Public Policy Evaluation in a Centralized State

Abstract

One feature of heavily centralized government with nationally uniform policies is the limited scope for policy experimentation. Policies cannot readily be trialled and there is little opportunity to learn from small scale implementation. Increasing centralization in Britain has been accompanied by a developing managerial ethos within central government. HM Treasury has demanded that new policies be given explicit objectives and an evaluation strategy be implemented. The inability to use randomized controls has led to creative evaluation strategies that stress the understanding of process rather than the measurement of outcome. With the desire to tighten further Government's grip on public expenditure, the demand is now for pilot schemes and impact analysis. Focusing on social assistance, the paper details the changing environment of policy evaluation in Britain, reviews non experimental methods that have been used and documents the first two policy pilots. The paper concludes with a discussion of the opportunities and dangers arising from the move towards experimentation.