pandemic policies, businesses were forced to bear the switching costs. Companies quickly adopted technology, created processes, and established workflows to transition into an era of remote work.

So switching costs have been paid. In spite of this, businesses are still moving back into the big expensive buildings. How could this be? Well, the simple answer is work from home is just less productive than work in the office, on average.

The Reversing Revolution

The numbers in the NBER study are staggering. An immediate 12% loss in productivity is a massive drop. Admittedly, this is only one study. But the study is bolstered by the fact that so many businesses are willing to hold on to large office buildings and call their workers back in. Finally, there are also good theoretical reasons why we would expect remote work would lead to lower productivity in corporations.

To understand these reasons, we need to think about teams. Consider two people who are moving the contents of their office into a moving van. Let’s say Jon can move 30 items an hour by himself. Patrick moves 20 items an hour by himself.

How much do you think they could do an hour if they worked as a team? You might be tempted to say 50, but that isn’t quite right. Think of how long it would take to move a couch by yourself. You could probably do it, but it’s awkward to carry alone, and it’d be hard to get it through doors and up or down stairs. Two people moving a couch are probably more than twice as fast as one person trying to do the same.

In other words, the team as a unit is more productive than the individuals’ work simply added together. It’s more likely that this increased efficiency will mean the team could move more, say 70 boxes, in the same amount of time. The whole is greater than the sum of the parts.

Economists call this team production.

Insofar as people work more cohesively in teams when they are physically present together, the benefits of team production will increase. It’s important to note that those who are in-office still have access to all the tools for teamwork remote workers do. You can Zoom the person in the room across the hall. But you can also walk down the hall. If walking down the hall enables better teamwork, it can mean a more productive team unit.

Team production comes with a problem, though. If Jon and Patrick can move a total of 50 boxes individually, but they move 70 when working as a team, who is responsible for the additional 20 boxes? In some sense, the team itself as a unit is responsible. The extra 20 boxes would not be moved without the team.

A similar analogy works in sports. How many touchdowns is a quarterback individually responsible for? You might say all throwing touchdowns belong to the quarterback, but do they?

In football, there are 11 players for each team on the field at any given time. How many touchdowns would a great quarterback be able to throw if you took just one of the eleven offensive players off the field? Probably zero.

At a professional level, defenses are just too good for the offense to succeed even if they’re missing just one player. It’s probably not right to say that quarterbacks are equally responsible for touchdowns as fullbacks, but it’s also wrong to say they are completely responsible. The team itself plays a role.

Returning to our boxes example, this creates a bit of an issue. If we know Jon can move 30, and Patrick can move 20, but the team itself can produce more than 50, how would we know whether Jon or Patrick decided to put in less effort?

In our fictional example, I specified the team could move a total of 70 boxes, but in the real world the maximum number of boxes a team can move is not known. If one of the two decided to not lift as hard or walk more slowly, maybe the total would drop to 60 boxes. Notice, 60 boxes is still more than the 50 that would have been done individually, but it’s less than the 70 possible.

In team production, it’s difficult to tell when someone is shirking (putting in less effort). Also the person who shirks does so at the expense of other members of the team. Whenever we don’t bear the cost of our decisions, we are more likely to make costly decisions.

Economists Armen Alchian and Harold Demsetz first introduced the problem, and it is now known as the team production problem.

So how is the team production problem solved? Let’s now say Jon and Patrick are movers for a moving company. The company wants to minimize shirking. What can they do? Well, the company could hire a manager—let’s call him Dan.

If the moving company wants to cut down the team production problem, they can offer Dan a financial reward for every box the team moves in excess of 50. Notice what this does. Dan now has a financial incentive to use all sorts of techniques and technologies to improve productivity and stop the team production problem.

He could try to motivate Jon and Patrick via good leadership. He could visually monitor them to see if he can catch someone slacking off. He could offer one or both financial incentives.

There is not one clear textbook answer to the team production problem, but paying individual managers to solve the problem on a situation-by-situation basis seems to be an effective solution.

So what does this have to do with remote work? My belief is that the ways managers often solve the team production problem require physical office spaces. For example, a shared company culture which motivates employees to put forth a little more effort may be easier to foster when you see your coworkers frequently and talk about your lives. Similarly, physical, visual monitoring may simply be a more motivating form of monitoring than others.

Many want it to be the case that remote work is more productive, but I chalk this up mostly to wishful thinking (though this is not to say that some people will not thrive in work-from-home situations, or that some jobs are not more conducive to working from home than others). Interestingly, the NBER study showed those workers who preferred working from home were actually even more unproductive than the average worker when working from home.

In any case, it appears the advantages of team production in physical space are still beyond the realm of being completely replaced. It’s possible that with technological changes, we may move to a place where remote work is common or universal, but the moral of the story is there is no guarantee. The extent to which it makes sense for people to do remote work varies depending on all sorts of technological, cultural, temporal, and job-specific factors. There is no clear end of history for the question of office vs. remote work.

Peter Jacobsen is a Writing Fellow at the Foundation for Economic Education.