In the mid-1990s New York City inaugurated its era of mass misdemeanors by pioneering policing tactics featuring intensive enforcement against low-level offenses as part of its quality-of-life and urban crime control strategy. These tactics have since spread across the country and around the globe. But the New York City experiment embarrasses our traditional understanding of how an expansion of criminal enforcement should work: as misdemeanor arrests climbed dramatically as part of an intentional law enforcement strategy, the rate of criminal conviction fell sharply. Using extensive, original data from a multiyear study, this Article exposes an underappreciated model of criminal administration in New York City’s processing of mass misdemeanors, one that makes sense of this trend. Misdemeanor justice in New York City has largely abandoned what I call the adjudicative model of criminal law administration—concerned with adjudicating specific cases—and instead operates under what I call the managerial model—concerned with managing people through engagement with the criminal justice system over time. The adjudicative model holds that courts stand between the proscriptions of substantive criminal law and the hard treatment of punishment by employing the criminal process to select the right people for punishment and to determine the proper amount. The managerial model does not depend on punishing individual instances of lawbreaking, but rather on using the criminal process to sort and regulate the populations targeted with these policing tactics over time. These findings challenge the “assembly-line” account so often associated with lower criminal courts, showing that misdemeanor courts engage in a tremendous amount of discretionary differentiation in the treatment of the people who flow through their operations. However, the basis of this differentiation is not what we would expect from the traditional adjudicative model of criminal law, namely guilt and blameworthiness.