The Effect of Lockdown Repeal on Socialization: Bayesian Multilevel Difference-in-Differences Approach

Abstract

The COVID-19 lockdown has had an unprecedented impact on people in various ways. This study evaluates the effect of lockdown repeal from both marketing and public-policy perspectives. Combining the Bayesian multilevel model with the difference-in-differences design, we find that a lockdown repeal has had a negative impact on socialization. Furthermore, the results show that those who have a low level of risk perception are less affected by lockdown repeal. Also, the negative effect of lockdown repeal varies depending on past socialization behaviors; that is, the lockdownrepeal effect is attenuated for those who socialized more than others in the past. Our findings contribute to the intersection of public policy and marketing literature and provide both academic and practical implications.