Britain's annual budget ritual should ideally be banished or at least take a radically different approach

The best surprise that Gordon Brown, the chancellor, could spring on us on March 21 would be to announce that this is the last traditional British Budget - the last national sporting occasion on which the chancellor produces tax changes like a rabbit from a hat and when analysts stay up all night to understand what is really meant.

The present chancellor has already had his fun with these occasions. He could have even greater fun by bringing them to an end and saying that his real Budget, like that of almost all commercial organisations and many other governments, will be the public spending plans to be unveiled in the summer.

Of course, there could be occasional finance bills. These would be used when the chancellor wants to improve the tax structure, or needs to raise more revenue, or can afford to return some cash to the taxpayer. A year with no tax changes - which under the current arrangements would be dismissed as a "boring Budget" - would then count as a success story.

The virtues of stability, which Mr Brown constantly preaches, apply to the tax system as well. An essential preliminary to any such reform would be the indexing of all tax rates and starting points not merely to inflation, as many of them are now, but to the level of incomes. That would halt the insidious process of what economists call fiscal drag, by which the tax proportion increases even when tax rates are not increased.

If a really radical change of this kind were in prospect, I imagine I would have caught a whisper. So let me put forward a second-best. Mr Brown should take heed of last year's conclusions of the Commons Treasury committee, on which Labour has a majority, and present the Budget in a clearer, more concise and less propagandist way. For instance, one should not have to guess that tax remissions come under the heading "Increasing Employment Opportunity" (as if Michael Portillo were already chancellor).

Above all, the financial effects of several budgets should be shown separately and not counted several times over. And tax and expenditure figures should be distinguished and shown on a single annual basis. "New money" should be distinguished from existing commitments and Mr Brown should stop rolling up expenditure plans for three or four years together to give an exaggerated impression to Labour MPs, who are not as stupid as ministers suppose.

What else? Many economists would actually like the chancellor to increase tax rates in the hope of encouraging interest rates and sterling to fall. Indeed, the European Commission has been lecturing the Irish government not to go ahead with plans for promised tax cuts to reduce economic overheating - almost certainly to no avail.

Apart from their political unrealism, such pleas are based on economic assumptions that are often falsified by events. A more moderate view is that fiscal policy should support monetary policy to the extent of not attempting to remit most of the current Budget surplus. This is estimated at £18bn ($29bn) for 2000-01, on the basis of unchanged policies, by the Green Budget published by the Institute for Fiscal Studies in collaboration with Goldman Sachs.

We have little true idea how much of this is structural and how much represents a cyclical boom effect. What we do know is how quickly surpluses can change into large deficits at the first whiff of recession. The IFS suggests a package under which the chancellor would be able to remit £2bn per annum of taxes while increasing public spending and still retaining a surplus of more than 11/2 per cent of gross domestic product.

How should the tax changes be allocated? There is one place in the tax structure where disincentives bite earlier and earlier, namely the starting point for the 40 per cent higher rate of income tax (equivalent to a consumption tax of well over 50 per cent). Although the higher rate has been unchanged since 1988, it has gradually bitten lower down the income scale.

This has mainly been the result of Conservative chancellors trying to show their "one nation" credentials by not adjusting the upper threshold in line with rising incomes. It was a futile gesture that gained them no political credit. But as a result, there are now 2.3m higher rate taxpayers, or three times as many as in 1980-81. Those affected are just the middle managers, the professionals and small businessmen on whom governments rely to spread the enterprise culture.

Moreover, the IFS pinpoints a way of raising the higher rate threshold, initially by £2,000 from its present £32,335 level, thereby reducing the number of higher rate payers by more than 400,000. It would be best to combine this relief with an increase in the Upper Earnings Limit (UEL) for employee national insurance contributions to the higher rate starting point. Doing so would eliminate the dip in the tax progression between the end of the UEL and the start of the higher rate.

The net revenue cost of the combined moves would be just over £600m per annum. A side effect would be to reduce the gains from the reform of the upper two deciles of the income distribution, leaving the chancellor with a margin of £1bn-£2bn to remit in more Labour-friendly ways.

Finally, I have a suggestion to make for all those dinner party conversationalists who say that they would rather not have the 1p income tax reduction already announced for 2000-01, and devote the sums to better public services. They should be enabled to do so.

Why should not income tax returns have a space in which the taxpayer could indicate his desire to forego the 1p relief (or any other relief on a future occasion)? The return could also contain an estimate of what the effects would be on his personal tax bill. There is no need to stop here.

Why should there not be a structured ballot attached to the income tax form in which the public-spirited taxpayer could express his preferences between extra expenditure on different public services such as health or transport? Then the new expenditure could be divided among public services in line with taxpayer preferences.

So far from being "rightwing", a move along these lines would reduce the biggest disadvantage of public over private spending, which is that the voter has to take a complete spending package from the winning political party and has no way of making his own choices as he can in his own personal shopping.

This reform might be called in the language of the spin-doctors "An Opportunity to Put your Money where your Mouth is". It would need to be carefully prepared months or years in advance of the actual announcement. But there is nothing impossible in it except to the bureaucratic mind.