Is Complexity in Taxation Inevitable?

J A Kay

Website: johnkay.com

IFS Working Paper 57, February 1985

The Deloitte Haskings & Sells Lecture, given to the Department of Accounting

at the University of Cardiff on 26th February 1985.

Taxation in Britain today raises about 135 billion pounds - between 35% and 40% of national income. Table 1 shows how this money is raised. The largest revenue raiser is income tax, which generates almost 30% of the total, and the two taxes on income, income tax and national insurance, together account for almost half. The other major contributions are value added tax, rates and north sea oil taxes, in aggregate accounting for 100 billion of the total. Petrol tax, corporation tax, duties on alcohol and on tobacco are all relatively modest revenue raisers - around 5 billion each.

My subject today is not where it comes from - the distributional effect of taxation - or even how it affects the economy as it does - the economic effects of taxation. I am concerned with how it comes. Many of you will be familiar with Butterworths Tax handbooks, which present in consolidated form the statutes that determine how we are taxed. You will know that they run to - literally - thousands of pages. Why is it all so complicated? Is that complexity inevitable?

There is a conventional wisdom that the reason tax systems are so complicated is because of constant political pressure from vested interests. If only we didn't run an open political system, it would be so much easier to run a decent tax system. Its an argument which is widely heard, and it's a very convenient argument. It's especially convenient for tax administrators. If only people would leave us alone, we could make a much better job of things, they can say in a world weary way. But, of course, people won't leave us alone. Complexity is all the fault of the politicians, with their many and contradictory demands. And there is no point in changing anything, because however attractive reforms might sound, the same pressures and the same politics would come and mess things up again.

And this conventional wisdom suits politicians, as well. It suits politicians because it puts the blame squarely on previous politicians which is where politicians always like to put it. It's all the fault of the earlier lot that the tax system is so complex. Vote for us, and it will all be different. Of course it never is, but that's the nature of politics.

Now elements of this - which I shall call the vested interest thesis - are all very familiar. Let me strike a controversial note right at the beginning by talking about VAT on books. We've heard a great deal on this subject in the last few months. But the truth is that very few arguments have been presented which couldn't equally well be applied to a whole range of other commodities. Isn't it important to have an educated and literate community? Of course it is. But it's also important to have a community that is properly fed, and clothed; housed, and furnished; that has access to computers, and to opera; and so on. Aren't books bought by organisations which serve important social functions, which are already under serious budgetary pressure? Of course they are. But so are many other commodities. The simple fact is that if you are going to raise between 35% and 40% of national income in tax there have to be rather strong and exceptional reasons for exempting any activity from making a contribution

Now I know that if I state out a case for imposing VAT on books I will be lynched out of Cardiff University and on to the next train back to London, so that I shall put the responsibility for these arguments on one of my colleagues. He was recently approached by a television company and asked what he thought about VAT on books. He said he was in favour of it and they talked for a few minutes. The producer then said he wasn't really the person they were looking for; they had assembled a group who were passionately opposed to imposing VAT on books and what they wanted was someone who was passionately in favour of VAT on books. My colleague demurred and said that no reasonable person could be passionately in favour of VAT on books. You might think that on balance, taking into account all the advantages and disadvantages, putting VAT on books was a better course of action than not putting VAT on books; but that was all. And sure enough the producer rang back an hour later and said that they couldn't find anyone who was passionately in favour of putting VAT on books and would he like to come on the show.

Now there is a point of substance, and general application there; the clamour of the vested interest tends to drown the quieter voice of reasoned analysis. Yeats' lines 'the best are in confusion, while the worst are full of passionate intensity' seems to apply to ever increasing areas of natural life and tax policy may be one of them. Certainly I find the worst moments at the policy seminars my Institute organises are those when someone from a bank gets up and says that the real problem is that banks pay too much tax; and then someone from an oil company will argue and point out that oil companies pay too much also; and the chorus will extend round the room. Moreover, they all will sympathise with each other. A former chairman of the Ways and Means Committee of the American House of Representatives summed it all up in a famous ditty:

"Don't tax you

Don't tax me

Tax the man

Behind that tree"

I've always wanted to follow a pantomime tradition and have that on a curtain which I could pull down when things got out of hand and make the audience join in.

The vested interest thesis has the advantage, too, that it points to its own direction of reform. What is needed, it suggests, is to sweep away all the reliefs and allowances which have gradually encroached on the tax system as a result of the pressures, and go back to a much simpler structure with a broader base and much lower rates. This is an idea which is always attractive to people who thinking about tax reform although not, in the main, to people who have thought about tax reform for very long. It is the motivation behind the current drive to reform both in Britain and in the United States.

I predict, with some confidence, that these reform packages will fail, on both sides of the Atlantic. Moreover, they will fail in ways which will appear to supported the vested interest thesis. Nigel Lawson will back off from fundamental changes to pension funds, taxation or to VAT under pressure from the exchange rate and his own backbenchers. The American package will be chipped away by a variety of interest groups. We will all conclude that political pressures are too great to make tax reform feasible, and drop the matter for another ten years or so. In Britain, at least, there is a ten year cycle for half baked attempts to reorganise the tax system.

But today I want to argue that the vested interest thesis is mistaken. Of course there are such interests at work in tax policy and, in the main, their influence on the design of tax structures is malign. But the idea that if only the Inland Revenue were left alone it would be capable of dispensing until most of the contents of Butterworths is nonsense; indeed it has never presented any coherent suggestions for doing so. It is not the politicians who imposed all this legislation on the Inland Revenue; it is, in the main, the Inland Revenue who imposed it on the politicians Today I want to explore how, and why. Because members of parliament do not spend each summer staying up at night in order to vote on and incomprehensible clauses of the Finance Bill because they want to stay away from their wives; they do so because they have been told by their technical advisers that this is the only way of making the system work.

But is this claim true? Is the vested interest thesis valid? Instead, I want to propose an alternative thesis. It is that there are some parts of the tax system that work well, and others that work badly. Most of that time spent late at night in Parliamentary committee, and most of the content of these volumes, is devoted to ineffectual reinforcement of the bits that work badly. Most revenue is in fact collected from relatively brief and comprehensible statutory provisions that require relatively little annual refinement and amendment. The appropriate direction of reform is to develop on and from the bits that work and to reduce the scope of the bits that do not work and reduce the dependence of revenue upon them.

Let me begin by explaining some of the weaknesses of the vested interest theory. Firstly, it is historically wrong. Anyone examining the history of the development of the British tax system would be more surprised by how little it reflected the influence of political pressures than by how much. Let's look at the three most notorious reliefs; the relief for life insurance policies (abolished in the last budget), tax exemption for pension funds, and tax relief for mortgage interest.

Life assurance relief was introduced by Gladstone, in 1853. Its history is a curious one. At the time, Gladstone was under pressure to tax earned income more favourably than investment income. The argument, which sounds rather strange to us now, was that earned income was more precarious than investment income; your earnings disappeared if you fell under a hansom cab, or if you retired, while the flow of revenue from your 3% consols went on for ever. Gladstone determined to oppose this idea (earned income relief was, in fact introduced fifty years later by Lloyd George) but hit instead on the idea of giving life insurance relief: it was intended as a form of depreciation allowance, which enabled you to provide for the replacement of your human capital on death or retirement. The life insurance industry opposed it at the time, on the very sensible grounds that it would distort the market for personal savings.

Tax exemption for pension funds dates from 1921. It was introduced as part of a search for fiscal neutrality. It was recognised, at a time when the tax threshold was way above average earnings, that the majority of members of occupational pension schemes would not be liable for tax on their own retirement savings. Thus tax exemption was extended to funds accumulated on their behalf, as it already had been to friendly societies for similar reasons.

Mortgage interest relief originated because relief used to be available for all interest payments - for the good logical reason that if you pay tax on the interest which you receive, you should be relieved of tax on the interest which you pay. As a matter of fact there used to be a tax on income from home ownership but this was shrunk to negligible amount and was finally abolished in 1962. Gradually tax relief was withdrawn from most other kinds of consumer borrowing - consumer credit has always been seen as somewhat disreputable - until the existing relief for loans for house purchase up to 30,000 was all that remained.

You will notice two characteristics of all these reliefs. Not one of them came into existence as a result of pressure from vested interests, even if vested interests are now active in their defence. Second, the rationale for each is long forgotten and no longer applies. The picture of perpetual raids and depredations on the integrity of the tax system by interest group pressures bears no relation to historical reality; bureaucratic evolution is the more accurate description

We can learn some lessons from this story, too, about when interest groups are and are not a problem. It is much easier to defend what you have than secure what you don't have. My guess, for what it is worth, is that the book lobby will probably succeed: but if it was arguing for zero rating to be introduced, rather than against it being taken away, I do not think it would have a chance. The charity lobby, with a much stronger pull on public sympathy, has achieved almost nothing in ten year of lobbying.

The reason is that good politics, or more accurately good bureaucracy, is above all about avoiding fuss, and trouble. Avoiding embarrassment is, above all, the cardinal objective; and this militates against change, either in introducing a relief or in taking it away once it has been given. The corollary is that if you take a robust view of this kind of fuss and trouble you have immense freedom to do what others see as impossible or impracticable. Take the last budget as example. Two principal victims of it were two industries - life insurance and leasing - which were previously seen as exceptionally powerful lobby groups. And they are; they are efficient, well financed and organised; and they have ready access to ministers, to civil servants, to party officers. In short, they have considerable capacity to create fuss. But that is all they have. The number of voters who are chairmen of life insurance or leasing companies is small and, in the main, they do not live in marginal constituencies. If you do not mind a certain amount of fuss; if you are determined to say no; you will quickly discover that there is in fact no bullet in the smoking barrel. And that was precisely what emerged last spring. Both these lobby groups quickly surmised that there was no point in fighting battles they were not going to win and devoted their energies to other questions.

The attack was easier, too, because it was part of a package of measures. If the only change you make is a technical change in the tax treatment of leases, the only people you will hear from the leasing industry. No one else will be interested; the capacity for effective fuss is maximised. If, on the other hand, you take on several groups at once you stand a chance of winning popular support for your adoption of a range of measures and you identify the complaint of each particular interest group as the self-centred special pleading which it almost certainly is. Paradoxically, you will win more victories if you take on several enemies at once than if you try to pick them off one by one.

And there is a lesson here for the book lobby. For anyone to introduct the taxation of books as an isolated measure would be the height of political folly. If it is to be done, it would be best done as part of a much wider package of extensions to the VAT base. And that is how, I predict, it eventually will be done.

There is one final point worth noting before we leave this discussion of the characteristics of an effective pressure group. The more technical a change is, the lower the degree of political influence over it. Let me give one, rather remarkable, example. The extension of VAT to food is generally considered too hot an issue for any politician to handle. But of course we already tax food, heavily, through the mechanics of the common agricultural policy. And before you blame that entirely on the EEC, let me remind you that for most of the last five years or so we have chosen to impose additional internal taxes on food by not revaluing the green pound with the actual pound and imposing positive monetary compensatory amounts. I can see eyes glazing ever already and that, of course, is exactly the point. The concept of VAT on food is comprehensible. Everyone knows what that means, at least roughly. But the mechanics of the common agricultural policy are beyond the understanding of all but the most masochistic genuises. Hence you can do things through one that you could never do through the other. It matters not that the effects are the same.

If I can make one last reference to books to drive home the point. The neatest suggestion I have heard for setting the Chancellor off this particular hook is that the budget should exempt books from value added tax. What man of learning could fail to applaud a Chancellor who exempted books from taxation? Nigel Lawson would be carried round Parliament Square on the shoulders of Gyles Brandreth and Margaret Drabble, before being buried in Westminster Abbey with other great patrons of literature. Those of you who understand the mechanics of value added tax, of course, will know that exemption would involve a much heavier burden of tax on books than exists at present.

Let me try to summarise the progress of the argument so far. Historically, vested interests have not had major influence on the structure of our tax system, and their effect has mainly been negative rather than positive - they have been able to stop adverse changes rather than to introduce favourable ones. Their effectiveness is easily exaggerated, especially for producer groups, which can be rather readily defeated by a determined minister. In relation to technical changes - even matters of fundamental, indirect, importance - ministers and civil servants have a relatively free hand to determine how the system evolves.

How then does it evolve? Why are tax systems so complex? Is that complexity inevitable? The first point to note is that it really isn't true that the whole of the tax system, universally and uniformly, gets worse. There are parts that work relatively well, and tend to get better; other parts that work worse. I plan to elaborate more on which are which, but let me give you the first clue; the ones that work well are those where the tax is based on transactions. Where the tax base is derived from money changing hands - the purchase of goods in a shop, or the payment of a weekly wage, for example - the associated legislation is relatively straightforward. Where the tax base requires calculation, or valuation, or assessment, then things become relatively complicated.

To see the force of the point, consider again some fiscal history, and let us go back before the origins of the present tax structure. Several hundred years ago the two principal taxes were import duties and stamp duties. The market economy was limited, and imports and property transfers provided on occasion for tax changes because it was then that recorded transactions took place while the King, or his agents, could observe and monitor. When emergencies took place - wars, for example - then two other types of tax were employed. One was a wealth tax; this worked well so long as wealth mainly took the form of tangible property - castles, land, cows - because the problem of assessment was fairly straightforward. Once society became more complicated, business trade and finance emerged, wealth became much harder to identify. Wealth taxes didn't work so well and were generally abandoned - a lesson which some would be fiscal reformers have yet to learn.

The second tax was of course inflation, which then took the form of straightforward currency debasement; the government reduced the metal content of the coinage, or simply issued new debased currency of its own (the idea of printing banknotes, or the equivalent, was slower to catch on).

Eventually, however, the revenue requirements of the modern state became greater than these devices would bear and for Britain the crunch came in the Napoleonic wars. Income tax was introduced in 1799. Its structure was re-organised by Addington at the beginning of the nineteenth century and the schedular system which he introduced has formed the basis of administration of the tax ever since. The schedules listed the sources of income. First, and most substantial, was property; then woodlands; interest; trade, profession or vocation (recall that the invention of the limited company was fifty years ahead); and finally offices or employments, which generally referred to crown appointments; the idea that employees were people of sufficient standing to be liable for income tax was not conceived of. Finally, there was a catch all for any other income which failed to fall within any of those schedules.

The list is redolent of the world in which the tax was introduced-one two centuries distant from today, and one in which financial affairs had a simplicity and a stability which vanished long ago. The tax was a tax on income, rather that a tax on the income of individuals; indeed it was thought unacceptable that anyone, even a tax collector, should know what a man's income was, and this was one of the rationales of the procedure of separate schedules. If you only had one piece of the jig-saw at a time you would not know what the picture on the puzzle looked like.

You will notice a feature of this structure is the absence of any definition of what income is. Income is exemplified, but not exhaustively, and nowhere described or defined. In 1799, anyone could recognise income when they saw it, but this was not necessarily to remain so. It was left to the Courts to identify what was meant by income, and it cannot be said that they have made a spectacularly good job of it. Income tax, said Lord McNaughten in what has become the key judicial dictum, is a tax on income; but this does not necessarily help a great deal. Income according to generally accepted accounting principles; well, sometimes, but not always and not necessarily and subject to appropriate exceptions.

Economists, as well as accountants, have struggled with definitions of income for long enough, and there are several volumes of tributes to their efforts and their lack of success. Perhaps the best known is that of Hicks. For Hicks, income is what a man can expect to consume during a period and still be as well off at the end of the period as he was at the beginning. This is a valuable concept, but one which is not enormously helpful to the tax collector. How does he establish how well off a man expects to be at the end of the period? And what is the taxman to do if these expectations are unreasonable? He may be relieved, or disappointed, to discover that Hicks came at the end of this discussion to the conclusion that "we should eschew the concepts of income and capital in economic analysis. They are bad tools, which break in our hands."

But it is premature, surely, to abandon the search for an operational measures. Fortified, perhaps, by another distinguished economist - Dennis Robertson - who wrote 'the goals and workhouses of the world are full of the people who gave up as a bad job the, admittedly difficult, task of distinguishing capital from income'. I was encouraged, when I agreed to give this lecture, to be sent a copy of an earlier one entitled 'Inflation Accounting - all the Answers' by Professor Whittington, but disappointed to find that it didn't wholly meet the promise of its title. I don't mean any disrespect there to Professor Whittington, because the central issue emerges very clearly from that lecture. It is that income is, in the last analysis, a subjective concept whose size depends on the judgment of the accountants who compile it and the particular purposes for which the measure will be used. This doesn't mean that the income concept doesn't have substantial practical value: it does, and this is the reconciliations of Robertson and Hicks. Robertson sees, quite correctly, that without an income concept as a guide to practical action individuals and companies are likely to make serious mistakes: Hicks observes that this practical guide cannot necessarily be given the precision of definition which he requires of it. Income, in short, is a necessary concept but one which cannot be given the precision or objectivity that some of its uses might require.

Foremost among these is its use as a tax base, and the principal source of the bulk of the extensive statutory provisions is successive, not very successful, attempts to give that precision to a concept which intrinsically lacks it.

Now it would be wrong to say that income tax doesn't work. After all, it does bring in more revenue than any other tax; and the bulk of that revenue comes from the part that Pitt and Addington certainly never envisaged - certainly never envisaged - the part that comes from PAYE. Now before anyone goes away and quotes me as saying that PAYE works well, I want to qualify that. It is unnecessarily cumbersome and expensive to administer both for employers and the Inland Revenue: largely incomprehensible to employees: and the fact that it is still not yet computerised is a national scandal. But although it is PAYE which brings in the bulk of the revenue, the proportion of the tax statues which is devoted to issues affecting PAYE type tax payers is negligible. And the reason for that is straightforward. For most PAYE taxpayers, the income on which they are taxable is the amount which is paid over to the worker in cash at the end of the week or to his bank account at the end of the month. There is a transactions base for the tax.

And if you look at the complications which arise in relation to the taxation of income from employment, you will quickly discover, that these are the exceptions which demonstrate the force of the general rule. There are two main areas of complication: pensions and other benefits paid after the end of employment: and fringe benefits and expenses of employment.

Pensions are the most significant benefit which arise as a result of employment, but which do not take the form of cash in hand. Instead, we tax them when, and only when they are received by the employee - when they actually do take the form of cash in hand. This arrangement is simple and convenient and works very smoothly, but is seen as an anomaly, and within the context of the income tax it is an anomaly. I've just written 70 pages on the complications of doing anything else, however, and I don't propose to bore you with that tonight.

The second difficult area I mentioned is the other one in which people get something other than cash from their job, and that is fringe benefits. Here too it is the absence of an immediate transaction base that creates the difficulty. If you look at other areas of income tax you will find, I believe, that the same principle applies. If there is a cash transaction, there is relatively little complication. If there is no cash transaction, then there is complexity.

But the more striking example of the theory that it is the transactions based tax that escapes legislative complexity is the contrast between income and corporation tax and VAT. Let me base my case on the answer to one simple question. I know there are representatives here tonight from Deloittes, and from other major accounting firms. Let me ask you what proportions of tax partner and manager time are devoted to income tax: to corporation tax: and to VAT. Let's stop for a moment while you think about it. And then compare these ratios with the revenues this year from the various taxes. In Table 2, I have undertaken a rather different calculation of million of revenue per statutory page. The differences are startling.

That's not all. Experience with VAT since its introduction in 1973 directly contradicts the preductions of the vested interest thesis of tax evolution. The theory says that the base should get steadily eroded by political pressures. Actually, all the major changes in it have been extensions and, lest you think we've been through a freakish period, recall that the same was true of the purchase tax that preceded VAT. Have the rules become steadily more complex? Not substantially. Do we add a few more pages to the legislative code each year? Well, its true that last year's Finance Act had several pages on VAT, much of it concerned until the extension to include building alterations-probably a simplification, on balance, and certainly seen as such by Customs and Excise. But you would have to go back several years before that to find any substantial fraction devoted to VAT.

But for a last word on the subject I turn to a different tax handbook - one of Tolley's - who publish a manual of tax avoidance, euphemistically called Tax Planning. They do have a chapter on VAT - one out of 54. They say, "the manual even includes a chapter on VAT. This may seem surprising since VAT is generally regarded as an unplannable tax". With that, I rest the elements of my case. Complexity in taxation is the result of departure from a base which is closely related to actual transactions. The greatest villain is the concept of income which, whatever its practical value as a guide to action, is too difficult in objective assessment, too remote from the transactions which lie behind it - or ahead of it.

These observations are not intended only as a critique of income tax. They apply, a fortiori, to a wealth tax. Many people are passionately opposed, in principle, to a wealth tax. I am not. I am, in principle, in favour of a wealth tax, especially if one could contemplate a substantial substitution of wealth tax for income tax. I am even willing to be at least mildly passionate in favour of a wealth tax. But I also recognise that such a tax and such a substitution is quite impractical. A wealth tax would end up as a tax on those elements of wealth which could be conveniently identified and assessed - certain kinds of tangible property, deposits, quoted securities. These are a small proportion of total wealth and one which would be likely to diminish. I do not doubt that the Inland Revenue could operate something called a wealth tax - as their counterparts in many other countries do. I do doubt that they could get substantial revenue from it, and that the experience of other countries supports.

If reducing complexity in the tax system involves a shift to transaction based taxes, what does that actually mean in practical terms? First, it means a shift from reliance on income tax to reliance on VAT. That means broadening the base of VAT, and increasing the rate. Back to books again, you may think, but actually the real issue here is not books but food, fuel, and perhaps housing. The obvious objection to these major extensions in the distributional impact of the change - low income households spend a much higher proportion of their budgets on food, fuel, housing; but one of my colleagues and I have recently demonstrated how, if the revenue from a broader tax base is used to increase income tax thresholds and national insurance and supplementary benefits, the overall effect can actually be more progressive than anything we have at the moment.

But a more fundamental shift is to switch income tax to a realisations basis - to move to what Americans have called a cash flow, or consumed income, basis or - since it amounts to the same thing - to levy direct tax on expenditure rather than income. The basic principle is that you tax business and investment activities - the area of income tax which gives rise to most complication - only when realisation takes place, and then on the basis of the net amounts withdrawn, whether of income or of capital. Tonight is not the occasion to elaborate these ideas, because I have done so at length elsewhere. What I want to do tonight is to answer the claim - frequently made - that fundamental tax reform is futile because whatever we do politicians will mess it up. It is not politicians who have messed up with system we have. It is administrators, trying to make the best of fundamentally unworkable principles.

This must sound like a note of pessimism on which to end this lecture. I don't intend it to be. Because I am clear that the answer to the question with which I began it is no. Complexity in taxation, or at least the degree of complexity which we have, is not inevitable. Tonight, I have tried to analyse why it is that we have so much complication, and why it is that existing political and administrative structures tend to operate in order to increase that complication rather than to reduce it. In going through that analysis, I hope you have seen that that complexity was not unavoidable, although the changes required in order to escape from it are considerable.

First, it is important to understand that it is not true that complexity is to any great extent the product of political pressures. It really is not the case that the system is complex because it is a finely honed instrument of sophisticated economic management, or that it is so complicated because it is so fair. It is mainly complicated because of attempts to buttress bits of the system that do not work and probably never could have worked.

Second, it is not true that the whole of the tax system is in this kind of mess and that it really doesn't matter what kind of tax structure you have because wherever you start it will it all quickly degenerate. Most of the complexity relates to elements of the income, capital gains and corporation tax systems which raise a relatively small proportion of total government revenue.

Third, existing political and administrative structures appear incapable either of generating worthwhile structional reform or of implementing them when imposed. Changes in the systems and criteria which determine policy are preconditions of a simpler system.

Here is information on how to order Drafting Trusts and Will Trusts and other books by James Kessler QC.

Home