Putting a Box around Context : when do we have the whole story ?

Pawson and Tilley’s Realistic Evaluation was published twenty years ago. It is timely to reflect on its origins and future directions. The basic ideas for Realistic Evaluation were first aired publicly at a criminology conference in San Francisco in 1991. There followed a series of papers in which the ideas were refined. Once Realistic Evaluation appeared in 1997 we wrote a few more papers together before moving to separate research agendas that build on the work: Pawson (mainly on methodology) and Tilley (mainly on practice and policy). Against expectations the ideas set out Realistic Evaluation, which appeared radical and subversive when first mooted, have become fairly mainstream with a welcome community of practitioners now working on them. The main examples used in Realistic Evaluation and in our early papers related to crime and crime prevention (Tilley’s main interest). The take-up has until recently been in other fields, most notably health (Pawson’s main substantive focus in recent years). The first part of this address will be retrospective. It will describe the genesis and development of realist evaluation. It will stress the dual concerns with methodology and with the production of findings that can be used to inform policy, practice and program development. It will highlight the serendipity involved in the production of the work, whilst also acknowledging that it appeared at a time that was ripe for fresh thinking. The second part of the address will look forward to issues that those involved in developing realist evaluation methods, in conducting realist evaluations (and reviews), and in developing and delivering policing and practices that are informed by realist thinking need to be thinking about. It will also highlight some obstacles that will need to be overcome if the potential benefits of realist evaluation are to be realized.