Jurisdiction Size and Local Government Effectiveness: Assessing the Effects of Municipal Amalgamations on Municipal Performance

Across the developed world, the last 50 years have seen a wave of local government mergers implemented in order to improve the performance of local governments. However, two methodological challenges make the estimation of merger effects difficult. First, endogeneity is a serious concern since mergers are rarely distributed randomly among local governments (selection bias) and since jurisdiction size may have an impact on performance, but performance is also likely to have an impact on the decision on whether to merge (reverse causality). Second, there is a lack of empirical indicators of service quality or performance. This means that most empirical studies focus on spending per capita as the indicator of scale effects. Conceptually most studies thus study economies of scale, while leaving the important question of effectiveness of scale unexamined. A number of recent studies of local government mergers rely on quasi-experiments (e.g. Allers and Geertsema, 2016; Blom-Hansen et al. 2016). While these studies address the endogeneity issue more effectively than previous studies, they fail to address the problem of empirical indicators. They focus on the effects of mergers on expenditures per capita and are thus limited to conclude on economies of scale. The important question of effectiveness of scale is left unexamined. In fact, we have no solid knowledge of the effects of local government mergers on the quality of local services. The contribution of this paper is to provide quasi-experimental evidence on this question. The research setting is the Danish local government reform in 2007. The reform was directed by the central government and constituted an exogenous shock to 239 municipalities, while 32 municipalities were left untouched. We thus have a quasi-experiment with pre- and post-treatment observations for both an experiment and control group. The paper focuses on public schools, which is one of the most important policy areas of the municipalities. Former studies have found no economies of scale in the school area, i.e., spending on schools did not drop in the amalgamated municipalities relative to non-amalgamated municipalities. But did mergers lead to increased effects of school expenditures? To address this question the papers focuses on students’ academic performance at the final examinations in lower secondary schools with appropriate controls for socioeconomic confounders. Potential scale effects in student performance are not likely to turn up instantaneously after merging municipalities, but should be identifiable 6-7 years after the merger. To test scale effectiveness we utilize the quasi-experiment of the Danish 2007 Local Government Reform in a DiD design based on a dataset with pre and post observations for student performance in each municipality combined with register data on school expenditures, the number of students, school sizes, number of teaching lessons, the number of students choosing private schools etc. as well as indicators of each municipality’s local fiscal environment.