There is widespread agreement that when the public is paying attention to an issue—that is, when an issue is salient—elected officials are more inclined to involve themselves in agency decision-making, limiting administrative discretion and centralizing policymaking authority. I argue, however, that the prospect of issue salience in the United States should be a stronger predictor of agency design and statutory delegation of administrative discretion. Using data on significant congressional enactments and federal agency design in the latter half of the 20^th Century, the analysis yields results that are consistent with the notion that issue salience is a long-term political risk from which political actors wish to insulate policy. In particular, issue areas that the public is most likely to consider “the most important problem” are linked to limited statutory delegations of policymaking discretion and the establishment of administrative agencies with insulating designs. If one accounts for this long-term measure of salience, then issue salience at the time a statute is enacted or an agency is established yields little or no predictive power. This study improves existing theory, identifies a problem with a significant body of research, and establishes long-term issue salience as a general predictor of U.S. statutory delegation and agency design, improving substantially on the explanatory power of the most highly touted predictors, such as legislative conflict and inter-branch partisan division.