Post by Ifrah Khanyaree
What's the science?
The COVID-19 pandemic has accelerated digital transformation across many industries and organizations. Within a matter of weeks of the onset of the pandemic, many office-based working adults shifted to working remotely full time. This week in Nature Human Behaviour, Yang and colleagues analyzed communication and working hours data from a large US tech company to find out the impact of remote work on employee collaboration and communication.
How did they do it?
The authors used anonymized email, instant message (IM), calendar, video/audio call, working hour data of 61,182 US Microsoft employees from December 2019 - June 2020, collected using Microsoft’s Workplace Analytics product. The authors then analyzed this data using a modified version of the traditional Difference-in-Difference model (DiD), which is a technique used in econometrics that measures causal effect between at least two sets of longitudinal data, where one group receives a ‘treatment’ and the other does not (the control group). This works because many of Microsoft’s employees were remote even before the pandemic hit; that group acts as the control group that also experiences the effects of working during COVID, but not the treatment (switching to remote work).
They used a modified version of DiD both because COVID affected both the treatment and control groups and because their model measured the effects of changes in two different treatment variables instead of one - an employee’s remote work status and also their colleague’s remote work status.
What did they find?
The authors found that the shift to remote work for all employees caused the communication network to become siloed: a decrease in cross-group communication but an increase in the connectedness of one’s own group. Remote work led to a substantial increase in unscheduled calls, emails, and instant messages, but a decrease in meeting hours, and total video/audio call hours. Synchronous collaboration, where more complex information can be conveyed, such as video calls, was decreased overall in favour of asynchronous communication, like emails or messages. Further, the total hours worked were increased. These changes were particularly enhanced for managers. Finally, connections between employees became more static, with fewer social connections being added or lost over time.
