Performance Gaps, Peer Effects, and Comparative Behaviour: Empirical Evidence from Swedish Local Government

Abstract

Abstract Centralized inspections face scepticism among local public managers, and voluntary evaluations have become a popular complement. This study uses the Swedish local government benchmarking programme (Kommunens Kvalitet i Korthet) to investigate what correlates with partaking in a benchmarking exercise empirically. This study finds evidence for temporal and spatial clustering: participants cluster geographically at an early stage of the exercise, new entrants tended to attach to existing clusters of participants rather than forming new clusters themselves. From event history analyses this study also finds that the proportion of participants among direct neighbours increases the conditional probability of entering the exercise. This suggests that public managers and local councils mimic the behaviour from nearby councils when it comes to the use of performance evaluations.