Introduction: Safety behaviors are key elements in reducing the spread of the COVID-19 virus, but have also assumed excessive proportions in form of panic buying groceries. This raises the question whether these behaviors are independent or related to each other. Adherent safety behavior including increased hygiene and physical distancing appears inherently adherent and prosocial, while dysfunctional safety behavior such as panic buying most probably emerges from other motives and contextual variables.

Methods: Data from 15,308 participants collected from March 10 to May 4, 2020, during the COVID-19 acute period in Germany, was analyzed to assess whether adherent and dysfunctional safety behavior are predicted by the same or divergent variables. Two multiple regression models are presented including various sociodemographic, trait, attitudinal, and COVID-19-specific variables as predictors.

Results: Some variables similarly predict both, adherent and dysfunctional safety behavior. Yet, adherent safety behavior is stronger predicted by COVID-19-related fear than generalized anxiety, while a trend toward a reverse pattern emerged for dysfunctional safety behavior. Adherent safety behavior was also related to higher trust in governmental actions to face COVID-19, subjective level of information, as well as use of public media and TV to remain informed on COVID-19. Higher age was related to dysfunctional, but not adherent safety behavior. Respondents living in rural communities report more adherent safety behavior than urban dwellers.

Discussion: Divergent psychological variables underlie adherent and dysfunctional safety behavior. This hints toward a theoretical separation with practical relevance in behavioral engineering and public health campaigning.