Prior to 2020, there were approximately 7 million individuals in the U.S. working remotely, which represented 3.4% of the population. While this doesn't seem like a significant proportion of the population, what is impressive is that this percentage of employees working remotely has tripled over the last several years. According to
Gallup's State of the American Workplace survey, 43% of Americans work remotely at least part of the time. Working from home (WFH), known by a number of different terms, including telecommuting, telework, and remote work, is becoming mainstream!
Of course, the COVID-19 pandemic changed things even more dramatically! Many offices and workplaces closed overnight and asked their employees to work from home, beginning in March, 2020. At the height of the pandemic,
more than half of American workers spent the majority of their work-day remotely at home.
There is a growing debate on whether U.S. workers should return to their place of employment and when. It's a great question, especially now that the COVID-19 numbers are finally on the decline (at least in the United States). The general impression is that U.S. workers have been just as productive working remotely from home as they had been working on site before the pandemic. However, we don't make major decisions such as this one based only upon general impressions! What does the data actually show?
Most of the studies conducted to date on this issue suffer from some fairly important methodologic flaws. A number of these studies have been surveys asking whether managers believe that employees working remotely are productive or not. For example, a
large survey of 1,579 Japanese firms in several different industries suggested that the average productivity of employees who work from home is 69% of those who work in the office setting.
I want to focus on two relatively new and important studies on the impact of the implementation of so-called "working from home" (WFH) policies on employee satisfaction and productivity. The first study (
"Does working from home work?") came out a few years prior to the pandemic, but it is the only (as far as I can tell) randomized, controlled study of WFH. Here workers at a Shanghai call center were asked if they wanted to shift to working from home four days per week, with the fifth day spent in the office. Approximately half of the employees (503) expressed interest in WFH, and of these, 249 were qualified to take part in the study (they had been working at the company for at least six months and had the appropriate Internet connections, equipment, and working space at home). The qualified workers were subsequently randomized (by lottery) to either the WFH or control group. The experiment lasted for 9 months.
At baseline, the two groups (WFH and control) were very similar in most respects. Compared to the employees who didn't volunteer for the study, the workers in the WFH and control groups tended to have longer commutes, less tenure in the firm, and less education. Productivity between the two groups was not different (WFH versus control, and volunteers versus workers who elected not to participate in the study). The only significant difference between the WFH and control groups was that the workers in the control group were more likely to have children (this may be an important difference - see next study!).
Notably, performance (as measured by productivity) improved by 13% in the WFH group, mostly attributed to a 9% increase in the number of minutes worked per shift, with a smaller percentage increase due to an increased number of calls per minute worked. A follow-up survey revealed that the WFH workers attributed this improved productivity to greater convenience - for example, they could get something to eat or drink very easily without necessarily dropping work. Quality was not sacrificed for quantity - quality was measured by at least two valid metrics and did not change during the experiment. Attrition decreased by 50% in the WFH group compared to the control group, and workers in the WFH reported higher satisfaction. Most importantly, across the company productivity in aggregate improved by between 20-30%, and the company saved about $2,000 per employee per year due to the reductions in required office space and reduced turnover.
Not all of the news was positive though. Interestingly, the rate of promotions (which were dependent primarily on performance) decreased by almost 50%. At the end of the experiment, the company elected to continue the WFH program and offered its employees the choice of either remaining at home or at work. Two-thirds of the workers in the control group (who had initially volunteered to work remotely at the beginning of the experiment) decided to stay in the office. Half of the workers in the WFH group changed their mind and returned to the office. When asked why they chose their minds, workers in both groups cited concerns about being lonely at home and missing out on the connections they shared with their fellow workers. In addition, workers in the WFH group felt that they were "out of site, out of mind" (which possibly explains the decrease in promotions).
The second study (
"Work from Home & Productivity: Evidence from Personnel & Analytics Data on IT Professionals") is an observational study in which productivity was measured in a large Asian IT services company following the abrupt shift from working from the office (WFO) to WFH in March, 2020 during the beginning of the worldwide COVID-19 pandemic. While this study was not a randomized, controlled study such as the one above, the investigators used a variety of methods and analytics to accurately gather and collect an extensive data set consisting of the total number of hours, how that time working was spent, and how productive the employees were during those hours. The study's results are based on data collected over a 17 month period (including both pre- and post-pandemic) on over 10,000 employees.
Employees significantly increased the total number of hours worked after WFH was implemented in March, 2020, and the bulk of this increase occurred outside of normal work hours (e.g. after 5 PM for a "typical" 9-5 PM work day). Compared to pre-pandemic, work output did not change. In other words, employees were working longer hours for the same level of work output, i.e. productivity decreased. Measured in this way, productivity decreased by 20% compared to pre-pandemic levels.
Okay, you could look at this with a glass half full or half empty. Half full first - perhaps individual employees were constructing their work day by taking personal time throughout the day and then working later in the normal work day to accomplish the same level of work output. Given that these were all salaried employees (which is different from the first study), the most important issue is that the normal work was still completed, right? Taking a half empty perspective, productivity decreased and we should worry that the employees were maybe working harder and longer for the same level of work output. It's hard to say which is the correct interpretation here, so maybe we should look at some of the other results.
Looking at how each employee spent their day provides some additional answers. During WFH, employees spent more time in formal and informal virtual meetings. Spending more time in meetings means that they had less time available to generate work output - hence the longer work day. They also spent less time networking (with colleagues and clients) and significantly less time in 1:1 meetings with supervisors. Remember the WFH employees' concerns about being "out of sight, out of mind" in the first study!
The study revealed some additional insights. Women were more negatively impacted by WFH than men (note that this has been observed in many, many other studies on the impact of the pandemic on the workforce). However, this was not simply related to having children in the home (the study investigators speculated that this was due to the gender-specific norms in that country that placed greater demands on women at home). Men too suffered a lower productivity when children were in the household (recall that in the first study showing greater productivity with WFH employees, the control group was more likely to have children). Employees with children at home worked significantly more hours than those that did not have children at home, so even when work output stayed the same, productivity decreased.
These two studies provide some important preliminary data that (1) working from home may increase feelings of being separate from the rest of the workforce (see increased feelings of loneliness in the first study, as well as decreased networking and 1:1 mentorship and coaching in the second one) and (2) the decrease in productivity (at least in the second study - remember that productivity improved in the first one) may require longer working hours with their own attendant consequences on work-life balance, satisfaction, etc.
There is no question that the pandemic has permanently changed our world. The shift to WFH that occurred almost overnight in March, 2020 has changed not only how and where we will work, but even also where we will live in the future (see
"The doughnut effect of COVID-19 on cities"). To what extent these changes will impact the greater society at large remains to be seen, and for now, only time will tell.