Introduction: Evaluation for Crime Prevention

Evaluation in crime prevention is almost universally called for. It is widely required by funding bodies, and, as a result, it is almost as widely attempted. It is technically very tricky both in relation to specific initiatives (Ekblom, 1990) and in relation to the sorts of partnership approach that are now widely promoted (Rosenbaum, this volume). Critics are agreed that standards are generally very low (Sherman et al., 1997; Ekblom and Pease, 1995; HMIC, 2000; Scott, 2001). There are some fundamental debates about how evaluation should be construed and conducted (Pawson and Tilley, 1997). There are also diverse stakeholders in evaluation, and a range of uses to which evaluations can be put. The papers in this volume are concerned primarily with the design and conduct of evaluations that are intended to help improve crime prevention policy and practice. Other possible purposes — for example to justify programs, to account for public expenditure, to celebrate achievement, or to encourage greater participation in crime prevention — are not much at issue here. Much literature on crime prevention has stressed the importance of evaluation for program improvement. The preventive process described by Ekblom (1988), the action research approach advocated by Clarke (1997), and the SARA (scanning, analysis, response, assessment) steps developed by Eck and Spelman (1987) to help operationalise Goldstein's (1979) problem-oriented policing all include evaluation as a key element. The purpose of evaluation in all cases is to provide feedback that will generate corrections to and refinements in crime prevention theory, policy and practice. The University of Maryland report to the United States Congress on evaluations in crime prevention (Sherman et al., 1997) and the British Home Office