Using emoji reduces dropout of remote workers: a causal analysis on GitHub

Remote work relies heavily on online communication methods including asynchronous text-based messaging, which is often associated with delayed feedback and reduced emotional support. Inefficient communication may ultimately lead to negative impacts on work-related outcomes such as a higher turnover rate. We propose that emojis can help mitigate such disadvantages and present a causal analysis on GitHub—a setting where remote workers work voluntarily and rely primarily on textual communication. Our analysis demonstrates that, after adjusting for confounding bias using propensity score-based methods, including emojis in at least one post in 2018 reduces the dropout risk in the following year by half (an absolute risk reduction of 0.052, p ≪ 0.001), and higher emoji usage intensity further leads to lower dropout risk. While the absolute effect of emoji use is greater for new contributors, the relative reduction in dropout risk is larger for developers with a longer contribution history. The effect of using emojis is also more pronounced for developers from collectivist or restrained cultural backgrounds, who may be more sensitive to group identity and social norms. Our analysis provides valuable insights for designing interventions to enhance the experience and work-related outcomes of remote workers.