- COVID-19 and the resulting lockdowns caused rapid and extreme changes to the working week.
- Researchers analyzed the working patterns of 15 million users on GitHub, and compared patterns in 2020 to the pre-pandemic 2015–19 period.
- Results showed that the pandemic initially caused GitHub users to work between 15 and 20 percent more on average.
- In some nations like India, the jump was even higher — 30 to 40 percent.
The move to remote work initially increased the average GitHub user’s hourly workweek by more than 15 percent and shifted work to nontraditional hours.
The pandemic caused a large number of workers to shift from working at the office to working at home. The rise of remote work may outlast the pandemic, but it is not clear whether it is affecting when and how much employees work.
In Labor Reallocation and Remote Work during COVID-19: Real-Time Evidence from GitHub (NBER Working Paper 29598), researchers Grant R. McDermott and Benjamin Hansen analyzed the working patterns of 15 million users on GitHub, the world’s largest platform for developing software code. They compared patterns in 2020 and the pre-pandemic 2015–19 period.
GitHub allows individuals and companies to manage versions of software and code by meticulously tracking individual code contributions. Because it includes a timestamp for code changes, the researchers were able to track when and how long code contributors worked. They found a dramatic change in working hours as soon as the pandemic hit.
In the years before the pandemic, weekends accounted for about 20 percent of GitHub users’ working hours. At the onset of the shift to remote work, this share jumped to more than 24 percent, an increase of about two working hours each weekend. Later in 2020, weekend work patterns began to drift back to their pre-pandemic levels, although out-of-hours work (early morning and late evening) remained elevated in multiple locations.
In addition to changing working schedules, the pandemic initially caused GitHub users to work between 15 and 20 percent more on average, or about eight hours per week assuming a 40-hour initial workweek. In some nations like India, the jump was even higher — 30 to 40 percent. The overall increase in working hours didn’t last, however. It trended back to pre-pandemic levels in most nations by July 2020.
The researchers also focus on six urban hubs — London, New York, San Francisco, Beijing, Bengaluru, and Seattle. They found that the reallocation of working hours across these cities preceded the local lockdown orders in each jurisdiction by several weeks, suggesting spillover effects across networks and geographies. There was further variation across cities in the return to pre-pandemic work patterns. London returned to trend by summer, coinciding with the lifting of lockdown restrictions, while Bengaluru remained above trend throughout 2020.
Men shifted their work schedules more quickly than women, suggesting that men may have benefitted more from the shift to remote work. Women reallocated work more slowly at the beginning of the pandemic, a finding that is consistent with other research showing that women bore the brunt of housework and childcare when offices and schools closed.