Global responsible tax: How complicated is simplification?

KPMG International hosted a virtual roundtable discussion in July 2026, as part of the Global Responsible Tax Program, to explore what it would really take to simplify tax systems and what might be lost, gained or exposed in the process.

Everyone wants tax to be simpler. It is a perennial challenge, with questions of certainty, compliance burden and basic comprehensibility consistently raised. A simpler system, many argue, would support growth, improve investment confidence and make it easier for taxpayers, businesses and revenue authorities to ‘do the right thing’. But the harder question is, what trade-offs are organizations prepared to make to get there?

Tax systems are asked to do many things at once: raise revenue, support growth, redistribute fairly, encourage certain behaviors, deter others, provide certainty, prevent avoidance and remain administratively workable for taxpayers, advisers and revenue authorities alike. Complexity is often the result not simply of bad design, but of competing political, economic and social objectives being layered into the system over time.

But how complicated can simplification be? Removing a relief may broaden the base and reduce distortion, but it may also remove targeted support. Fewer rules may make a system easier to understand, but they also may create more discretion, uncertainty or perceived unfairness. Harmonization across borders may help businesses and tax authorities operate more consistently, but it may also reduce national flexibility. Strong anti-avoidance rules may protect revenue and public trust, but it can also add complexity for those simply trying to comply.

Technology and AI add a further dimension. Digital administration, data analytics, real-time reporting and automation may offer ways to make tax easier to navigate and administer. But if tax rules become more automated, who is responsible for interpretation and how can taxpayers understand and challenge decisions? If systems become easier to comply with but harder to see, has tax really become simpler? 

Held under the Chatham House Rule, the discussion brought together academics, policymakers, business representatives and civil society voices. The write-up below summarizes the personal views expressed and does not necessarily represent the position of any organization, including KPMG International Limited or any KPMG member firm.

Is complexity about how much legislation exists, or how effectively people use it?

It was agreed by many in the discussion that legislative length (or indeed number of taxes) is a poor proxy for the practical difficulty of a tax system. The length of the UK’s tax code is often cited in exasperation, but a lengthy provision may be relevant only to a narrow group and provide those taxpayers with valuable clarity. Conversely, a short or principles-based rule may create considerable uncertainty if it leaves important questions to interpretation.

It is therefore important to distinguish between the volume of law and its effectiveness.

The relevant test may not be how many people ultimately claim a relief or fall within a provision, but how many must spend time determining whether it applies to them. A narrowly used rule need not create widespread complexity if its intended users can identify and apply it easily.

Clear drafting was seen as particularly important. Terms such as ‘reasonable,’ ‘normal’ or ‘just’ may permit flexibility, but they can also create blurred lines. Several participants favored rules that allow taxpayers to reach a clear yes or no answer in as many circumstances as possible.

Periodic pruning of obsolete or duplicative provisions was supported, but not as a substitute for improving the usability of the law.

Is the deeper economic problem unpredictability rather than complication?

As the discussion developed, many participants increasingly framed the challenge as one of uncertainty. Complexity and uncertainty are related, but they are not identical. A technically detailed system can still be relatively predictable, while a shorter and more discretionary system may produce divergent outcomes across regions, officials or taxpayers.

Different forms of uncertainty were identified. Legal uncertainty concerns whether legislation provides an intelligible answer. Administrative uncertainty arises when the interpretation or behavior of a revenue authority is difficult to anticipate. Policy uncertainty reflects the frequency of rule changes and the possibility that taxes, reliefs or thresholds may be altered with limited notice.

Long-term predictability was considered particularly important for investment. A business may be able to comply with a complicated incentive, but it may still decide not to invest if it cannot be confident that the relief will apply, that its claim will be accepted or that the policy will remain in place.

One participant also raised the cognitive burden created by uncertainty. Businesses must devote management attention to interpreting rules and anticipating challenges, while governments and tax authorities must understand the interaction of numerous provisions before making further changes.

Greater access to binding clearances was also suggested as a means of reducing the cost and risk associated with uncertainty over difficult legal interpretation and boundary questions.

Technology can streamline compliance, but can it simplify the underlying policy?

A central distinction emerged between the design of a tax and the process used to administer it. Either can be simple or complex independently of the other.

Process simplification may offer the most immediate opportunities. Third-party reporting, pre-populated returns, integrated data systems and faster digital filing can reduce the information taxpayers must gather and submit themselves. Better use of third-party data may also narrow the tax gap by reducing non-reporting. However, some participants cautioned that the burden does not disappear; it may be transferred to banks, platforms, employers or other institutions required to provide the information. Where businesses are required to provide more third-party data, this should be considered holistically alongside opportunities to streamline or remove other tax or wider regulatory reporting requirements.

It was also suggested that administrative complexity can also arise from limitations in revenue authority systems, particularly where these are fragmented or have not been upgraded for a significant period, requiring businesses to devote substantial time and resources to resolving issues that are not caused by the underlying tax rules.

Several participants argued that machine-readable rules will increasingly require more structured conditions and clearer decision points. AI may assist by summarizing legislation, navigating guidance and identifying relevant provisions. However, one participant was wary that AI was the right way forward for tax legislation as its simplistic style can leave things out and hallucinate. 

What would change if tax rules were built around small and medium-sized enterprises (SMEs) rather than adapted for them afterward?

The burden of tax complexity was described by one contributor as highly regressive. Large organizations can draw upon in-house tax teams, established relationships with revenue authorities and external advisers. Small and microbusinesses often lack comparable resources, making the same rule considerably more expensive for them to understand and administer. One contributor argued that tax jargon makes life harder and that language should be recognizable and understandable to taxpayers.

Some participants argued that governments frequently ‘think large first’: rules are designed around the structures and capabilities of major organizations, with exemptions or simplified treatments added later for SMEs. This can produce a patchwork of thresholds, elections and special regimes that itself becomes difficult to navigate.

The discussion of research and development incentives illustrated the problem. Some smaller businesses were said to be withdrawing from claims because the process, uncertainty or risk of challenge outweighed the expected benefit.

Ideas included allowing microbusinesses to base tax more closely on their financial accounts, with reduced access to bespoke reliefs in exchange for lower compliance costs. Such approaches would involve trade-offs. Taxpayers might gain simplicity but lose opportunities to optimize their position.

What can jurisdictions learn from one another without assuming that systems are inherently transferable?

International examples demonstrated that simplicity takes different forms.

Singapore was cited as an example of a highly streamlined filing process supported by employer reporting and a relatively narrow range of taxable personal income. An individual with straightforward circumstances may be able to complete a return in minutes.

At the same time, it was noted that this administrative simplicity is connected to wider choices about the tax base, redistribution and the role of government. A process cannot be separated entirely from the policy system it administers.

India was mentioned as an example of technology and third-party data being used to pre-populate income information, facilitate electronic filing and accelerate processing. This illustrated how administrative reform can generate significant practical benefits even within a large and complex economy.

The European Union (EU)’s emerging tax simplification agenda was discussed as an example of reform focused less on wholesale redesign than on cleaning up an increasingly layered body of rules. Participants pointed to the proposed tax simplification omnibus and the recast of the Directive on Administrative Cooperation as attempts to consolidate successive amendments, remove obsolete provisions and reduce overlaps between reporting regimes.

Some proposals were seen as relatively straightforward administrative improvements, while others were more ambitious, including reducing withholding tax frictions on certain cross-border payments and reconsidering whether some anti-avoidance rules remain necessary where newer regimes, such as Pillar Two, already address similar risks.

The central constraint is that EU tax measures generally require unanimous agreement among member states. This makes simplification both legally and politically demanding. The likely outcome is that the most pragmatic measures may progress, while more ambitious proposals are diluted, delayed or removed. The EU case therefore illustrates a broader lesson from the discussion: simplification is often easiest where it involves consolidation and duplication, but much harder where it requires governments to surrender discretion or accept a different balance between competitiveness, enforcement and national autonomy.

Can tax systems be improved without destabilizing economic activity?

Some participants were skeptical that wholesale reform could be implemented safely or sustainably. Rewriting numerous elements simultaneously would place significant demands on legislative drafters, revenue authorities, software providers, advisers and businesses. It could also create an ‘economic freeze’ as taxpayers delay decisions while waiting to understand the new system.

The discussion therefore pointed toward incremental reform. Better use of third-party data, removal of duplicative reporting, clearer language, periodic review of thresholds and targeted improvements to areas of disproportionate burden may be more achievable than a single comprehensive redesign.

Incrementalism is not without risks. Repeated small changes can themselves add complexity if they are not guided by a coherent direction. Participants’ comments therefore suggested a need for an enduring framework: reforms should be assessed against accessibility, certainty, administrative cost and economic impact, even when introduced separately.

Some participants were skeptical that wholesale reform could be implemented safely or sustainably. Rewriting numerous elements simultaneously would place significant demands on legislative drafters, revenue authorities, software providers, advisers and businesses. It could also create an ‘economic freeze’ as taxpayers delay decisions while waiting to understand the new system.

The discussion therefore pointed toward incremental reform. Better use of third-party data, removal of duplicative reporting, clearer language, periodic review of thresholds and targeted improvements to areas of disproportionate burden may be more achievable than a single comprehensive redesign.

Incrementalism is not without risks. Repeated small changes can themselves add complexity if they are not guided by a coherent direction. Participants’ comments therefore suggested a need for an enduring framework: reforms should be assessed against accessibility, certainty, administrative cost and economic impact, even when introduced separately.

Contributors

• Mukul Asher, Professorial Fellow, Lee Kuan Yew school of Public Policy, National University of Singapore
• Bill Dodwell, Former Tax Director at the Office of Tax Simplification and Editor in Chief, Tax Adviser Magazine
• İrem Güçeri, Associate Professor of Economics and Public Policy, Blavatnik School of Government (Oxford University)
• Daniel Herring, Head of Economic & Fiscal Policy, Centre for Policy Studies
• Peter Jelfs, Professor of Accounting (Professional Practice), Brunel Business School
• Andy Richens, Senior Technical Writer, Croner-i Tax
• Andy Scott, Principal Tax Adviser – Economic Policy, CBI
• Sam Sherwood, PhD candidate, Brunel University
• Sir Edward Troup, former HMRC Permanent Secretary
• Raluca Enache, Head of KPMG’s EU Tax Centre, KPMG in Romania
• Chris Morgan, Head of In-House Tax, KDN
• Conrad Turley, Head of Global Tax Policy, KPMG International
• Becky Holloway, Programme Director, Jericho
• Neal Lawson, Partner, Jericho