Introduction

Our interest in evaluation has been arrived at almost by accident. Perhaps it is the same for everyone. Driving our particular focus on this issue have been our recent experiences as local evaluators for a series of young offender projects. Reflecting upon those experiences has brought us to a recognition of the close (and mutually reinforcing) relationships between the government's strategy for young offenders and the process of evaluating the resulting ‘interventions’ within which we were implicated as evaluators. As we argue in this chapter, there is a clear and direct connection between the new and quasi-scientific language of ‘programme evaluation’ operating within a strict and supposedly ‘evidencebased’ discourse of ‘what works’ and the new ‘actuarial–interventionist’ logic of contemporary youth justice (Feeley and Simon, 1992, 1994). Here, we attempt to develop a critique of both, although never losing sight of their intimate connection.

Setting a context

Before developing our critical commentary upon the evaluation process that accompanied the rolling out of the Youth Justice Strategy, following the 1998 Crime and Disorder Act, it is important to describe something of this strategy and the central role of evaluation within it. The government was clearly committed to an ‘evidence-based’ approach to tackling youth crime, as elaborated by the Audit Commission Report of two years earlier (Audit Commission, 1996). In the same vein, the rigorous cataloguing of outputs and rapid dissemination of ‘what works’ findings regarding demonstrable ‘good practice’ which might be replicated in new settings were very much part of the overall strategy (Goldblatt and Lewis, 1998; Home Office, 1998; Hope, 2002). Accordingly, when the newly established Youth Offending Teams (YOTs) were presented with an opportunity to submit special project bids to the Youth Justice Board (YJB), an evaluation component (of around 5% of each project funding bid) had to be identified.