When telework really works
(Image: Tale / Shutterstock)
I recently read an important report in Harvard Business Review about an experiment conducted at a Chinese company called Ctrip, China's largest travel agency, on a topic that has been the subject of much debate and controversy in the U.S. government – telework.
In the nine-month experiment, which involved employees of the company's call center, staff was divided into two groups -- one that continued doing their regular job at the office, and another that did their regular work from home.
Importantly, in terms of the validity of the experimental results, the company did not do what is often done in such circumstances -- namely, solicit volunteers for telework and then compare their performance with those preferring to continue at the office. The problem with interpreting comparisons of volunteers with non-volunteers is there may well be differences between the type of person who opts in that itself influences performance.
Instead, what the company did was solicit volunteers, and then divide that group based on whether individuals had an odd- or even-numbered birthday. The comparison was thus between two groups who had both volunteered for telework.
The two groups continued to work the same shifts, with the same managers and same equipment, as before the experiment, allowing a before-and-after comparison for each group. The company also kept extensive, computerized records of the times employees were actually working, sales they made, and the quality of their interactions with customers.
The results were fascinating. During the nine months of the experiment, productivity of the teleworkers increased 13 percent compared with those employees' baseline, while the productivity of those staying in the office remained unchanged. Most of the productivity gain came from increased working hours, due to fewer breaks and sick days among teleworkers. Productivity per minute worked was also somewhat greater, however -- employees stated in a survey that the quiet of working at home made them more efficient.
Turnover during the nine-month experiment was also 50 percent less for teleworkers than for those in the office. (Telework, of course, also economizes on office space, one of the most important original reasons this company had undertaken the experiment.)
There was an unexpected result, however: After the experiment was done, half the teleworkers asked to return to the office, and three-quarters of those who had stayed in the office turned down an offer to begin to telework. Why? People complained they were lonely when they worked at home -- long live the office cooler (or its Chinese equivalent)! This suggests that perhaps telework should be a part-time thing (though there may be issues with duplicating office-assigned equipment some employees use).
These results should be carefully noted by federal personnel managers and policymakers. There are two reasons, but important ones, to worry that the results would not be transferable to a government environment.
First, this company had good quantitative performance measures for work input, output and results. If an agency does not have those, the productivity results might be negative in the absence of the ability of a boss to observe employee behavior directly.
Second, these jobs were performed by individuals in isolation. For work that must be done in teams, physical interaction of team members may well be better than a virtual team environment.
But the overwhelming news from this important study is that teleworking, under the right conditions, really works. I believe this study should ignite a movement to find more federal workplaces where it can be used. Federal personnel managers and policymakers, get going!
Posted by Steve Kelman on Feb 19, 2015 at 11:37 AM