The limits of performance assessments of public bodies: external constraints in English local government

Policy makers have an interest in evaluating what government and taxpayers get for their money. In the United Kingdom, central government has developed elaborate oversight mechanisms involving a plethora of indicators, targets, and inspections to assess public sector performance across a wide range of services. The underlying assumption is that auditees are ultimately responsible for the performance on which they are measured. This rationale might neglect the fact that they face external constraints not influenced easily by their decisions. The research presented here scrutinises one such evaluation approach, the Comprehensive Performance Assessment (CPA) of local government in England, for the significance of such factors. Using panel-data analysis involving the Index of Multiple Deprivation 2004 and data from the first three CPA rounds in 2002 – 04, the research finds that deprivation in education, crime, and living environment had a significant negative effect on local authorities’ overall CPA scores, while discretionary expenditure had a positive but not significant effect. However, the effects differ widely across the five types of authorities that exist. The model also reveals that there is no statistically significant evidence to suggest that authorities controlled by a single party perform better than those without, and that, although Conservative councils have higher CPA scores in simple bivariate analysis than those controlled by other parties or with no overall control, these differences disappear in multivariate analysis.