How is policy made when goals are not widely shared, attentionattention is fleeting, choicechoice is biased by framing effects, and collective benefits differ from individualindividual interestsinterest? Under these conditions, the utility of synoptic rationalityrationality is limited. I analyze two alternatives, bounded rationality and multiple stress/garbage cans, to argue that process affects policy content, optimal solutions are utopian, and greater ambiguity and time constraints likely lead to a politicized, disjointed process.