As the Government struggles over its white paper on regional government, it is important to dispose of some commonly-held assertions about regional government.
It is frequently said regional government would resolve the constitutional problems posed by devolution to Scotland and Wales. In particular, it would smooth out the difficulties caused by the MSPs' ability to determine legislation covering England and Wales.
Patrick le Gales, joint author of Regions in Europe, suggests that instead of a Europe of the regions there is 'a Europe with certain regions'. He illustrates, with some notable exceptions, 'regions as a level of government have generally remained relatively weak. They have weak autonomy, weak resources, weak political capacity and weak legitimacy'.
Advocates of regional government rarely specify what elected regional authorities will actually do. They assume regional government is a good thing, but without indicating the extent of their remit. There is little point in having regional elections unless the assembly elected commands attention and power.
There is already concern about the levels of turnout in local, European and national elections; why would regional elections secure a higher turnout?
Even with all the publicity surrounding London's 2000 mayoral election, the turnout was slightly lower than the preceding 1998 borough elections. In Wales the turnout was about the same as in the previous local elections.
That such regional elections cannot produce strong turnouts does not augur well for elected regional bodies elsewhere, especially where regional identity is far more uncertain.
Unless regional authorities are given high-profile powers and functions, the election process will attract little attention. The minimum required would be the equivalent authority commanded by the Welsh Assembly, but again no one is suggesting this.
It is more usual to argue in favour of regions' strategic role, but it is not clear what this means or how any strategies would be implemented. It has been suggested the regions take over the functions of the government offices for regions or regional quangos, but this could prove problematic.
On the whole, government offices for the regions do not have responsibility for services, but supervise or allocate resources - such as housing investment programmes - to councils or other bodies. To have a substantial impact, regional bodies would have to take over central government's responsibilities for quangos, as well as the quangos themselves.
It is argued regional bodies would not reduce the powers of local government. This is difficult to believe as the creation of regional assemblies would be accompanied by the extension of unitary councils. If that means the abolition of counties, then most of their functions would go up to the regions rather than down to the smaller unitary councils.
It is often suggested regions should be based on areas covered by the government offices of the regions. This would result in regions double the size of those in most other European countries and ignores economic, social and political realities. For example, Cornwall would hardly welcome being governed from Bristol. Merseyside and Manchester have their own distinctive identities, but would be represented by one regional authority. The south-east and the East Midlands have hardly any identifiable regional identity, but both have strong sub-regional distinctions within them.
Our greatest concern is with the proposal that regional government be introduced area by area as voted for in referendums. But the best justification for regional government is the need for decentralisation, and that would involve fundamental change in the organisation of central government, including the number of government ministers. It is fantasy to believe there would be such a change if just two or three elected regional assemblies were established initially.
Central government must give regions serious thought before it puts forward any proposals. The danger for local government is that - whatever is said by ministers at the outset - the consequences of regionalisation will not be decentralisation downwards but regionalisation upwards.
George Jones is professor of government at the London School of Economics and John Stewart is professor of local government at Inlogov, University of Birmingham