Wednesday, March 25, 2020
The coronavirus outbreak of 2020 has precipitated a shock where millions of workers around the globe have been forced to work from home under conditions of social isolation. In this paper, we exploit the coronavirus shock in China to estimate differences in content contributions to online communities by workers who work from home (WFH) relative to workers who cannot work from home (CWFH). Our pre-period comprises the time frame of the Chinese New Year holidays when most workers were on vacation in their hometowns. In the post-period, all workers were under mandatory lockdown facing conditions of social isolation. However, while workers in certain industries such as the internet industry, software industry, etc. were asked to WFH, other workers in industries such as airlines, computer hardware, and manufacturing could not work from home. We estimate a difference-in-differences specification comparing open content contributions by WFH workers to contributions by CWFH workers in the post-period. Using data from a popular question-and-answer platform, we find that WFH workers posted 19% fewer answers than CWFH workers. Moreover, relative to CWFH workers, WFH workers answered less popular questions. We also provide suggestive evidence (using natural language processing tools) for a plausible mechanism: psychic costs related to working from home without adequate supporting arrangements. Finally, we present correlational findings regarding how online contributions are higher under conditions of both WFH and CWFH compared to physical work in the office in a prior period. Our findings contribute to literatures on working from home and online content contributions by workers and have managerial implications for managing WFH under conditions of social isolation.