Homicide statistics are a widely accepted metric of security and democracy. This article argues for a focus on how bodies come or do not come to be counted – of what happens before states enumerate. The experience of São Paulo relates that how many people die and how many do not is connected to the governance of an organised crime group known as the PCC. The punishment practices of the PCC and groups like it throughout Latin America reshape the lived paradigm of governance over life and death, albeit in concealed ways. Statistics are produced by and are productive of a de jure state, different from the state de facto. The acceptance of state-made homicide figures, whether for analysis, visualisation or political claims, is consequential for the future of lived security and social science knowledge production.
