New public management (NPM) arguments on strategy and structure suggest that performance is enhanced if large organizations are disaggregated into smaller units. the NPM perspective reflects the views of public choice theorists who claim that big organizations are unresponsive to public needs, inefficient and fail to achieve their formal goals. These arguments have underpinned many recent changes in the structure of public services at both central and local levels.
This paper uses data on six local government services to test the NPM hypothesis that there is a negative relationship between scale and performance. Five dimensions of performance are analysed: service coverage, quality, speed of provision, efficiency, and administrative effectiveness. Scale is measured through indicators of service output, caseload and needs. the impact of scale is tested in multivariate statistical models which control for other potential influences on variations in performance across local authorities.
Only around half of the statistical evidence suggests that the smallest local units are the best performers. In addition, even when performance does decline with scale, this trend is reversed in the very largest units. Thus, contrary to NPM arguments, the biggest organizations are seldom the poorest performers.