Governance and Development Review home page Centre for the Future State website
Most recent reviews on the site Complete list of reviews Search reviews by subject Contact the Governance and Development Review

< Go back


Varieties of Multilevel Governance


Decentralisation is the term we normally use to talk of the shift of (some) power from the central state toward multilevel governance. In recent years we have seen a very high level of interest in decentralisation, a great deal of actual decentralisation in many developing countries, and a large new literature on the subject. Specialists and practitioners, heavily engaged with the details of particular cases, may not always be able to stand back and detect the most interesting or important general patterns. Liesbet Hooghe and Gary Marks have made a very important addition to our intellectual armoury by suggesting that, amid all the diversity, there are two quite distinct and coherent approaches to decentralisation, based on different concepts of governance. They label these two approaches Type I and Type II respectively.

The defining characteristics of Type I are (a) the sub-central jurisdictions are multipurpose; (b) the memberships of these sub-central jurisdictions do not overlap; (c) there are a fixed number of levels of sub-central jurisdictions; and (d) the whole system follows one uniform design. The simplest such systems might involve, under a national government, a set of provincial authorities each with broad ranging, identical powers and responsibilities; and clear non-overlapping responsibilities for a given territory and population. In most large countries there will be at least one other level of territorial public authority, which relates to the provincial level in the same way that the provincial level relates to the national level. These are the systems we tend to think of as 'normal'.

By contrast, Type II systems are defined by (a) task-specific jurisdictions; (b) overlapping memberships; (c) an unlimited number of jurisdictional levels; and flexible design. More 'traditional examples' would include many empires, and the United Kingdom itself, in which the Channel Islands, England, Northern Ireland, Scotland and Wales each have a different relationship to the Crown. More relevant contemporary cases are found in many large urban areas, in which responsibility for different functions and services - schooling, police, road transport, rail transport, bus services, hospitals, planning, electricity, gas, water etc. - are divided among many different bodies, each organised in their own way, with different territorial boundaries, different numbers of sub-levels etc.

To tidy minds, Type II systems at first sound messy, and very much a second best. But Hooghe and Marks show that the evidence does not support such prejudice. Each type of system can work well in some contexts. Hooghe and Marks tell us something about their relative strengths and weaknesses, and the circumstances to which each might be most appropriate.


Many social science concepts are useful mainly to help social scientists explore theoretical issues and debate with one another. But some concepts can have great practical value: not because they directly tell us anything new, but because they identify a pattern that might otherwise have been missed, and help and stimulate us to think in original ways about familiar problems. I anticipate that this modestly labelled concept of Type I and Type II governance will be in the latter category.

Source: Hooghe, L. and G. Marks, 'Unravelling the Central State, but How? Types of Multi-level Governance',American Political Science Review, Vol. 97 No. 2, (2003), pp. 233-43. American Political Science Review is published by the American Political Science Association.

Keywords: decentralisation, institutions.

Commentator: Mick Moore, IDS, (July 2003)

< Go back

      Top | Home | Latest reviews | Archive | Subject search | Contact | CFS site