UK tax system complexity: causes and consequences

People often complain that the tax system is ‘too complex’. What does this mean, why does it happen and why does it matter?

What is tax complexity?

Although the term ‘tax complexity’ is widely used, with the complaint usually being made that the tax system is ‘too complex’ – people rarely complain that the system is ‘not complex enough’ – the concept of tax complexity turns out to be quite elusive when one tries to pin it down.

To start, what do we mean by ‘the tax system’? By a ‘tax system’, I mean the set of tax laws/rules that define:

• the various types of agent – companies, individuals and households – involved in paying taxes;
• the types of transactions that these agents might undertake that give rise to taxes or reliefs;
• the rates and duties that apply to these transactions; and
• the set of administrative procedures that agents have to go through by way of providing information, completing tax returns, paying tax, and undergoing investigations where tax returns are challenged.

It is important to note that the UK tax system – the set of rules defined by UK tax law – is not the same as the tax system faced by all UK taxpayers. For very many individuals and for quite a lot of companies with rather simple tax affairs, the relevant tax system will indeed be the UK tax system. But for a large number of individuals and companies, the relevant ‘tax system’ will be some part of the international tax system that involves the tax rates and the rules and administrative conditions applying in many different countries. Which countries might matter will vary from taxpayer to taxpayer, depending on the variety of countries in which they actually conduct transactions or might contemplate conducting transactions.

Given the above definition of a tax system, it is helpful to distinguish two different dimensions of complexity: design complexity; and operational complexity.

Design complexity

This reflects the number of different commodities/transactions that are taxed and the number of different tax rates that apply to each of these. This latter encompasses a number of different issues:

• whether, and how, for any given commodity, the rate at which it is taxed varies with the total amount or value of the commodity that the taxpayer buys or sells, e.g. income tax rates, stamp duty rates;
• whether, and how, for any given commodity, the tax rate paid by some taxpayers differs from that paid by others, e.g. whether incorporated businesses pay tax at a different rate on their profits than unincorporated businesses; and
• even if, for given commodities, there is a single tax rate paid by everyone, whether and how the rate paid on one commodity differs from that on another, e.g. why some goods are zero-rated for VAT and others are not.

It might be thought that complexity arises because of the number of different tax rates; and that, if everything were taxed at exactly the same rate, then this would be the least complex tax system. But this is potentially misleading. When you tax commodities, you essentially raise the price of things that consumers buy – e.g. bread – and lower the price of things that they sell – e.g. labour. The net result is that consumers get less bread for every hour they work. So if you tax everything at exactly the same rate – with 20% tax on income and 20% tax on bread – then, in very simple terms, you are essentially doing the same thing twice over: reducing the amount of bread that people get per hour of work. You could achieve the same outcome by having a 40% tax on income and a 0% tax on bread; or a 0% tax on income and a 40% tax on bread; or any combination in between.

What this very simple example suggests is that a tax system in which there is a single rate of tax – in the example above, 20% – applied to all commodities may to all intents and purposes be equivalent to one in which there are two different tax rates – 40% on income and 0% on VAT. In this case, can we say that one system is more complex than the other, or should we say that they are equally complex? Indeed, it might be argued that a system in which there is a single rate of tax applied to all commodities is more complex than that in which there are two different tax rates – in the example, a single rate of tax of 40% on income, coupled with a zero rate of tax on all consumer goods – simply because, in the second case, there are far fewer things that are effectively being taxed.

At the other extreme, you could think of a tax system that not only taxes all commodities at different rates, but also has non-linear taxes with multiple bands and rates for various commodities, and these tax rates and/or bands can vary by household and firm type. We might all agree that this will have an extremely high level of complexity. But how do we decide on the ‘right’ amount of different tax rates to have? Any given tax system has multiple aims:

• to raise revenue to fund public expenditure;
• to promote fairness by having progressive income taxation, and/or taxing less heavily those things that are consumed predominantly by the poor, and more heavily those things that are consumed predominantly by the rich; and
• to promote economic efficiency (growth and productivity), which involves taxing ‘bads’ such as pollution, rather than ‘goods’ such as work and savings; and where ‘goods’ must be taxed, taxing more heavily those things that are more ‘sticky’ (less price sensitive).

The traditional treatment of tax design by economists focuses on these three objectives, and assumes that taxpayers are fully compliant. However, in the current climate of concern about tax avoidance, it is important to recognise a fourth objective: to reduce opportunities for non-compliance through avoidance and evasion.

The economic theory of tax design helps us understand how to best design a tax system that achieves these objectives. Associated with this ‘ideal’ system will be some level of complexity – in the sense that there will be some degree of progressivity and that different commodities are taxed at different rates. This can provide a benchmark against which we can decide whether a given tax system is ‘too/unnecessarily complex’. So some degree of design complexity is an inevitable consequence of any tax system that has the aims of raising revenue, redistributing income, and doing so in a way that is the least damaging to productivity and growth.

Now any given tax system will typically have a design that is far from optimal, as defined above. This may not always be associated with excessive complexity – it may just be that the tax rates are wrongly set; for example, the rate of tax on some ‘bads’ may be too low, while that on some ‘goods’ is consequently too high. Often, though, tax systems do end up having too many different tax rates and reliefs, as politicians pursue additional objectives. These  may have a strong political imperative at a particular moment of time, but this recedes as the economic and political climate changes, leaving the rate or relief in place. This results in the need for periodic overhaul and reform.

However, policy purposes can change over time, and therefore so too can the complexity of the ideal system. This means that an apparent increase in complexity does not necessarily represent the system’s becoming ‘unnecessarily complex’. When vehicle excise tax was first introduced, it was done so as a relatively efficient revenue raising measure. Over time, car ownership became a necessity rather than a luxury, and so became relatively price insensitive. More recently, vehicle duty has been seen as a tool to help achieve environmental objectives; and, as a result, the rate of duty has been differentiated by class of vehicle. Whether this is an example of the system becoming too complex depends on whether you think there are better ways of achieving the environmental objective.

Operational complexity

The second dimension of tax system complexity is its operational complexity. This essentially reflects how costly it is for an honest taxpayer to comply with the informational, filing and payment obligations of the tax system.

No all such costs could be said to arise because of the complexity of the system. For example, for taxpayers with cash-flow issues, there may be the costs of meeting the payment obligations. There is an inevitable fixed cost in time and money in filing a tax return, however complex the system.

But some costs do arise due to the complexity of the system. Operational complexity reflects how easy it is for a taxpayer to map the various transactions they undertake – and the terms in which they understand these transactions – into the categories used by the tax system and the language in which these are described. To some extent, this aspect of complexity will relate to the tax design complexity discussed above. Other things being equal, the more distinctions that there are between the different categories of transaction and the tax rates these attract, the more costly it may be for taxpayers to make sense of their tax obligations.

But operational complexity could arise for other reasons:

• The fit between the terms in which the taxpayer conducts their affairs and the way in which the tax system treats different transactions could be low. The tax system may treat as different certain types of transaction that the taxpayer treats as identical; or treat as identical transactions that the taxpayer regards as different.
• The language that is used to define transactions may be difficult for taxpayers to understand. There is an understandable desire by HMRC to write tax law and guidance in a language that reduces legal ambiguity, and that will survive challenges by lawyers and courts. This can often sound rather stilted, though, and it may not be the language in which individuals understand or describe their affairs. There may be more effective ways of combining the two objectives – using the legally correct terminology, but giving an illustration in more common language which will be accurate in the vast majority of cases.
• Inconsistencies in tax law/guidance. At one point there were five different definitions of a ‘child’ in the US tax code.
• Taxpayers may not fully perceive or understand the logic behind the various steps they have to take to complete their tax returns. The complexity can be reduced, by giving taxpayers as many opportunities as possible to answer a simple question and then skip a great number of steps that do not apply to them.

While these factors can contribute to the operational complexity, there is an extent to which this complexity will fall over time, as taxpayers learn about the tax system and become more familiar with its definitions. So a fifth aspect of operational complexity has to do with the frequency of changes.

The discussion of tax design complexity distinguishes the inevitable complexity of an ‘ideal’ tax system, and the unnecessary complexity of a given tax system. The same distinction could apply to operational complexity. There may be certain irreducible information requirements that a tax authority needs from taxpayers.

Over time, these irreducible informational requirements can change – for example, because of changes in technology that allow HMRC to capture information provided in one context and apply it in many others, thus reducing the need to capture essentially the same information repeatedly. In addition, smart apps may be able to gather information from taxpayers in a relatively costless way and use this to populate a tax return.

The costs and consequences of complexity

There are a number of reasons why tax complexity matters.

Distortions

If the design of the tax system is unnecessarily complex, it could create unwarranted distortions that discourage productive economic activity. This has costs that can, in principle, be measured as lost gross domestic product (GDP). However, I stress again that there is no automatic link between complexity and the distortionary costs of the tax system.

Non-compliance

Tax complexity can create opportunities for tax avoidance that can create significant costs to the economy in terms of both reduced efficiency and fairness. The efficiency losses arise for a number of reasons, including:

• in the presence of avoidance, tax rates have to be higher than otherwise in order to raise given revenue;
• very bright people are being employed to devise, and then to detect and counter, elaborate schemes of essentially paper transactions to move money around and reduce tax liabilities;
• there is increased inequality – as these schemes are expensive, it is typically the better off who can avail themselves of them.

Nevertheless, it is important to recognise that tax avoidance may actually be a way of reducing some of the potential distortionary costs induced by excessive complexity.

Compliance costs

Complexity can increase the costs of complying with the tax system, due to the additional time spent by taxpayers in meeting their obligations, or the financial costs of using professional advisors. While not all compliance costs arise because of complexity, nevertheless the factors giving rise to operational complexity will give rise to compliance costs.

Legal uncertainty

Operational complexity can potentially give rise to legal uncertainty. This arises when taxpayers do not fully understand their true tax liabilities; how certain transactions should be treated for tax purposes; and the basis on which HMRC comes to a different view on how their affairs should be treated, when it challenges the tax return.

Increased compliance costs and legal uncertainty largely affect the taxpayers faced with the consequences of complexity; but the costs of distortions and non-compliance fall on everyone. With the recommendations of the Smith Commission and calls for further devolution of tax powers, these issues are extremely pertinent and likely to become more so.