Tax Talk - Question and Answer session

Reflections from the Bench 

Dame Susan Glazebrook[1]

 

1          Introductory

Having seen the law from both sides, first as a practitioner advising clients and later as a judge shaping New Zealand's jurisprudence, what are the main changes you noticed once you joined the Bench?

I think the way the question is phrased probably highlights the key differences between being a lawyer and being a judge.

Judges, at least until overturned on appeal, decide cases.  Their decisions settle the dispute between the parties, both on the facts and the law, with real consequences for those parties.  Judicial decisions also have wider consequences.  The legal issues decided will bind those administering the law and judges in lower courts if the later cases cannot be distinguished.[2]  And decisions will be relied on both by those subject to the law and by those advising on the law.  By contrast, advice given as a lawyer will have real consequences for the particular client being advised but no wider effect.[3]

In terms of process, when giving advice a lawyer is reliant on the facts as outlined to them by the client and on the documentation provided.  Of course, lawyers will try to make sure they have a full idea of the facts by questioning the client and pressing them on the existence of other documents.  But, in the end, they are reliant on the client.  And they are reliant on their own experience and research for the giving of legal advice.

Judges, on the other hand, will of course use their own experience and research in coming to their decisions but they have the benefit of having both (and sometimes more) sides of the dispute, both facts and law, argued before them in the course of the adversarial process.  Lawyers can and should try and replicate that when giving advice – in other words try and see all possible sides of any possible argument – but it will never be quite the same as occurs in a courtroom setting.   

When it comes to end result, advice will never be able to be definitive because the lawyer does not have the final say.  Ultimately, the final outcome is for the courts but, if something does not get to court, there are often interim decision makers who will effectively have the final say for the client.

On the other hand, judges do not have the luxury of saying something like “the better view is x but there is a possibility that it may be held to be y.”  Judges are obliged to decide the cases that come before them impartially and according to law.  This is a responsibility taken very seriously. 

I would say, however, that both judges and lawyers have an ultimate responsibility to the law that duty must be respected even when advocating for a client.  Therefore, lawyers and judges in effect work in partnership to uphold the rule of law. 

In terms of the difference between being a lawyer and a judge from my personal perspective, the main one was becoming a generalist again after years of specialising in taxation as a lawyer.  This had its terrors as you can imagine - especially presiding over criminal trials when I had virtually no experience of criminal law in practice.  Luckily the practice of throwing judges in the deep end as soon as they were appointed had stopped.  It was at least three weeks before I found myself presiding over a multi accused high profile criminal trial.

Becoming a generalist again did, however, remind me that the law is a connected whole and that it is dangerous to lose sight of this which can happen if you become too specialised and only interact with other specialists.  

 

What do you wish tax lawyers better understood about judicial decision-making?

One common criticism, not just in the tax field, is that the courts, when a case comes before them, do not take the opportunity to deal with the issues arising in a broader sense, either to clear up uncertainties in the law or to develop the law.  

There is a good reason for this and it relates back to the role of the courts as adjudicators.  It is the role of the courts to decide the cases that come before them.  Courts do not choose their controversies.  They are also limited to the material (at least the factual material) that the parties choose to put before them.

Inevitably the arguments of counsel will be framed in terms of the particular interests of the parties in the particular case.  And that is also true of the judges involved – their consideration of the law will be in the context of the particular facts before them.  Going too far outside of the case to make more general statements about the law has the danger that the law as propounded will not work in other factual circumstances.[4] 

The courts are not of course necessarily limited to the legal arguments put forward by counsel and will sometimes go outside them, subject to issues of natural justice and, in tax cases, subject to any statutory bars.  There may also be fairness issues if arguments are raised for the first time on appeal as the evidence may have been different had they been raised earlier.

One example of the Supreme Court | Te Kōti Mana Nui going outside the way the case had been argued in the courts below is Trustpower.  That case related to whether the expenditure obtaining resource consents for proposed electricity generation projects was deductible.  In its leave judgment the Court said that it could not be artificially constrained by the argument that the only question was whether the consents were “feasibility expenditure”.  The parties had accepted that the then current Interpretation Statement was correct in treating “feasibility expenditure” as being on revenue account.  The Court also said that Trustpower should also clarify what it submitted the taxation treatment of expenditure relating to the consents would be, should one or more of the projects come to fruition.[5]

There was an issue whether the wider issue was able to be decided by the Court because of the way the statements of position had been framed in that case.  In its substantive decision the Court left open whether or not a court could decide a challenge on a basis going outside the propositions advanced in the statements of position because it held that the wider point was in fact within the statements of position.[6] 

 

2          Nature of Tax Law

Turning to taxation law and its specific nature, in your view, is tax law “special” in a way that justifies treating it differently from other areas of law?

In terms of special litigation rules, I think that most of the special rules relating to tax litigation, such as strict time frames and limiting parties to arguments already heralded, are now unnecessary.  This is because the rationale for them has disappeared with modern court procedures, including case management by the courts.[7]  I accept that it is a good idea to encourage the parties to identify at an early stage the arguments that are to be made but this is already required by the new court rules for all cases.  And, if issues limitation also binds the courts, which was left open in Trustpower, then there is a real concern as to the precedential value of a decision where the full legal picture has not been argued or decided.

I think therefore that tax litigation should be treated the same as any other litigation.[8]  This includes taxpayers being able to initiate litigation in the ordinary way at any stage in the disputes resolution process.  Otherwise, it is the Revenue that effectively chooses the cases that come before the courts. 

Not having special rules for tax litigation would mirror the approach to the interpretation of tax statutes.  Up until around the first world war, the interpretation of taxation statutes was treated differently from other statutes (apart from criminal statutes).[9]  In this period, taxation statutes were interpreted in favour of the taxpayer against the state on the premise that taxes are extractions from which the subject should as far as possible be protected.  Following the first world war, tax statutes were construed literally just on the words used with no presumption either way.[10]  The more modern approach, however, is to treat tax statutes like all other statutes and to interpret them in a purposive fashion as in fact directed by the Legislation Act 2019 and its predecessors.[11]

Do you think that the technical nature of tax law means that there that there should be specialist tax judges or a specialist tax court?

One major advantage of a specialist court or specialist judges is a build-up of expertise and therefore greater consistency of decisions.  Specialist judges will also have a deeper understanding of the policy underpinnings and of the wider implications of their decisions.  They will also have an ability to grasp technical detail more rapidly, leading to efficiency gains.

In terms of disadvantages, there has been a concern about the domination of the area by a limited number of judges and therefore a risk of the development of idiosyncratic interpretations of the law and also of judges becoming vulnerable to interest group manipulation.  Over-familiarity with the subject matter can also lead to stagnation, particularly if there is no opportunity for cross-fertilisation from other related areas of law.  It is also suggested that it can be healthy for specialists to have to justify their assumptions, premises and settled ways of thinking to non-specialists. 

Aotearoa | New Zealand has long had the specialist Taxation Review Authority (from 2024 the Taxation and Charities Review Authority).  But the Authority does not hear all first instance tax cases and in recent years there have been so few cases that I doubt that the judges would have been able to build up any expertise, unlike the Authority members when I first started out as a tax lawyer. 

In terms of the High Court | Te Kōti Matua and the appellate courts, I think there is merit in the argument that we should emulate the system in the Federal Court of Australia where judges can opt into particular specialty groups, including taxation and intellectual property.  They then undertake to acquire expertise (if not already expert) and to keep up to date in that area.  First instance cases in specialty areas are allocated to a member of the relevant group.  In the appellate division of the Federal Court it is my understanding that there is an attempt to have at least one specialist judge on any court hearing on appeal. 

It seems to me that such a system within the context of a general court alleviates many of the disadvantages of specialisation.  There will always be the opportunity for cross-fertilisation from other related areas of the law as the specialty group members will be sitting on other cases.  If the group is relatively large with a proper opportunity for ongoing professional development, this should alleviate the risk of stagnation and also concerns as to the perpetuation of idiosyncratic interpretations of the law and interest group capture. 

The High Court | Te Kōti Matua has now gone some way towards the Federal Court system with the new commercial list and commercial panel but has not yet moved towards having specialist groups for particular areas of law.   

I do caution, however, that the benefits of specialisation can be exaggerated.  Rational people do not go to court for their own amusement or by choice.  They do so because either the law, or its application to the particular facts, is unclear.  This means that many of the points that end up in court will be ones that even a specialist will not be familiar with.  Someone with prior specialist knowledge may be quicker to come up to speed than a non-specialist, but come up to speed they must.  Also specialist judges are still judges.  The longer they are on the bench, the further away from the coalface of advising clients they get.  This necessarily affects their perspective.[12]

 

3          Development of Tax Law

One of the foundations of modern New Zealand, tax administration is Sir Ivor Richardson’s Organisation Review Report of 1994.  Do you have any reflections on the effect of that report as it now operates?

First a disclaimer.  Since my appointment as a judge some 26 years ago, I have had no personal experience with Inland Revenue administration, apart from that gleaned through tax cases before the courts and through generally keeping an eye on any major developments. I can therefore only provide very limited comments. 

One of the most important initiatives has been the rewriting of the Income Tax Act 1976.[13]  This project was already underway at the time of the Organisational Review but it was seen as an essential part of the improved processes recommended by the committee.[14]  The rewrite project concentrated on plain English drafting and contained some innovations, such as flow charts to make the Act easier to understand.[15]  But the main benefit was re-organising the Income Tax Act on a conceptual basis and grouping related issues into parts, with a structure that allows amendments to be added seamlessly. An empirical study conducted after the final stage of the rewrite concluded that project had been a success.[16] The fact that the Act was easier to understand must have made it easier to advise clients and also easier to administer.

The Review also led to a new dispute resolution process. To give an idea of the issues that arose before that review, taxpayers often were given no information about the reasons for assessments or amended assessments.  There was inconsistency around the country in the interpretation of the law and in dispute resolution practice.  It was frequently impossible to get anyone other than the inspector involved to look at the matter.  This meant that the sometimes quirky views of individual inspectors could keep a client tied up in a dispute that could easily have been sorted out with an independent look.  Often there was a refusal to meet face-to-face to discuss the issues and correspondence was not answered in a timely fashion, or sometimes at all.[17]  While the structured dispute resolution process has some faults, it has clearly been an improvement on what happened before.    

Another important initiative was a new organisational structure with three main divisions: adjudication, policy and operations.  And I mention also the Generic Tax Policy Process.  But for detailed expert commentary on this and other tax policy and administration issues I refer you to a book edited by Professor Craig Elliffe and others that is due to be released in February next year.[18]

 

Can you comment on the fact that many significant tax issues appear to be resolved through the disputes process, settlement, taxpayer rulings or Inland Revenue guidance, rather than through contested judgments. 

There is no doubt that there has been a dearth in tax cases before the courts in recent years.[19]  There are a number of reasons for this, including some of the initiatives I have just described such as legislation backed by coherent tax policy.  The existence of binding rulings has also played a part, as has the radical improvement in Inland Revenue technical tax bulletins and other publications which educate taxpayers on the law and their obligations.

But another important factor is the combination of the penalty and use of money interest regimes and the elaborate and costly steps involved in the tax dispute resolution regime.  While in many cases the disputes resolution regime has led to better decision making, it is likely also to have “burned off” potential litigants who are forced to enter settlement arrangements or just to accept the outcome of the internal disputes resolution process.

It is true that, at the end of the pre-litigation disputes resolution processes, there is a formal second look at the issues through the Disputes Review Unit.  But however independent the Unit is within the structure of the Inland Revenue, it is still not a court.  It is an internal administrative mechanism without the true independence of the courts (both in actuality and perception) and without the public scrutiny of its decisions in terms of the open justice principle, which is so fundamental to our system of justice.

Access to the courts is particularly important in an area such as taxation which impacts on the relationship between the citizen and the state.  It is also important for the development of the law, which occurs, albeit with caution and incrementally, even within the confines of a statute-based system.  Access to the courts is important most of all for the definitive interpretation of the law which provides certainty and ultimately aids compliance.  Court cases also provide the basis for the disputes resolution process, settlements as well as the rulings regime and Inland Revenue policy statements.

As fewer cases reach the courts, the more the Revenue will be relying on their own interpretation of the law in an echo chamber with no impartial review and adjudication.  This has major implications for the rule of law.

 

4          Tax Avoidance Jurisprudence

The Supreme Court has developed a distinctive approach to tax avoidance in Ben Nevis and subsequent cases. Do you have any reflections on that approach? What is your view of the development of tax avoidance jurisprudence over your career?

I have discussed the history and evolving approaches to tax avoidance in an article published in 2014, as well as analysing the tax avoidance cases in the Supreme Court | Te Kōti Mana Nui.[20]  Just to recap, I saw the approach to tax avoidance as an exercise in statutory interpretation and essentially the same exercise as reconciling any specific provisions in any statute with general open textured provisions.

In Ben Nevis the majority held that the taxpayer first has to satisfy the court that the use made of the specific provision was within its intended scope.  If that is shown, the second stage of the inquiry is whether the use of the specific provision has altered the incidence of income tax in a way that was not within the contemplation and purpose of Parliament.  There is no limit to the factors that can be taken into account at this second stage and these can include considerations of artificiality, circularity, commercial reality and economic effect.[21]

I discussed in that article the concerns that had been expressed by commentators that the Parliamentary contemplation test may simply allow a judge to substitute the judge’s own view of the arrangement under the guise of answering the question of Parliamentary contemplation and purpose.  The, at least partial, answer to this concern was that the second stage of the inquiry must be grounded in the words of the statute interpreted in the context of the Act as a whole and any relevant wider context.  And that it is not the role of the courts to fill the policy gaps left by Parliament or to substitute their own view of what should be the policy.  The role is statutory interpretation and not one of rewriting the legislation.

I also discussed in the article the extent to which the Ben Nevis test was in fact a new approach and concluded that it did not signal a complete shift in how the courts approach tax avoidance but that it did firmly endorse an economic substance approach as against the legal substance approach that had been used in some earlier tax avoidance jurisprudence. 

The latest Supreme Court | Te Kōti Mana Nui case on tax avoidance was Frucor.[22]  I dissented in that case.  My judgment speaks for itself and it is not appropriate that I comment further.[23]

 

5          Legislative Drafting

It is probably fair to say that modern New Zealand tax legislation has become increasingly detailed and complex. What are your thoughts on the current legislative drafting approach (both tax and non-tax)? Do you think modern tax statutes are still fit for purpose from a judicial interpretation perspective, or has the level of detail made the law harder rather than easier to apply?

In terms of legislation generally, there has in recent years been a concerted effort to make statutes more understandable and accessible. The modern approach to drafting is based on three core principles: that legislation should be fit for purpose (effectively based on sound policy), constitutionally sound (for example consistent with fundamental values and principles) and accessible for users (easily understand and easily found).[24]  The rewrite process in relation to taxation legislation was essentially based on the same principles and, as I have already said, has been deemed a success.

Since the question is posed in the way it is, I am assuming there is a general perception that tax legislation has become more complicated and difficult to understand since 2007.  If this has occurred, it could have arisen by a failure at the first stage (for example piecemeal reforms that have not been fully thought through in overall policy terms) or a failure at the third stage in terms of the style of drafting.  

Tax statutes are drafted by Inland Revenue | Te Tari Taake and not the specialist Parliamentary Counsel Office (PCO) which drafts all other legislation. [25]  This can be seen as justified because of the technical nature of tax statutes.  On the other hand, the PCO has drafting expertise and an overview of the whole statute book.  In 1999 a select committee recommended that drafting of tax statutes be transferred back to the PCO.[26]  A subsequent independent review of tax law drafting recommended that Inland Revenue retain responsibility for drafting bills but with increased use of PCO training and resources.[27]  A compromise solution therefore.

If indeed unwarranted complication has started to creep back into revenue statutes since the 2007 Act was passed, then this might possibly be the time to consider again whether drafting responsibilities should be transferred back to the PCO.[28]

 

6          Artificial Intelligence

Artificial intelligence is extremely topical. What role do you think AI is likely to play in tax law over the next decade — particularly in interpreting complex legislation, assisting with disputes, and shaping how taxpayers and advisers understand Inland Revenue guidance?

For advisors, there is no doubt that the generative AI tools currently available can lead to major efficiency gains and particularly in managing large quantities of data.  Empirical research conducted by Professor Daniel Schwarcz, who delivered the Borrin Lecture[29] this year, has shown that human lawyers with access to AI tools perform better than those without such access.[30]  There is also no doubt that the tools will continue to develop and that some of the issues with them, such as the infamous hallucinations, will diminish if not altogether disappear.  

There are guidelines on generative AI issued by the New Zealand Law Society for lawyers[31] and also court guidelines on the use of generative AI in courts and tribunals.[32] While couched as guidelines, lawyers would be well advised to adhere to them.  Two of the most important aspects relate to confidentiality and ensuring and checking accuracy of any AI generated product.

This highlights an important aspect stressed by Professor Schwarcz in his lecture: the need for qualified and experienced humans, both for checking the output but also for ensuring the quality of inputs (to avoid the classic “rubbish in, rubbish out” scenario).[33]  This highlights the need for training of lawyers to work with AI.  I would suggest too that firms would be very short sighted indeed if they neglected the recruitment and training of junior lawyers.  Otherwise, they will not develop the skills necessary for being the future guardians of the rule of law.    

Turning now to taxpayers, AI certainly has the potential to improve taxpayer access to information on taxation.  The problem is that the tools available to the general public are prone to error, either because of hallucinations or because the taxpayer is not sufficiently skilled to enter the right inputs. You will no doubt have had the experience of taxpayers coming to you for advice armed with pages and pages of misleading information where you have to spend as much, if not more, time explaining why the material is wrong as you do giving them the proper advice.  There is, however, a real potential for well-designed consumer self-help tools to be developed.[34]      

 

Does the increasing use of AI by government agencies create any rule of law concerns?

The use of AI by government agencies, including Inland Revenue, has the potential to assist in improving productivity and efficiency and especially where there is a vast amount of information and data involved.[35]  Generally, the use of AI by government agencies only causes rule of law concerns when it is used for automated decision making. 

There are some spectacular instances of technology failures which have caused real injustice. Two examples are the UK Post Office Horizon scandal which resulted in hundreds of sub-postmasters being wrongly prosecuted.[36]  And the so-called Robodebt scandal in Australia, which concerned a compliance and debt recovery programme to claw back supposed overpayments to welfare recipients.[37]  It used a process known as “income averaging” to assess income and entitlement to benefits.  This not only did not produce accurate results but it did not comply with the relevant legislation.[38]  It was not the only aspect of the scheme that did not comply with the law.  The scheme was also unfair in that the onus was on the welfare recipient to show the debt was not owed and there were attempts at a “cover-up” lasting some years.[39]  The Royal Commission that reported on the scheme concluded in very strong terms that:[40]

Robodebt was a crude and cruel mechanism, neither fair nor legal, and it made many people feel like criminals.  In essence, people were traumatised on the off-chance they might owe money. It was a costly failure of public administration, in both human and economic terms.

An example going the other way is Operation Protego, a scam that occurred in 2022 and 2023 in Australia.  It was promoted on social media and involved fraudulent GST claims from non-existent businesses. This resulted in an estimated $2 billion of false claims being paid out before the Australia Tax Office (ATO) shut down the loophole.  It appears that the problem was a combination of automatic refund payments and flaws in the fraud detection processes.[41]

In line with OECD guidelines, Aotearoa | New Zealand has developed various guidelines on the use of AI in the public sector.  There is a general framework on the use of AI, which is broadly defined, in the public sector.[42]  The overall vision is to: “Adopt AI responsibly to modernise public services and deliver better outcomes for all New Zealanders”.  There are then five key principles including human-centred values, transparency and accountability.  There are also specific guidelines with regard to generative AI with similar principles, and including security, transparency, lack of bias and privacy.[43]  l mention also the Algorithm Charter developed by Statistics NZ (Inland Revenue is one of the signatories).  This charter includes commitments about transparency (meaning that it is explained how decisions are informed by algorithms), a focus on people, to respect privacy and human rights and that human oversight is maintained. 

Some commentators have suggested that non-binding guidance of this type is insufficient to manage the risks.[44]  I do note that the Australian National Audit Office is responsible for overseeing and evaluating the responsible use and governance of AI across Australian federal government agencies and that an audit of the ATO’s AI governance was conducted in 2025.[45]  I do not understand there to be a similar audit programme specifically focused on AI use in Aotearoa | New Zealand.  

 

7          Future of Tax Law

Putting aside AI, what developments do you think will pose the greatest challenges for tax law over the next decade?

Judges, in their core work of deciding cases, are normally examining past events.  They do have to keep in mind the wider implications of their decisions but predicting the future is not one of their core skills.

Without therefore claiming any particular expertise, I would suggest the following: the blurring of national boundaries in business but also increasing protectionism, digital business models and the gig economy.  Increasing and evolving regulatory requirements are also likely.  Indeed, this was the major concern that arose in the 2025 Deloitte survey of tax professionals.[46] 

On a wider scale, we have we have geopolitical conflict, climate change, including extreme weather events, and possible future pandemics.  We live in what is called a NAVI world, one characterised by constant disruption.  Businesses, including tax professionals, need to plan for such a world, which means planning for a wide range of scenarios.  And it is important to treat this as creating opportunities as well as challenges.[47] 

 

8          Final Question

Looking back over your career, what are two or three of your favourite tax judgments, from New Zealand or overseas, that you would recommend attendees read or re-read?

I know a few people who read judgments for pleasure but I have never been able to understand that urge.  Give me a good detective novel any day or, if I am feeling intellectual, something from the Booker long list.   

I am very much in the camp of read what you need when faced with a particular issue – you get much more out of reading a judgment when you are trying to work out whether and how it applies to a particular set of facts.

That does not mean that you should not keep abreast of developments in the law and in tax policy.  And not just tax law as other areas of law will be relevant to any tax issue.  And not just in New Zealand as international developments are also important.  But the sheer volume of material will mean that usually you will be relying on headnotes, summaries or conferences to keep up to date rather than reading judgments. 

 

Footnotes