Harm is a recurring theme in the social sciences. Scholars in a range of empirical areas have documented the deleterious outcomes that at times emerge from social structures, institutions and systems of governance. Yet these harms have often been presented under the rubric of ‘unintended consequences’. The outcomes of systems are designed to appear devoid of intentionality, in motion without any clear agency involved, and are thus particularly adept at evading accountability structures and forms of responsibility. Drawing insights from decades of social theory – as well as three illustrative examples from Australia's health, welfare and immigration systems – this article argues that many social structures are in fact intended to cause harm, but designed not to appear so. In presenting this argument, we offer a theoretical framework for conceptualising harm as actively administered. We also challenge scholars from across the social sciences to reconsider the partially depoliticising narrative of ‘unintended consequences’, and to be bolder in naming the intended harms that permeate social life, often serving powerful political and economic interests.