In this book, Nancy Cartwright, Eileen Munro and John Pemberton introduce a new method for assessing whether plans for how to affect change produced their intended outcome, or whether they are likely to do so in the future. The method offers the prospect of a step-change improvement in the accuracy of policy assessments, based on a new pluralistic theory of causation. This theory, which goes beyond existing ones, synthesises seven tried and tested familiar component accounts so as to license identification and systematisation of a wide range of evidence types. The authors outline well-grounded improvements to methods for policy development and assessment by the systematic use of real-world examples, including notably that of child welfare. Their book will be valuable for the burgeoning audience concerned with the critical issue of how to develop and implement policies that work across domains from welfare to education and economics to medicine.