ABSTRACT
In the aftermath of a nuclear attack, shelters can offer potentially important protection. How well they fill that role depends on a set of interdependent decisions made by the individuals and organizations that must prepare and use them. We look at three such decisions. For each, we begin with formal analysis of the consequences expected from different possible actions. Those analyses are, then, reviewed in terms of how individuals facing these choices will perceive them, given the information that they are likely to have. The first example suggests that preparing a home shelter according to guidelines from the Department of Homeland Security may not pass a cost-benefit test. The second example explores the use of readily available information about a blast to infer how urgently shelter should be sought. The third example considers when shelters should be left, suggesting that individuals with the best shelters and slowest evacuation speeds should evacuate last, if they have the provisions needed to remain. In each case, helping people to protect themselves requires prior risk analyses and communication development.
