At the height of the local election campaign in 1991, Chris Patten, the then Tory party chairman, confessed himself 'gobsmacked' that Labour had costed out, in detail, a cheaper alternative to the 'community charge'.
Within a week the Government had been forced to publish hastily compiled 'exemplifications' of what council tax levels might have been in each district in England, Wales and Scotland if it had been in force in 1991/92.
These figures - based on a 'desk survey' of properties by district valuers at October 1990 prices - had originally been intended for internal consumption by ministers and civil servants only.
Now, with the full valuation exercise virtually complete, the Government has again been wrong-footed by its failure to publish an update of its provisional estimates of 1991.
This is despite the fact that Britain's 459 district and islands councils have all been individually notified, as required by law, of the results of the valuation exercise in their own areas.
In the absence of official information, the Labour-led Local Government Information Unit (LGIU) scored an early propaganda coup last week with its prediction that the average council tax in England would be some pounds 60 higher than previously estimated.
This was based on a sample survey of 13 authorities, which suggested - wrongly, as it turned out - that the value of properties for council tax purposes was some 11 per cent less than the Government had estimated last year.
The Government had at that time - to the surprise of the building societies - put the average house price in England at pounds 80,000, in Wales at pounds 60,000, and in Scotland at pounds 53,000.
The LGIU prediction appeared to confirm press reports last month that the valuation had put many more homes in the lower-value council tax bands than previously expected.
What these early reports overlooked - and what the Government at first did little more than hint at - was that the valuation exercise had also identified significantly more properties than previously anticipated.
Preliminary drafts of the valuation tapes were already in the hands of the local authority associations, and by the end of last week the Association of London Authorities published the first print-out.
That print-out formed the basis of Labour's widely reported press briefing last Monday, from which a new myth emerged - that individual authorities' council tax bills would rise or fall next year as a direct result of charges to their tax base.
What all the reports ignored is that the only valuation figure that will have any direct effect on next year's council tax levels is the national average change from the previous estimate.
This is because the Government's revenue support grant system compensates individual authorities pound-for- pound for any local shortfall in their taxable resources.
So it doesn't actually matter that Hackney's tax base is down by 22.4 per cent on previous estimates, Swansea's by 18 per cent, or that the Scilly Isles' is up by 29.5 per cent. The only way a relatively low tax base can affect local taxation is by increasing the gearing effect when a council's spending exceeds or undercuts its government-set standard. These facts are now being stressed by the Department of the Environment - which dismissed the press reports as 'absolute nonsense' - and were confirmed by several sources familiar with the intricacies of the grant mechanism.
'Having a low tax base is not bad news,' said one finance expert. 'It won't make any difference to the average tax bill,' said another.
What the national picture shows, the DoE finally disclosed this week, is that the total number of domestic properties in England is 19,949,803 - a rise of just more than one million on the previous estimates.
But this 5.4 per cent rise in property numbers is more than offset by a 7.1 per cent fall in their council tax valuation, with more properties falling into lower tax bands. The net effect is to reduce English local authorities' overall tax base by about 2.1 per cent.
This slight shrinkage would have added pounds 8.65 (or 2.2 per cent) - not pounds 60, nor pounds 80 - to the Government's 1991/92 target council tax of pounds 400 for a two-person household in a Band D property, worth between pounds 68,001 and pounds 88,000 at 1 April 1991. The proportions in each band are:
Band A (up to pounds 40,090) 26.3 per cent;
Band B (to pounds 52,000) 19.1 per cent;
Band C (to pounds 68,000) 21.8 per cent;
Band D to pounds 88,000) 14.5 per cent;
Band E (to pounds 120,000) 9 per cent;
Band F (to pounds 160,000) 4.9 per cent;
Band G (to pounds 320,000) 3.7 per cent;
Band H (over pounds 320,000) 0.6 per cent.
In Wales, where the same eight bands relate to lower values, the total number of properties is up by 3.6 per cent to 1,200,516. But valuations are down by about 5 per cent on previous estimates, giving an overall reduction of between 1 and 2 per cent.
In Scotland, property numbers are up by 3.1 per cent to 2,165,900. But there is as yet insufficient published information on bandings to calculate the overall change.
Councils are currently going through the final phase of the valuation exercise before the 23,315,000 households in Britain are officially notified of their council tax bands in early December, prior to the setting and issuing of bills next March.
This involves a check on the Valuation Office figures and an estimate - largely based on poll tax registers - of the likely numbers of exemptions and of 25- per-cent discounts for single-adult households. The Government had originally estimated that 33 per cent of households would receive discounts, reducing the total tax base by one-twelfth. So here again, any changes in the national - but not the local - picture would affect the target tax.
So what, if anything, does the recently completed valuation exercise tell us about likely average council tax levels for next April? The short answer is, very little.
The Government's notional target charges for a two-person Band D property - pounds 400 in England, pounds 270 in Scotland, and about pounds 120 in Wales - related to the 1991/92 tax year and are already out of date. The equivalent, unpublished, figures for the current year would probably be about pounds 460 in England, pounds 306 in Scotland, and pounds 160 in Wales. And ministers will be setting new targets when they announce the 1993/94 revenue support grant settlement later this year.
But the target tax - officially called the Council Tax for Standard Spending - is simply the amount the Government thinks councils should charge if they spend in line with Whitehall's guidelines.
This year, despite a tough capping regime, English councils exceeded the guidelines by a total of pounds 313m, equivalent to an extra pounds 19 per Band D property or pounds 8.87 per head on the poll tax. In Wales and Scotland the excess per head/property was even greater. Even more crucially, the target council tax makes virtually no allowance for non- payment. Last year's pounds 400 figure for England assumed a non-collection rate of only 0.2 per cent.
This year the average English council has allowed about 8 per cent for such 'other adjustments', and last week's government figures on poll tax collection levels simply serve to confirm the accuracy of this estimate.
Councils are likely to exercise similar caution for the first year of the council tax, potentially adding a further pounds 36 per Band D property.
The third major flaw in the pounds 400 target is that it assumed that some pounds 1bn of transitional relief for households facing large increases in poll tax bills would be recycled as revenue support grants, reducing the headline council tax.
In practice, the Government will need at least this sum - equivalent to nearly pounds 60 per Band D property in England - to protect individual losers from the change-over to council tax.
So headline council taxes - if not the average amounts actually payable - will need to be that much higher.
All these factors, coupled with the reduction in the valuation base, point to an average headline council tax in England of about pounds 585 for a two-person Band D property at 1992/93 prices. This compares with an average English poll tax of pounds 562 per couple.
Factors pointing in the other direction include the Government's determination - voiced by Michael Portillo, now Chief Secretary to the Treasury, last year - not to let high bills 'blow the council tax out of the water'.
In the English shires, counties and districts will be keen to keep bills low as they vie for public support in the current review of local government structure. The county councils have the added pressure of four-yearly elections next May.
Another factor is the possibility under the new system that high gearing will allow some shire districts, such as Rochester, actually to set a negative council tax that would at least partly offset the county precept.
A fourth factor - the direct result of the recent valuation - is that because more properties are in the lower tax bands, the average amounts payable in England and Wales will relate not to Band D but to Band C, where bills are reduced by 11 per cent. In Scotland the average could be as low as Band B, with a 22-per-cent reduction.
The irony is that if Chris Patten had not been gobsmacked in April 1991, Michael Howard, the Secretary of State for the Environment, would not now be under such pressure to deliver a pounds 400 council tax in April 1993.