British singer George Michael
British singer George Michael performs on stage. Michael was one of the 1,600 people identified as customers for Liberty - a tax avoidance scheme. Reuters

We now have 18,000 pages of tax law in Tolley's tax handbooks but it seems even our prodigious quantities of tax law may not have been sufficient to prevent 1,600 wealthy people avoiding a tax liability on a total estimated income of £1.2bn.

These people attempted to realise artificial tax losses by investing in the Liberty partnerships. We will not whether they succeeded before March next year, when the tax scheme gets to the courts.

This is the latest in a long line of publicised cases. All too often it is demonstrated that the hedge fund manager who said "my cleaner pays more tax than I do" is far from being alone as a wealthy non-taxpayer.

Of course the government has not just sat on its collective backside while all this has been going on – even though there is a question mark about delays in closing down the Liberty scheme itself.

We now have the naming and shaming of tax cheats, codes of practice for banks, a general anti-avoidance rule, and every year the proud claim by the Chancellor in his Budget that he is saving £x million with new anti-avoidance measures.

But just as a leaky inner tube can get so bad that more patches cannot prevent the air escaping somewhere else, and a new tube is required, so we need to stand back and look at our tax system. We need to decide if we should take a fresh approach.

There is indeed an alternative approach – which is that our tax law should aim to identify the economic consequences of what people have done, and tax them accordingly.

The Liberty investors had no economic loss because the partnerships had equal income and expenses – but their plan was to make expenses tax deductible even though the income was tax- free.

At the moment we have a fragmented and broken tax system. As a structure it does not have sound foundations, and yet we wonder why the upper floors are buckling. If we had a solid base we could have clearer tax law, and it would be easier to identify and tackle unacceptable avoidance.

Our highly complicated anti-avoidance law is not currently solving our problems. Indeed such law can be so obscure that it is difficult to discern its purpose, and it can actually encourage tax advisers to look for ways round it. And if they do get round the detailed rules no one knows for sure whether it is then appropriate to invoke the general anti-avoidance rule.

There are some amusing examples of how complexity gets the tax- man's knickers in a twist. Section 179 of The Taxation of Chargeable Gains Act 1992 applies where a company leaves a group having acquired an asset from another company in the same group.

I would encourage anyone who has more than a passing interest, and access to the law, to take a look at how this section reads now, after all the amendments made in recent years. For practical purposes the section is a complete mess.

In 2000 some amendments were made, but some of these were wrong, and had to be corrected by a revised paragraph 47 to Schedule 30 of the Finance Act 2000 inserted by section 79 Finance Act 2001, which was deemed to have had effect from 2000.

The original errors were not spotted by anyone before the Finance Act 2000 was enacted, because of the overriding complexity. Other examples can be given of this sort of thing.

Of course we also have the problem of too many reliefs and allowances. But analysis shows that our problems go deeper than this.

We should start by asking the basic question – what are we trying to tax? If our objective is to tax real profits and give tax relief for real losses no-one would now dream of starting with our current mish mash of rules to achieve that objective.

Is reform pie in the sky? I don't think so.

But if we are to make progress Treasury ministers, their officials, and the tax profession, need to engage with the problem of avoidance at a more fundamental level than has happened to date.

David Martin is a research fellow at the Centre for Policy Studies (CPS).