The name said it all. Closed worlds like those of education or railways tend to adopt their own special bureaucratic language. So church schools become “voluntary aided”; others are “controlled”.

It still amazes me that, for the first radical type of school in a hundred years, Kenneth Baker chose such a dismal name. Not “charter schools”, giving them a history and a sense of purpose. Not “self-governing schools”, a mouthful but at least an accurate mouthful. His choice was “grant-maintained school” - leaden bureau-speak that only served to emphasise the schools’ dependence on grants while obscuring their source.

Were GM schools a success? Of course. They by-passed the LEA and the debilitating, wasteful cycle of committees, councillors and bureaucrats, enabling headteachers and governors to take real responsibility. Most schools that went GM flourished, with a stronger ethos, more motivated teachers and better results.

There were also important side-effects. Forced to re-evaluate their services, LEAs became more cost-effective and school-focused in everything from maintenance to payroll.

But was the policy a success? Here the answer has to be no. If GM schools were right, why weren’t there thousands? Overall, the policy failed, and for three reasons, two of them self-inflicted.

First, there was overwhelming LEA hostility, even from Conservative-run councils. Headteachers were threatened, governors bullied, parents intimidated. We had to strengthen the law just to ensure fair ballots. Some LEAs were completely unscrupulous in defending their monopoly.

Second, the Conservative government gave less than whole-hearted support. From the start the policy seemed clouded in ambiguity. Was GM status simply an option - as the Education Reform Act itself suggested - or was it ministers’ preferred status? Did we want every school to go GM or enough to be a catalyst in every area?

There were other ambiguities. The attraction was supposed to be additional recurrent expenditure. So why did we proffer such extravagant capital funding? In the early years the GM sector’s capital allocations were hard to justify.

And those ballots. If we wanted all schools to go GM in the end, what was the point of all that balloting? To make them vote every year until they got it right?

Third, there was local management of schools. At precisely the same time as we were promoting the advantages of voting for total self-government, we introduced another policy that gave schools 85 per cent self-government. They could have almost all of the budget and still stay with the LEA family.

Wised-up authorities cottoned on fast, delegating just enough to their schools to keep them from balloting. As the early capital carrots disappeared, the advantages of going GM became much more marginal.

This combination of outright LEA hostility and lack of single-minded political commitment by the government proved fatal. We got GM schools all right but far too few of them - and almost all in the wrong places.

In the shires, suburbs and the South, headteachers with strong governing bodies went for the full advantage. But in the cities and the North there were only a few scattered beacons too feeble to light up the surrounding darkness.

We should have been bolder. The great missed opportunity was the l992 general election. That was the point at which a fresh Conservative government could have justifiably declared the experimental phase over. We should simply have deemed all secondary schools GM, and left the authorities to sort out their primaries.

But we funked it. We worried about the implications for local government; we were wary of setting up regional offices of the Department for Education; we fussed over uniform funding formulae across the LEAs and for five long years self-government for schools withered on the vine.

The Labour government hasn’t wasted any time with ballots. The new categories of foundation and community schools (is there a difference?) were simply put to governors. Parents who specifically voted for GM status have not been allowed to vote on discarding it.

Has Labour learnt from GM? Their visceral opposition to the l988 Act soon evaporated. By the 1997 general election they had even come round to promising greater delegation. If they can turn this policy into reality, then GM schools will not have died in vain.

Good schools shouldn’t be run by councillors or by ministers. Self-governance, in an open, accountable system where parents have real choice, has to be the way forward. Next time we’ll do it properly - and we’ll choose a better name.

Briefing 22-26 Michael Fallon was a Minister at the Department for Education from 1990 until 1992. He is now Tory MP for Sevenoaks, Kent.