ABSTRACT
The emergency management chaos that occurred in the wake of Hurricane Andrew resulted in a host of relatively standard public administration recommendations (NAPA 1993), including improved planning systems, better coordination, clearer "trigger" mechanisms for federal and state action, and specification ofmore measurable objectives for emergency managers. Even if they do all these things, however, in an actual event political leaders and emergency managers still must be able to cope with many unpleasant surprises; for example, the inability to obtain the information that ex ante plans require. If unable to improvise and react well to a crisis as it unfolds, they can make the crisis more intense by inducing confused decisions or counterproductive behavior. In fact, our analysis of crisis decision making supports a growing body of research that emphasizes the "intergovernmental paradox of emergency management." Local government is that level of government least likely to perceive emergency management as a key priority and to pay sustained attention to it, but it is handed more and more responsibility for handling disasters (Cigler 1984; Schneider 1993; Schneider 1992; Wolensky and Wolensky 1990). In the case of Andrew, all federal, state, and city plans called for Metropolitan Dade County to be the premier agent of information and collective action during the initial crisis phase, which we define as lasting six days - the two days before the storm, the day of the storm, and the three aftermath days before the military was called in.
