Nudge-based messages have been employed in various countries to encourage voluntary contact-avoidance and infection-prevention behaviors to control the spread of COVID-19. People have been repeatedly exposed to such messages; however, whether the messages keep exerting a significant impact over time remains unclear. From April to August 2020, we conducted a four-wave online survey experiment to examine how five types of nudge-based messages influence Japanese people's self-reported preventive behaviors. In particular, we investigate how their behaviors are affected by repeated displays over time. The analysis with 4241 participants finds that only a gain-framed altruistic message, emphasizing their behavioral adherence would protect the lives of people close to them, reduces their frequency of going out and contacting others. We do not find similar behavioral changes in messages that contain an altruistic element but emphasize it in a loss-frame or describe their behavioral adherence as protecting both one's own and others' lives. Furthermore, the behavioral change effect of the gain-framed altruistic message disappears in the third and fourth waves, although its impact of reinforcing intentions remains. This message has even an adverse effect of worsening the compliance level of infection-prevention behaviors for the subgroup who went out less frequently before the experiment. The study's results imply that when using nudge-based messages as a countermeasure for COVID-19, policymakers and practitioners need to carefully scrutinize the message elements and wording and examine to whom and how the messages should be delivered while considering their potential adverse and side effects.
Keywords: Altruism; Behavioral economics; COVID-19; Infectious diseases; Nudge; Panel survey experiment.
© The Author(s) 2021.