Work Interruptions and Creativity: Evidence from a Natural Experiment

Abstract

We studied and estimated the causal effects of work interruptions owing to unexpected events on the subsequent creative production of ideas in firms. To do so, we exploited a natural experiment – unexpected stops in production plants that led to interruption of work for some workers but not for others. In our main event, this interruption led to four paid days off work. We collected data on weekly submissions of workers to an idea management system, building a balanced worker week sample. We used coarsened exact matching (CEM) to build a matched sample (n = 3,921) and a difference-in-difference design to investigate whether individuals who have been exposed to an interruption are producing more ideas than others. We found that an interruption leads to 46% more (and 3% more valuable) ideas in the three weeks following it. This significant increase is consistent across many robustness checks, alternative estimation methods, and a replication in a second natural experiment. We put slack time leading to freed cognitive capacity forward as possible explanation for our effect. We also introduced boundary conditions to our findings: First, the positive effect of interruptions is more pronounced for less creative innovators (+14%). Second, interruptions have a negative effect (-22%) on creativity if they are not associated with slack time, and individuals shift their attention and cognitive capacity somewhere else as a consequence of the disruption (as in the case of a strongly adverse event).