Disaster Warnings and Compliance: The Impact of Social Process Factors over Time

Abstract

Disaster warnings have proved to be only partially effective in bringing about behaviors that enhance survival. For disasters that are ongoing and associated with repeated warnings, determinants of compliance may change over time. To test this argument, a theoretical working model is developed and tested on the basis of data collected from a field survey of a national representative sample of Israeli households (814). Respondents were asked if they complied with the warning to enter a sealed room (i.e., shelter in place) at the initial and toward the end of a three month long ballistic missile attack on Israeli civilians during the First Gulf War. Compliance decreased with time. We then sought to generate a parsimonious model to explain the drop in compliance by focusing on a variety of factors that included “message” oriented and social process variables. Results of the logistic regression models demonstrated that compliance initially was determined by prior levels of preparedness whereas determinants of behavioral compliance three months later broadened to include gender, risk perceptions, and the source of knowledge for actions to the threat. These results strongly suggest that initial compliance with warnings depend on having available fundamental survival assets. With repeated warnings over time, social process variables significantly affected compliance decisions. Efforts to increase disaster survival should therefore consider the time continuum in terms of the impact of social process variables on complying with warnings.