Not only are local authorities very big, but they have also been heavily bureaucratised. While other European countries expanded their welfare states using social welfare agencies, charities and housing associations, the British solution was to blow up the existing local government bureaucracies like balloons. This created absurdly giganticist town halls in the late Seventies, when Glasgow ran 170,000 council houses and Birmingham 120,000 homes from remote and centralised housing departments.
Historically, British town-hall bureaucracies have also been heavily professionalised and departmentalised, their boundary disputes creating impenetrable barriers for local citizens. Each department tends to answer to a committee of councillors set up on functional lines (education, housing, social services, highways), so that bureaucratic divisions have been reflected in political terms as well. In theory, responsibility is partitioned across multiple departments: in practice, no one is running the overall show.
Decentralisation efforts in the UK and Europe have been designed to bring down the scale of local government organisation, to make it accessible, to involve local residents more in decision-making, and above all to combat the fragmentation and obfuscation of local government bureaucracy and committees.
Changes in information technology are among the many things favouring decentralisation. Twenty years ago, to get a computerised rent collection system you had to stack up housing bureaucrats in atower-block office with a mainframe computer in the basement. Now officers with networked PCs can work on each local housing estate. Equally radical changes have taken place in budgeting, control systems and the management of service quality - along the lines of private-sector firms like McDonald's.
So decentralisation is far from a "loony left" idea. Liberal Democrat and "new Labour" councils are both keen on getting local government closer to the people. And the Audit Commission stresses the need for more management decentralisation.
In Sweden, the authorities who were most unhappy with the service from council staffs, and hence keenest on contracting out, are the ones who have adopted radical decentralisation. One pattern there (similar to that planned for Walsall) is for the local council to appoint people to a large number of budget-holding neighbourhood councils. It may spend its money with the local-authority service, or assign it to private contractors. Decentralisation is a key to making the purchaser/provider distinction work.
Elected councillors retain ultimate control of policy and budgets, but local residentshave a say in running their neighbourhoods. Yet if things go badly wrong at neighbourhood level, the council can put them right. This averts the problems encountered in Liberal Democrat Tower Hamlets in the early Nineties, where neighbourhoods were controlled by a very few local councillors - thus vulnerable to capture by the British National Party.
The British public likes the idea of decentralisation. In March 1994, a Channel 4 pollconducted by ICM and the London School of Economics' Public Policy Group asked whether neighbourhood councils were a good or a bad idea. People responded by a huge margin in favour, with 81 per cent saying it was a good idea and 66 per cent saying they would participate or make use of them. There was strong support also for local referendums, phone-in lines and more easily accessible information.
All of this means the decentralisation issue will not go away - whatever happens to Walsall's latest initiative.
The author is Professor of Government at the London School of Economics and chairs the LSE Public Policy Group.