Federalism, Decentralisation and Performance
Researcher Prof. Dr. André Kaiser, Diplom-Verwaltungswissenschaftler Niels Ehlert
Project Description Recent economic and social sciences literature has not sufficiently taken into account the fact that political systems may vary with regard to the organisation of power along two dimensions. According to William Rikerís definition (1975), federal systems are primarily characterised by a constitutionally guaranteed division of authority between central state and member states. Yet it is always to do with establishing institutions for the separation of powers. By centralisation, one understands the extent to which the implementation of political programmes is controlled by the central government. Decentralisation respectively means the degree of implementation is controlled by other political institutions. The jurisdiction may be either functional or geographically organised. From this, the project poses the following research questions:
  1. How do national states behave with regard to the unitarism-federalism dimension and the centralisation-decentralisation dimension? What connections exist between the two dimensions?
  2. What difference do the two dimensions make concerning state activity? How do they affect the performance of the political system?
The project assumes the two following working hypotheses that completely go against the majority of federalism-theoretical literature: decentralisation, and not federalism, brings decided performance advantages (H 1). Within a unitaristic framework, decentralised policy implementation is tendentiously more efficient than under the condition of territorial separation of powers (H 2).

While the theses generally emphasise the advantages of federalism and decentralisation over centralised political control, it isnít theoretically clear which of the two above-mentioned dimensions is actually responsible for the claimed performance advantages. Therefore, the project will record a broad spectrum of indicators to measure the success of political programmes. The statistical analysis may at best provide us with good clues about which institutional arrangements are most likely to lead to good and bad performances. However, the causal mechanisms for this cannot be identified by this alone. We also require, firstly, a sound theoretical derivation and, secondly, an empirical test in detailed case studies. The research project should therefore connect statistical analysis and detailed case studies along the lines of the "nested approach" (Lieberman, 2002).
Funding The project will receive knock-on financing from the Centre for Empirical Economic and Social Research while a proposal for third party funding is being drawn up.
Data Database of indicators of political performance for OECD countries
  • Country-comparative case studies
  • Regression Analysis
Publications link