Our Purpose

Prevention has an evidence problem, but not the one most people assume. The research exists. Decades of work across public health, criminology, developmental science, and implementation studies have established that many forms of harm can be prevented. We know more than we have ever known about what drives violence, exploitation, poor health outcomes, and system failures. We understand the critical periods in human development when intervention has greatest effect. We have robust evidence on what works.

The problem is not knowledge. It is translation.

Evidence remains locked in academic literature, inaccessible to the practitioners and policymakers who need it. Policy is developed without realistic plans for implementation. Commissioning cycles reward short-term outputs over long-term prevention. Agencies work in silos while the problems they address cut across every boundary. Interventions are designed for average populations rather than the specific communities they serve. And too often, the people most affected by harm have no voice in designing responses to it.

Prevention Works exists to address this translation failure. We believe that evidence should reach the people who can use it. That strategy without implementation is empty. That prevention requires collaboration across systems because no single agency owns the conditions that produce harm. And that the goal is not simply better programmes but better systems: institutions that learn, adapt, and intervene before damage is done.

This is practical work. It involves training, consultancy, evaluation, knowledge exchange, and the slow labour of building networks and partnerships. It requires honesty about complexity and humility about what we do not yet understand. But it starts from a conviction that prevention is possible, that it is more effective and more economical than responding after harm occurs, and that the barriers to achieving it are organisational and political, not scientific.

Our purpose is to remove those barriers, one partnership at a time.