As a work of political theory, When the State Meets the Street is unusual in that it combines normative reasoning with ethnographic fieldwork. Political theory is often criticized for operating at too high a level of abstraction, specifying ideals of justice or democracy that bear little relation to the political world in which we live. I believe that a turn to ethnography can serve to anchor a more grounded approach, situating philosophical reasoning within a more realistic understanding of our political institutions.Political theory should look to ethnography not just for additional context or meaning or texture, but also to uncover new normative questions that are worth asking.Seen from a distance, it may seem like all that happens when the state meets the street is the application of legal directives. Seen from up close, however, one starts noticing that the process is far more indeterminate and that the individuals involved have real discretion. Ethnography thickens the plot.With that, a new terrain for normative reflection opens up. What values should street-level bureaucrats be sensitive to when wielding their discretion? How can we justify that discretion and reconcile it with the tenets of democratic government? How can we hold them accountable?Street-level bureaucrats are the first to take the blame when things go wrong. We sometimes forget, however, that the constraints they operate under are largely of our own making. They reflect the amount of resources that we, as a democratic public, have chosen to devote to public services.Describing what happens when the state meets the street is like holding a mirror to ourselves as a polity. If we do not like what we see, we need to reconsider our own values and priorities.This provides us with a different starting point for reflecting on democratic politics. We usually think first about what laws and policies we should have, and only then about how to implement them. Once we know what we want the state to do, we can figure out how to do it.But what if we flipped things around—thinking first about how we would want the state and its officials to interact with citizens and working our way back, from there, to questions of management style, institutional design, and policy choice?