86 Finance & economics The Economist October 30th 2021
Thepyjamarevolution
I
n february2020 Americans on average spent 5% of their work
ing hours at home. By May, as lockdowns spread, the share had
soared to 60%—a trend that was mirrored in other countries. Ma
ny people, perhaps believing that working from home really
meant shirking from home, assumed that office life would soon
return to something like its prepandemic norm. To say it has not
turned out that way would be a huge understatement.
Most office workers remain steadfastly “remotefirst”, spend
ing most of their paid time out of the office. Even though a large
share of people have little choice but to physically go to work, 40%
of all American working hours are still now spent at home. In mid
October American offices were just over a third full, suggest data
from Kastle Systems, a security firm. From Turin to Tokyo, com
mercial areas of cities remain substantially quieter, compared
with precovid norms, than residential ones. Economists are try
ing to work out what all this means for productivity.
Perceptions about the future of office work are changing. Last
year British government ministers exhorted workers to get back to
the office; now they are quieter. Wall Street banks, often the most
enthusiastic advocates for inoffice work, are toning down the
rhetoric. According to a monthly survey by Jose Maria Barrero,
Nick Bloom and Steven Davis, three economists, bosses expect
that in a postpandemic world an average of 1.3 days a week will be
worked from home—a quarter more than they expected when
asked the same question in January. Even that could in time prove
to be an underestimate. Workers hope they will spend closer to
half their working hours at the kitchen table.
A few factors explain why remotefirst work remains domi
nant. Many people remain scared of contracting covid19, and
thus wish to avoid public spaces. Another possibility is that work
ers have more bargaining power. In a world of labour shortages, it
takes a brave boss to make people take a sweaty commute five days
a week (workers view being forced to be in the office fulltime as
equivalent to a 5% pay cut). There is a more intriguing possibility,
however. Work that is largely done remotely may be more efficient
compared with an officefirst model.
The past year has seen an explosion of research on the econom
ics of working from home. Not all the papers find a positive impactonproductivity.A recentpaper by Michael Gibbs of the University
of Chicago and colleagues studies an Asian itservices company.
When the firm shifted to remote work last year average hours rose
but output fell slightly. The authors ascribe part of the decline in
productivity to “higher communication and coordination costs”.
For instance, managers who had once popped their head round
someone’s door may have found it harder to convey precisely what
they needed when everyone was working remotely.
Most studies, however, find more positive results. Mr Barrero
and his colleagues’ surveys cover a large number of firms, rather
than just one. Only 15% of homeworkers believe they are less effi
cient working in this way than they were on business premises be
fore the pandemic, according to a paper published by the team in
April. A study released that month by Statistics Canada finds that
more than half of “new” remote workers (ie, those who normally
worked outside the home before the pandemic) reported complet
ing about the same amount of work per hour as before, while one
third said they got more done.
Economists have less insight into why remote workers might
be more productive. One possibility is that they can more easily
focus on tasks than in an office, where the temptation to gossip
with a coworker looms large. Commuting, moreover, is tiring.
Another factor relates to technology. Remote workers, by necessi
ty, rely more on tools such as Slack and Microsoft Teams. This may
allow bosses to coordinate teams more effectively, if the alterna
tive in the office was wordofmouth instructions that could easily
be forgotten or misinterpreted. Patent applications for work
fromhome technologies are soaring, while American privatesec
tor investment in it is growing by 14% yearonyear.
Yet the popularity of remotefirst work presents a puzzle. If it is
so wonderful, then why is there little evidence of a shift towards
“fully remote” work, where firms shut down their offices altogeth
er? Companies that have chosen to do this are in a tiny minority.
The number of people moving to cities such as Tulsa, in Oklaho
ma, which is positioning itself as the global capital of remote
work, remains small.
Perhaps it is only a matter of time before everyone who can
goes fully remote. A new study in Nature Human Behaviour, how
ever,suggests that firms have good reason to hold on to their office
buildings, even if they are used less frequently. The paper studies
the communications (including instant messages and video calls)
of 60,000 Microsoft employees in 201920. Remote work makes
people’s collaboration practices more “static and siloed”, it finds.
People interact more with their closest contacts, but less with the
more marginal members of their networks who can offer them
new perspectives and ideas. That probably hurts innovation. The
upshot is that fully remote teams might do quite well in the short
term, but will ultimately suffer as innovation dries up. What a way to make a living
How best to collaborate, then, in a remotefirst world? Many firms
assume it is enough for everyone to come into the office a few days
a week, since this will lead to people bumping into each other and
talking about ideas. Others, backed by stronger evidence, say that
managers must be more intentional, and bring people together
with the express purpose of discussing new ideas. Firms will have
to experiment as they get used to a new way of working, and the
precise arrangement may vary dependingonthetype of work.
What seems clear, though, is that offices willstillhave a role after
the pandemic—even if they are mostly empty.nFree exchange
Remote-first work is taking over the rich world. A growing body of research hints at why