This Article interrogates a critical, yet understudied, regulatory design choice the Supreme Court makes in each criminal case raising constitutional questions about law-enforcement conduct: not what the Constitution requires but how to implement its requirements. In particular, the Court must decide whether to address its decision directly to rank-and-file officers or instead to political policymakers, such as legislators and police administrators, who in turn will regulate officers on the street. In the former, dominant model, termed here first-order regulation, the Court tells officers precisely what they can and cannot do. In the latter model, second-order regulation, the principal objective instead is to enunciate constitutional values and create incentives for political policymakers to write the conduct rules. Framed differently, the Court, as principal, enlists political policymakers as its agents in the regulatory enterprise. Although first-order regulation predominates, a careful search uncovers bits of second-order regulation in spaces such as inventory searches and interrogation, and analogies in fields like employment and desegregation. The Article claims that second-order regulation should, in some domains, benefit suspects and criminal defendants in the aggregate by increasing the expected value of their constitutional protections. It should meanwhile facilitate efficient prosecution of the guilty. The benefits of agency, in other words, should in some cases outweigh the costs. Shifting rulemaking responsibility from the Court to political leaders harnesses certain comparative advantages of political institutions and permits experimentation in search of workable, well-tailored safeguards to protect constitutional rights. Even more important, social science research suggests that affording law-enforcement greater opportunity to participate in its own regulation encourages “buy in” that leads to improved compliance. The agency costs, in contrast, including “slippage” in the form of potentially underprotective rules, can often be reduced to tolerable levels. After making the general case for the second-order approach, the Article sketches where it should work especially well or poorly. It then reimagines several of the Court’s first-order decisions in a second-order model. Finally, it suggests a role in second-order regulation for other potential catalyst institutions, such as legislatures and state courts.