The natural response is simply to change the system. A new piece of anti- avoidance legislation here, a new tax relief to encourage the thing we really want more of there. Lobby groups will call for the changes and it seems so easy. After all, tax systems are sets of rules, so when we change them, individuals and companies have to respond.
The rules do have to be obeyed, but that doesn't mean we achieve the outcome we want. If we give a tax subsidy for corporate research and development (R&D) spending, we will see more R&D spending in company accounts, but much of it may simply be a redefinition of activities that are happening anyway. If we seek to encourage investment by giving new tax reliefs for one way of financing investment (say, by raising equity) we will see a shift in financing from debt to equity rather than necessarily any more total investment. If we seek to encourage investment by a temporary increase in investment allowances (as in the early 1990s), we will see investment that was going to happen anyway packed into the period of temporary allowances, with low investment either side.
Once all the relabelling, refinancing, and rescheduling has occurred, and much of the cost has also gone to those who were already doing whatever is to be encouraged, there may be some genuine change in real behaviour which meets the true objectives of the reform.
None of this should be taken to imply that the tax system now is perfect and shouldn't be changed. What we have now is unfair, inefficient and massively complex and distortionary. But responding by piecemeal change without an absolutely clear underlying strategy risks simply making things worse. And before any change is made it is vital to consider all the ways in which it won't work, will be abused, diverted and distorted. Because, however good the intentions of policymakers, many taxpayers will seek to find the best ways of legitimately minimising their tax liabilities.
We can see these problems spectacularly in the field of the taxation of savings. Early in their period of office, the last Conservative government felt that more funding for risky small companies would be desirable, so they introduced a tax relief called the business expansion scheme (BES). The BES eventually grew rapidly, but ways were found of using the scheme to invest in highly non-risky assets such as fine wine and property. By the close of the scheme, universities were using BES to finance student accommodation. The BES was eventually abolished.
Schemes aimed at the other end of the savings market have had similarly chequered experience. Personal equity plans (PEPs) were introduced in 1987 aimed at encouraging direct equity investment. Very little happened until the government relented and allowed individuals to hold unit trusts in PEPs. Tax-exempt special savings accounts (Tessas) were introduced with the aim of encouraging those on lower incomes to save. Being on a low income will typically mean that you can't afford to save at all, let alone put money away for the five-year minimum holding of a Tessa. Little wonder that only 3 per cent of the poorest 20 per cent of the population has a Tessa, and that most of the money that went into Tessas seemed simply to come from pre-existing saving rather than be new in any way.
And these schemes can appear to be successful as Tessas did, although probably producing little new saving. The best example was perhaps personal pensions. When these were introduced, the government thought that an extra tax incentive would be necessary to encourage people, although still expecting a maximum of half a million takers in the first year. Five million individuals took out personal pensions in the first year, taking the large incentive payment on offer, and only a third saved any money of their own. The cost to the government was some pounds 9bn, a quarter of the annual NHS budget.
What does the taxation of saving look like now? We have the enterprise investment scheme, occupational pensions and personal pensions, PEPs, Tessas, National Savings, direct holding of shares, and interest-bearing accounts. All these have different tax treatments, and are listed in rough order of generosity of tax treatment.
And yet saving is above all something where we would not want the tax system to be determining choices. We save because we want to be able to consume in the future. So we care little about the vehicle in which we save, except about the rate of return we expect, the risk and how accessible it is. Different tax treatments will have an enormous impact on how we save. But we would want the choice of savings vehicles to be determined by the underlying investments, not the tax regime. So, rather than the variation we see in tax treatment, we would be better off with a uniform tax structure.
Perhaps surprisingly, we have moved some way towards that in recent years, but could move further. If the tax-free lump sum in pension taxation were removed, and the remainder of Miras taken away, these two assets, which form the great bulk of personal wealth and saving, would be taxed much as Tessas and PEPs are now. The large remaining distortion would then be the continued taxation of interest income.
Taxing interest income is economically inefficient when we do not tax the return to saving in pensions, housing, or equity held in PEPs. It also penalises those on low incomes, who if they save at all, tend to save in interest-bearing assets. So the Individual Savings Account (ISA) fits well into a coherent strategy for taxing saving, and is generally welcome. Tessas were not attractive to those on low incomes because of the five-year lock in, and ISAs should avoid that.
But the initial proposal for ISAs sought to pay for tax deductibility of interest income by imposing limits on the lifetime level of contribution to ISAs and what could be brought over from Tessas and PEPs. Aside from administrative and compliance problems (which could be substantial), this falls into the trap of not taking seriously enough how individuals would respond. The idea that those with more than pounds 50,000 in a PEP would leave the excess hanging around waiting for the taxman when they had the option of enterprise investment schemes, venture capital trusts, additional voluntarily contributions to pension schemes, or buying a larger house, to say nothing of more adventurous tax planning, is almost funny.
If the ISA proposal is amended coherently during the consultation period, the final scheme will be welcome. Even then, ISAs are unlikely to increase aggregate savings much, or to encourage many on low incomes to save for the first time. But they would have the merit of fitting into a sensible aim, equal taxation of all form of savings. It is these general aims we need to be clear about, rather than frenzied production of impressive sounding schemes that add complexity and distortion while achieving little. The test for the March 17 Budget is how coherent it will seem in five years, not how impressive it sounds on the evening of Budget Day.
Andrew Dilnot is the director of the Institute for Fiscal Studies.