Since the summer of 2014, Americans have seen more videos of violent interactions between police and non-Whites than ever before. While the interpretation of some specific incidents remains contentious and data on police use of force are scant, there is evidence that racial disparities in policing exist even when considering racial disparities in crime. The traditional civil rights model of institutional reform assumes that racial bigotry is the primary cause of these disparities; it attempts to address problems through adversarial litigation, protest, and education. This article offers an expansion of that model—one based on insights from behavioral science—that facilitates a less adversarial approach to reform and allows one to be agnostic about the role of racial bigotry. The new behavioral insights model focuses on identifying the contexts—called identity traps—that can escalate negative interactions between police and communities, as well as ways to interrupt them.