TUUP Appellant

and

THE COMMISSIONER FOR THE Respondent SOUTH AFRICAN REVENUE SERVICE
Coram: Bishop, AJ (Prof P Surtees and Mr T Ledwaba, assessors)

Dates of Hearing: 23 and 25 October 2023

Date of Judgment: 17 January 2024

J U D G M E N T

BISHOP, AJ
VAT vendors are entitled to deduct from their output tax liability to SARS the input tax on goods and services they acquire for making taxable supplies. An advocate can claim the VAT added to his expenditure on renting chambers or buying paper as a deduction on the VAT he owes. For purely commercial enterprises this presents little difficulty – all the VAT paid on goods and services for the business is deductible.

But the Value-Added Tax Act 89 of 1991 does not permit the deduction of input tax on goods and services acquired for non-taxable supplies. VAT paid on goods and services to make “exempt supplies” is not deductible. The reasoning is that the supplier of exempt supplies should be the ultimate bearer of the VAT costs, rather than the person who receives the exempt supplies. Under section 12(h)(i), the supply of “educational services” is one such exempt supply. So is the supply of board and lodging to students.1

Some VAT vendors provide only taxable supplies, and some provide only exempt supplies. But others provide both taxable and exempt supplies. Universities are one example. Their primary business is to provide education and related services to students. But they also conduct commercial research and make other taxable supplies. Often the same goods and services are used for both – the same printer is used to print educational materials and research materials; the same leased building is used for lectures and commercial experiments. How does a university determine how much of the VAT on its inputs it is entitled to deduct from its VAT liability?

Through section 17(1) of the VAT Act. It empowers the Respondent (SARS) to determine the “ratio” of goods and services a vendor intends to use for taxable compared to other uses. The question is not what percentage of the business, the turnover, or the profit is for taxable or supplies, but what ratio of its inputs the vendor uses to make its taxable versus non-taxable supplies.

‌1 VAT Act section 12(h)(ii).

In 2011, Higher Education South Africa (HESA) approached SARS for a section 17 class ruling2 for all universities, including the Taxpayer. SARS granted the requested ruling (the VCR). It separates universities’ research activities, allowing them to claim VAT inputs at different rates for different types of research. For all other expenses, it creates a formula for calculating the ratio at which universities can claim VAT input deductions. It then sets a cap: no matter the outcome of the formula, universities cannot deduct more than 12.5% of the VAT they spent on goods and services. The VCR was agreed between HESA and SARS.

The Taxpayer is before this Court because it is unhappy with SARS’s decision not to permit a deviation from that class ruling. Back in 2009, before it agreed to the VCR, it concluded two agreements with the Developer – a Head Lease and a Sub-Lease. The Taxpayer leased land to the Developer, which was then obliged to build a student residence on the land. The Taxpayer would then lease the land back from the Developer in order to use the improvements as a residence for its students.

The issue in this application is how, if at all, that agreement affects the Taxpayer’s ability to claim VAT input deductions. The Taxpayer argues that its obligations under the Head Lease are taxable supplies and it should be entitled to claim the VAT on those obligations as input tax deductions. SARS contends that the agreement as a whole is for the purpose of providing the student residence – an exempt supply – and therefore the Taxpayer is not entitled to any input tax deduction.

‌2 SARS can issue rulings to a class of vendors under section 41B of the VAT Act. Section 41B reads:

“(1) The Commissioner may issue a VAT class ruling or a VAT ruling and in applying the provisions of Chapter 7 of the Tax Administration Act, a VAT class ruling or a VAT ruling must be dealt with as if it were a binding class ruling or a binding private ruling, respectively: Provided that-

the provisions of sections 79(4)(f), (k), (6) and 81(1)(b) of the Tax Administration Act shall not apply to any VAT class ruling or VAT ruling;

an application for a VAT class ruling or a VAT ruling in terms of this section shall not be accepted by the Commissioner if the application-

is for an advance tax ruling that qualifies for acceptance in terms of Chapter 7 of the Tax Administration Act; and

falls within a category of rulings prescribed by the Minister by regulation for which applications for rulings in terms of this section may not be accepted.”

In addition to this central dispute, the Taxpayer raises another complaint – is the 12.5% cap in the VCR lawful? It takes the view that SARS cannot impose caps on apportionment ratios under section 17(1), but only apply formulae or methodologies and must live with the results. SARS disagrees. It contends that the Taxpayer has not properly pleaded its attack on the VCR and that there is, in any event, nothing wrong with the 12.5% cap which fairly reflects how universities operate.

I conclude that SARS is right on all counts. The Taxpayer did not properly plead an attack on the VCR. Even if it had, the attack is bad. The purpose of the dual agreements with the Developer was the provision of exempt supplies, and the Taxpayer is not entitled to any input tax deduction. Even if it were not, accounting for the Head Lease in the section 17(1) apportionment ratio would have anomalous results inconsistent with the purpose of the provision.

I begin by describing the factual and legal building blocks for the case: section 17(1); the VCR; the Head Lease and Sub-Lease; and the Taxpayer’s application. I then deal with the challenge to the VCR. Finally, I explain why the Taxpayer is not entitled to a departure from the VCR.

Section 17(1)
The VAT Act permits VAT vendors to deduct “input tax” from their VAT liability, which it defines to include:

“(a) tax charged under section 7 and payable in terms of that section by-

(i) a supplier on the supply of goods or services made by that supplier to the vendor; or

where the goods or services concerned are acquired by the vendor wholly for the purpose of consumption, use or supply in the course of making taxable supplies or, where the goods or services are acquired by the vendor partly for such purpose, to the extent (as determined in accordance with the provisions of section 17) that the goods or services concerned are acquired by the vendor for such purpose.”

Section 17 is the provision that allows SARS to determine “the extent … that goods or services are acquired” “in the course of making taxable supplies”. Without the currently irrelevant provisos, and with the core part underlined, section 17(1) reads:

“Where goods or services are acquired or imported by a vendor partly for consumption, use or supply (hereinafter referred to as the intended use) in the course of making taxable supplies

and partly for another intended use, the extent to which any tax which has become payable in respect of the supply to the vendor or the importation by the vendor, as the case may be, of such goods or services or in respect of such goods under section 7(3) or any amount determined in accordance with paragraph (b) or (c) of the definition of ‘input tax’ in section 1, is input tax, shall be an amount which bears to the full amount of such tax or amount, as the case may be, the same ratio (as determined by the Commissioner in accordance with a ruling as contemplated in Chapter 7 of the Tax Administration Act or section 41B) as the intended use of such goods or services in the course of making taxable supplies bears to the total intended use of such goods or services.”

Section 17 requires the determination of a “ratio” of input tax that is deductible. That ratio is the relationship that “the intended use of such goods or services in the course of making taxable supplies bears to the total intended use of such goods or services”. If a vendor uses 50% of all the goods and services it acquires to make taxable supplies, and 50% to make exempt supplies, the ratio is 1:2 or, more colloquially, 50%. For every R1 it spent on those goods and services, it would be entitled to deduct 50c from its VAT liability.

Section 17(1) allows SARS to determine the ratio, and expressly permits a class ruling under section 41B. Section 17(1) does not dictate to SARS how it must calculate the ratio. It does not require it to use any particular formula or methodology. And it plainly envisages that the method will differ between vendors and between classes of vendors.

In practice, SARS has issued Binding General Ruling 16 that requires vendors to use the standard turnover method to calculate the apportionment ratio under section 17(1). Without a specific ruling for a class or a vendor under section 17(1), they must use that method. BGR16 provides that the method only applies if it is “fair and reasonable to the vendor’s business activities.” The vendor must determine if that is the case. If the vendor determines the method “is not fair and reasonable, it is the vendor’s further responsibility to approach the Commissioner for an alternative method.”

The VCR
Until 2011, universities followed the standard turnover method, or obtained their own individual ruling from SARS for an alternative method. In February 2011, HESA applied to SARS for a class ruling to apply the varied input based method for all public universities. It argued that the standard turnover method did not accurately reflect the way in which universities used goods and services for taxable and exempt supplies. There was also inconsistency between universities as some had more favourable VAT rulings than others. HESA wished to resolve both problems. The Taxpayer was a member of HESA throughout.

SARS agreed, and engaged with HESA and the universities to make a new section 17(1) ruling that would apply to public universities as a class. SARS led the evidence of Mr Rain who headed SARS’s team that considered HESA’s application. There were three important elements to Mr Rain’s testimony which were supported by the underlying documents.

Universities receive most of their income from student fees and government grants. The primary “third stream” of income is research. Universities receive private or government contracts to conduct research, or recover money from selling the results of their research. There are some universities that conduct far more research than others. One of the goals of the class ruling was to fairly treat both types of universities.

SARS determined that, given the importance of research, it should be treated separately from universities’ other income-generating activities. It investigated the different ways in which universities conduct research and divided them into different categories depending on the extent to which it was for educational or commercial purposes:

[19.1] Basic research was entirely for education purposes and no input VAT could be deducted.

[19.2] Applied research could serve both educational and commercial purposes and would receive a 50% deduction.

[19.3] For contract research, it would depend on the involvement of students. If they were involved, the research partially served an education purpose and a 50% deduction would be allowed. If no students participated, the research was 100% for making taxable supplies and all input VAT could be deducted.

Having removed the primary third stream of income, SARS then determined what remaining percentage of universities’ goods and services were used for other taxable supplies. For example, some universities rented residences during holidays, or rented out lecture halls for private events.

After much debate about which formula to employ, SARS and HESA settled on the varied input based method which uses the following formula:

y

=

  a    

×

100  

(a+b)

1

In the formula, y is the apportionment percentage, a is VAT incurred wholly for purposes of taxable supplies, and b is VAT incurred wholly for purposes of exempt and non- supplies.

SARS decided to focus on the operation of non-research universities (like the Taxpayer) to avoid double counting the impact of research. The question was what ratio of universities’ general overheads did they use for taxable versus exempt supplies. SARS went through the cost centres of a sample of universities. They were trying to identify what expenses, excluding research, were for the purpose of taxable supplies. The investigation demonstrated that the majority of universities used less than 10% of their overheads for making taxable supplies. After discounting for anomalies, no non-research university used more than 12.5% of its overheads for taxable supplies.

The anomalies included, for example, a university that owned but did not manage a shopping centre. It did not influence the general overheads of the university, and so had to be excluded from the calculation. Including it in the calculation would distort the ratio by massively increasing a in the formula, even though the University dedicated little to no resources, and therefore incurred little to no VAT, in managing the centre. The consequence of the distortion would be that the university would claim a higher percentage of input VAT deductions on all its expenses than it in truth used to make taxable supplies.

SARS and HESA therefore agreed to a 12.5% cap which was “more taxpayer friendly” because it was the highest ratio any university had demonstrated on the varied input method. It also equated to the highest percentage non-research universities achieved on the varied turnover-based method. The cap served two purposes. It ensured that high value transactions that did not affect universities’ overheads – but would significantly increase a and therefore y in the formula – could not have an outsized impact. It also ensured a degree of fairness between universities, and ensured no university had an outsized competitive advantage by being able to deduct VAT at a much higher rate.

A final wrinkle to mention is that the VCR excluded certain expenditure. Specifically, it excluded “expenditure on any capital goods or services acquired, unless acquired under a rental agreement/operating lease.” I mention this because there is a debate between the parties about the application of this exclusion to the Head Lease.

SARS’s investigations to make the ultimate ruling took 18 months. Mr Rain’s evidence that the process was “thorough” was undisputed. The Taxpayer admits that it “provided some input in the process … when requested for information”, but claims “it was never directly

involved in offering a suitable basis for apportionment for itself.” The VCR was first issued on 31 July 2012, and errors were corrected on 1 August 2012. The ruling has been renewed on the same terms regularly since then, and remains in force.

While it is a class ruling under section 41B, it remains possible for any taxpayer to request a separate, individual ruling if it believes that the application of the VCR is not “fair and reasonable”. As I explain below, that is what the Taxpayer did.

The Taxpayer claims that the VCR is more beneficial to research universities than it is to it. Mr Rain disputed this. He argued that accounting for research separately should mean that the universities were treated equitably. The cap ensured that research universities did not benefit unduly. The Taxpayer also argued the VCR was a compromise and was meant to be temporary. Again, Mr Rain demurred. He said SARS would not have spent 18 months to make a temporary ruling, and that the VCR had been regularly renewed on the same terms and conditions. As SARS introduced evidence, and the Taxpayer did not, and as I found Mr Rain an impressive witness, I accept his testimony on these issues.

The Head Lease and the Sub-Lease
The Head Lease and the Sub-Lease were concluded on the same day – 4 December 2009. They were – as I explain in more detail later – conditional on each other. Each would come into operation only on the coming into force of the other (in addition to various other conditions).

Under the Head Lease, the Taxpayer leased property to the Developer for 20 years. The consideration for the lease of the property was twofold. The Developer would construct a student residence worth at least R84.5 million (defined as “the Improvements”), and it would pay the Taxpayer R25 000 (excl VAT) in rent each year. The Taxpayer was required to make an initial payment of R19.220 million three days after the construction, described as a “contribution to the Improvements”.

The Sub-Lease then required the Developer to lease the property – with the residence it had built – back to the Taxpayer for fifteen years. The Taxpayer would pay the Developer an escalating rental starting at R877 340 in March 2011, and ending at R2 493 209 at the Sub- Lease’s end in 2026 (both VAT inclusive). It would also be responsible for interior maintenance, and pay the Developer operating charges to maintain the exterior and provide security.

The relationship is a financing arrangement. The Developer makes its profit on the construction of the residence under the Head Lease through the rental paid over 15 years under the Sub-Lease.

The Taxpayer would be responsible for managing the Sub-Lease, which would require substantial expenditure. But it accepts that all expenditure under the Sub-Lease is for the provision of exempt supplies. By contrast, the Taxpayer accepts it incurs negligible expenditure to manage its obligations under the Head Lease.

The Ruling Application
In 2019, the Taxpayer applied for a ruling under section 41B. It argued that the VAT incurred in acquiring the supplies under the Head Lease should be included in the calculation of its apportionment, which would then exceed the 12.5% cap, which SARS should allow.

In essence, its argument was that its supply of the property to the Developer under the Head Lease was a taxable supply. The consideration was the annual rental, and the construction services. The Head Lease was a barter transaction – construction services for possession of the property. The rental the Taxpayer paid under the Sub-Lease reflected the market value of the improved property for both agreements and therefore the value for which it had to account for output tax for its supply under the Head Lease. The Taxpayer acquired the leasehold improvements (the residence) not to provide accommodation to its students as you might think, but to enable it to make the taxable supply to the Developer under the Sub- Lease. Accordingly, the VAT on the improvements, which is equal to the VAT of the rentals it pays under the Sub-Lease, must be accounted for as input tax and included in the apportionment formula under the VCR. As this would increase the ratio under the varied input formula above 12.5%, that cap should be lifted to cater for the Taxpayer’s special circumstances.

If you find this logic confusing and circular, you are not alone. It imagines away the Sub-Lease other than for the purpose of calculating the value of the Taxpayer’s supply of the improved property (which the Developer has built, and can only use to provide a student residence to the Taxpayer under the Sub-Lease). But that was the Taxpayer’s argument.

Importantly, the Taxpayer did not argue that the 12.5% cap was inherently unlawful. It did not challenge the legality of the VCR. It accepted it as a given, but asked for an exemption in its special circumstances. It asked to add the impact of the expenditure under the Head Lease to either the apportionment ratio calculated under the VCR, or 12.5%, whichever was

the lower. So it accepted that, but for its particular complaint about the Head Lease, the 12.5% cap could be applied, even to it.

SARS decided not to grant the ruling. The Taxpayer objected. SARS dismissed the objection. The Taxpayer appealed, and so the dispute found its way to this Court.

SARS’s reasons for not granting the ruling are threefold. First, it argues that the Head Lease and the Sub-Lease are intertwined transactions and their joint purpose is the provision of exempt supplies. Second, it contends that, even if the supplies under the Head Lease are taxable, including them would distort the apportionment ratio. Third, it argues that the supplies are capital expenditure not under a rental agreement, and are therefore excluded by the VCR.

Throughout this process, the dispute remained one primarily about whether and how the Head Lease should be factored into the Taxpayer’s section 17(1) apportionment ratio. As I detail below, the Taxpayer did complain about the VCR. But it never asked for relief declaring it invalid. The relief the Taxpayer sought in both the objection and the appeal were to grant the ruling it sought in its initial section 41B application.

The Hearing
This matter was initially set down for five days. On the first day, the Taxpayer indicated that it did not intend to call any witnesses. SARS called only Mr Rain, who testified about how and why the VCR was adopted.

The Taxpayer then submitted its heads of argument which challenged not only the refusal of the ruling application, but the validity of the 12.5% cap. The Taxpayer argued that it was irrational and ultra vires for SARS to impose a cap on the apportionment ratio that universities could claim under section 17(1). It also persisted with its arguments that the Head Lease should be factored into its ratio, which should be allowed to exceed 12.5%.

SARS argued that the Taxpayer had not properly pleaded its attack on the 12.5% cap and ought not to be permitted to advance it. It argued that, in any event, the attack was bad in law. It also persisted with its three objections to the initial ruling application.

There were a number of issues that, depending which way they were decided, would be determinative of the application one way or the other. I asked counsel if I should decide only those necessary to resolve the dispute, or if I should express an opinion on all the issues. Mr Abba SC cautioned minimalism; Mr Ben SC encouraged me to address everything. Of the five issues – pleading, ultra vires, exempt supplies, distortion and capital expenditure – I address all except capital expenditure.

The Validity of the 12.5% Cap
The first question is whether the 12.5% cap is lawful. That raises two distinct questions: Was the legality of the 12.5% cap properly put in issue in the pleadings? If it was, has the Taxpayer laid a basis for reviewing and setting it aside?

Did the Taxpayer plead an attack on the 12.5% cap?

Before we can consider whether SARS has the power to impose the 12.5% cap, we must decide whether the issue is properly before us. The dispute before us, like all legal disputes, “must be resolved on the issues raised by the parties”.3 Those issues are defined by the pleadings – the Rule 31, 32 and 33 Statements – and “parties will be kept to their pleadings, where any departure from the pleadings would cause prejudice or prevent a full enquiry.”4 At the same time, “pleadings are made for the court and not the court for pleadings.”5 Where an issue is fully ventilated without prejudice to the parties, then the Court should not avoid deciding a question because “the pleadings had not been as explicit as they might have been.”6

We must therefore sail between the Scylla of unfairly departing from the pleadings, and the Charybdis of using the pleadings as an excuse not to determine a dispute properly raised and ventilated.

Mindful of those two concerns, there are five reasons this Court should not decide the ultra vires challenge.

First, the relief the Taxpayer sought is inconsistent with a finding that the 12.5% cap is unlawful. In paragraph 104 of its Rule 32 Statement, the relief the Taxpayer sought was to uphold its appeal and altering SARS’s decision “to one approving the Ruling Application.”

What did the Taxpayer seek in its Ruling Application? Under the heading “Request”, the Taxpayer asked SARS for the following:

“Kindly confirm by way of a binding ruling in terms of section 41B of the Act, that [the Taxpayer] may add the additional percentage resulting from the inclusion of its leasehold improvement supplies in its apportionment formula to the lower of its

actual VAT apportionment ratio as calculated in terms of the VCR; or

‌3 Africa Cash and Carry v Commissioner, South African Revenue Service [2019] ZASCA 148; [2020] 1 All SA 1 (SCA); 2020 (2) SA 19 (SCA); 82 SATC 73 at para 53.

‌4 Ibid.

‌5 Ibid.

‌6 Ibid.

the 12.5% cap as provided for in the VCR.”

The Ruling Application accepted the validity of the 12.5% cap. It relied on it as a mechanism to calculate the apportionment formula. The Rule 32 Statement refers back to the Ruling Application to define the relief sought. The finding that the Taxpayer now seeks – that the 12.5% cap is unlawful – is inconsistent with its own relief.

Second, the 12.5% cap is contained in the VCR and the Taxpayer never asked for a finding that the VCR itself was in any way unlawful. Its application and its appeal were always structured as a request for a departure from the VCR, the validity of which was otherwise assumed. It is an ordinary rule of pleading that if you wish to challenge the validity of an administrative act, you must identify the decision you are attacking, the facts upon which the cause of action is based, and the legal basis for the review.7

‌The closest the Taxpayer comes to foregrounding the attack in its pleadings is the following excerpt from its Rule 32 Statement:

“In terms of section 17(1), SARS has the (delegated) power to issue an apportionment ruling determining the methodology to be applied by the vendor. SARS does not … have the (delegated) authority to impose an arbitrary numerical limit on the result such methodology renders. Imposing such a limit is ultra vires the powers of SARS.”

It is one thing to complain that numerical limits are arbitrary and ultra vires. It is quite another to seek relief declaring those numerical limits invalid for that reason. Despite alleging that the 12.5% cap is “arbitrary” and “ultra vires”, the Taxpayer never sought an order that it should be declared unlawful. The relief it did seek assumed its validity. In that context, this passage in the Rule 32 Statement cannot be interpreted as introducing a challenge to the VCR. It is introduced as an aside to the main challenge – the refusal to grant a departure from the VCR.

Third, multiple statements in its Ruling Application positively indicate that while it did not like the 12.5% cap, the Taxpayer did not challenge its validity:

[56.1] “Although we note that the imposition of an arbitrary cap on the apportionment ratio resulting from the method approved by SARS in terms of section 17(1) of the VAT Act, we do not seek to address this issue as part of this particular application.”

‌7 See, for example, Bato Star Fishing (Pty) Ltd v Minister of Environmental Affairs and Tourism and Others [2004] ZACC 15; 2004 (4) SA 490 (CC); 2004 (7) BCLR 687 (CC) at para 27.

[56.2] “For the purposes of this application, we accept that the TUUP VAT apportionment ratio would be capped to the agreed 12.5%.” What the Taxpayer asked for was that it should be allowed to claim a higher apportionment ratio "despite the 12.5% cap”. That accepts the cap is otherwise lawful.

[56.3] It calculated the apportionment ratio it sought using the 12.5% cap. It argued that if its ratio (excluding the leasehold transaction) was 16%, “the 12.5% cap would apply”. If including the leasehold transaction added an additional 5%, that would be added to the 12.5%, not to the 16%. This again accepts the general validity of the cap.

The ruling application is the foundation of the dispute. It may be that the Taxpayer was entitled to depart from the ruling application in its objection or its Rule 32 Statement. But then it must do so expressly so as to alert SARS that it is making a different case, and challenging a different, additional decision. It did not do so.

Fourth, the validity of the 12.5% cap features nowhere in SARS’s Rule 31 Statement. As its validity was not questioned in the application or the appeal, that makes sense. In responding to the passage quoted in paragraph [54] in its Rule 33 Statement, SARS baldly denied that the apportionment ratios were ultra vires.8 That is understandable because it could not have known this would become a lynchpin of the Taxpayer’s case. How could it when the Taxpayer never challenged the VCR and sought relief relying on the 12.5% cap?

The Taxpayer led no evidence, so SARS was still not aware that the Taxpayer was directly attacking the validity of the VCR when it led its evidence. Mr Rain did give some evidence that was relevant to justifying the cap. But it seems at least possible that, had SARS known the cap was directly in issue – rather than merely berated in passing – it would have advanced further evidence to defend it.

I do not accept that this is a purely legal issue as the Taxpayer argued where it could be raised at any point without prejudice. As I explain below, a percentage is a “ratio”, so there is no simple linguistic reason the cap is unlawful. Whether section 17(1) permits this particular type of ratio is a mixed question of fact and law. It depends on whether caps are necessary to achieve the purpose of section 17(1). The answer to that question depends, at least in part, on evidence of how taxpayers generally (and universities specifically) use supplies for mixed

‌8 It also pointed out that, if they were, the result would be that the Taxpayer would be required to apportion its mixed-purpose inputs under the standard turnover-based model.

purposes and whether formulae alone can accurately capture the extent to which those supplies are used for exempt or non-taxable supplies.

The result is that SARS has not properly defended the 12.5% cap because it was only placed on notice that it was required to do so when the Taxpayer submitted its heads of argument. That is prejudice. It would be unfair to decide the issue when SARS has not been properly alerted to the challenge.

Fifth, the VCR was granted at the request, and with the agreement, of HESA, and applies to all universities. A declaration that the 12.5% cap is unlawful would affect all universities’ VAT liability. While removing the cap could only increase the amount of input tax they could deduct, one of the reasons SARS proffers for the cap is fairness between universities. Whether that is a valid concern that can justify the cap is an important question. Those universities, whose section 17(1) VAT ruling will be altered if this Court declares the 12.5% cap unlawful, have not been notified, or afforded an opportunity to be heard. Nor has HESA.

To be clear, I do not hold that the Court is precluded from considering the 12.5% cap in the absence of all the universities. The issue was not argued. But the broader impact of the relief the Taxpayer seeks on the entire class, without clearly identifying the issue in its pleadings, cautions against allowing it to advance the argument.

In conclusion, it is not open to the Taxpayer to now seek to attack the 12.5% cap as being inherently ultra vires. SARS was denied a fair opportunity to defend it, and this Court does not have all the information before it to decide the question. Nonetheless, in case I am wrong on that issue, I consider whether the 12.5% cap is permitted by section 17(1).

The 12.5% cap is a lawful ratio

The Taxpayer’s argument that section 17(1) does not authorize the 12.5% cap rests on two inter-related claims: it is not a “ratio”; and SARS is obliged to apportion according to the actual ratio of the intended use of goods and services for making taxable supplies without any artificial cap.

I do not agree with the first argument that a percentage is not a “ratio”, or that SARS is limited to adopting formulae or methods. 12.5% is “the quantitative relation between two amounts showing the number of times one value contains or is contained within the other”.9 12.5% is just a different way of writing 12.5:100.

There is nothing in section 17(1) that says SARS must use a formula, and cannot specify a percentage. Nor is there any reason to interpret “ratio” to mean “formula”. I agree with SARS that it should have the flexibility to use formulae, methodologies, or directly impose ratios. What will be appropriate will depend on the nature of the vendor and its business. As long as the outcome reflects the relationship of the vendor’s use of taxable and mixed supplies, it is permissible. Interpreting section 17(1) to limit SARS’s discretion to calculate apportionment does not serve any identifiable purpose of the provision.

In any event, any formula SARS adopts under section 17(1) must ultimately produce a percentage or a ratio. The formula is the means to the end. But the end will always be a ratio, not a formula. That is the case here. The VCR states that “[i]n calculating the apportionment ratio” universities must use the approved methodology. The methodology defines “y” as “the apportionment percentage”. That percentage is what ultimately determines a university’s VAT liability.

SARS is entitled by section 17(1) to determine that ratio in any manner that reflects “the intended use of such goods or services in the course of making taxable supplies bears to the total intended use of such goods or services”. That is what it did here – the ratio of 12.5% is the outcome of the method it used to investigate universities, as described by Mr Rain.

The Taxpayer’s second, and more compelling, argument is that the 12.5% cap is arbitrary because, even if a university apportions more than 12.5% of the goods and services it acquires to make further taxable supplies, under the VCR it cannot claim the full input VAT to which it is entitled. That is true. But to my mind it misses the point.

First, the cap is an administrative tool, not immutable dogma. The universities (including the Taxpayer) approached SARS through HESA and asked for a class ruling under section 17(1). Nobody forced them to do that. They asked for it. Having been asked for a class ruling, SARS investigated how universities use their goods and services to provide both exempt and non-exempt supplies. It came up with a detailed, agreed method that includes the cap.

‌9 Concise Oxford English Dictionary (12 ed, 2011).

The 12.5% cap has the effect of a default rule. It is the maximum apportionment universities are allowed to claim without a separate ruling under section 41B. If a university can demonstrate to SARS that they are in fact using more than 12.5% of their overheads to make taxable supplies, and that the 12.5% is therefore not “fair and reasonable” then SARS will permit it to claim a higher ratio. The cap exists because, after extensive investigation, SARS determined that it is the highest ratio that universities will use for taxable rather than exempt supplies. If a university is using more than that, there is something unusual, and it should seek a separate ruling from SARS. That is what the Taxpayer did here – it asked for a specific departure based on its specific facts. In my view, there is no unfairness or irrationality in setting a default position for a class, but permitting deviation on a specific showing.

Second, the cap was not pulled out of thin air – it was based on SARS’s careful study of the industry and agreement with HESA. SARS first concluded that research projects were a special case and should be treated separately. It then determined that the variable input based method should be used for universities’ other acquisitions. But it went beyond that; it investigated universities’ cost centres to evaluate how universities actually used goods and services to determine how much they used on taxable rather than exempt supplies. That inquiry revealed that (save for exceptional cases) most universities were under 10%, and the highest percentage was below 12.5%. 12.5% was therefore taxpayer friendly. It was “sense- checked” against the ordinary turnover-based model, which yielded similar results.

Setting the cap at 12.5% was anything but arbitrary. It was based on detailed study and agreement, including with the Taxpayer. It was set at a level that reflected the actual practice of universities. Relying blindly on a formula without considering whether that formula produces results that align with practice would be far more arbitrary than the course SARS followed.

Third, but why, one may ask, was there a need for a cap at all? Because the variable- input formula SARS and HESA agreed on for non-research activities could produce anomalous results. Where a university had a project that had high input tax costs, but negligible overheads, it could result in a far higher apportionment ratio under the variable input formula. The example Mr Rain gave was a university that owned, but did not operate, a shopping mall. The result would be that the university could deduct a higher percentage of input tax on all its expenses that was disconnected from how it actually used those goods and services. The 12.5% cap was designed to avoid the risk that in some cases the methodology would not reflect reality.

It is important to remember what is at issue here. What SARS is trying to figure out is how much of a university’s running costs are being used for exempt supplies, and how much for taxable supplies. The ratio determines what percentage of input tax it can deduct on all the taxable goods and services it uses in a year. A high value transaction that does not meaningfully alter the university’s overheads should not entitle it to claim a higher apportionment. It would effectively permit it to deduct input VAT it used to make exempt supplies.

What TUUP wants to achieve is exactly what SARS sought to avoid. SARS realized that certain unusual transactions or practices, if fed mechanically into the formula, would produce distorted outcomes. They would permit universities to claim an apportionment far in excess of the actual percentage of its overheads they used to make taxable supplies. As I explain in detail below, that is exactly what would happen if the Taxpayer was allowed to take into account input tax from the Head Lease and abandon the 12.5% cap.

This reasoning does not “beg the question” as Mr Ben SC argued. He contended that whatever result the variable input formula produces must reflect the ratio required by section 17(1). But SARS’s evidence is that, in some cases, the formula does not produce results that achieve the purpose of section 17(1). If the result is 12.5% or lower, it is unlikely to be an anomaly and the formula is sufficiently accurate. If it is higher than 12.5%, then it is higher than any of the universities that it evaluated in detail. It will not allow that result of the formula, unless the university can justify it on a special showing with evidence beyond the inputs into the formula. Mr Ben SC’s approach wrongly treats the variable input formula as the only acceptable means for determining the apportionment ratio. As I have already explained, it isn’t.

Fourth, the Taxpayer criticises SARS for considering fairness between universities in setting the 12.5% cap. This is a rationality challenge, rather than an ultra vires one. But there are, in any event, two answers to this criticism:

[79.1] This was not an individual ruling, but a class ruling. Whatever the position may be when SARS makes a ruling for a specific vendor under section 17(1), when a class seeks a ruling, fairness must necessarily play a role. Section 17(1) is not mechanical; it does not requires SARS to apply any particular method or formula to reach the ratio. It affords SARS a discretion. There may be multiple different methods that all meet the basic criterion of section 17(1). In choosing between those methods, SARS can lawfully consider fairness between vendors within a class. And, as I pointed out earlier

and as the Taxpayer did, nothing prevents any member of the class from seeking its own, specific ruling.

[79.2] It was not the only reason SARS imposed the cap. It also sought to avoid distorted outcomes. That was a sufficient reason to justify the decision.10

In sum, the Taxpayer’s attack on the 12.5% cap is bad on its merits. If it had pleaded it properly, I still would have rejected it.

The Impact of the Head Lease
That brings us back to the Taxpayer’s original claim – that it should be entitled to deduct the input VAT on its supplies under the Head Lease. Rather than consider the merits of the Taxpayer’s argument, I focus on two of the reasons SARS gave for refusing the ruling application.

For the Purpose of Making Exempt Supplies

The Taxpayer seeks to treat the two agreements as separate agreements with separate VAT consequences. It argues that the Head Lease is, in effect, a barter transaction. The Taxpayer rents the land to the Developer. The Developer pays for possession of the land by building the residence.

Mr Ben SC relied on the SCA’s judgment in Respublica to argue that “the VAT consequences of a supply must be assessed by reference, first and foremost, to the contractual arrangements under which the supply is made.”11 Applying that rationale, the Court in Respublica held that “one cannot legitimately attribute to Respublica’s supply, governed as it was by its own contractual terms, the characteristics of an altogether different supply of accommodation to third parties under separate contracts”.12 The Taxpayer says the Court should close its eyes to the Sub-Lease, and analyse the VAT consequences of the Head Lease purely on its own terms.

‌10 The Taxpayer’s challenge, to the extent it was pleaded, was an ultra vires challenge. It did not take the point that a single bad reason means that the whole decision is irrational. Westinghouse Electric Belgium Societe Anonyme v Eskom Holdings (Soc) Ltd and Another [2015] ZASCA 208; [2016] 1 All SA 483 (SCA); 2016 (3) SA 1 (SCA) at para 45. In any event, I find that the reliance on fairness between universities was not a bad reason, and so does not taint the other, good reason.

‌11 Commissioner for the South African Revenue Service v Respublica (Pty) Ltd [2018] ZASCA 109; 81 SATC 175 at para 12.

‌12 Ibid at para 14.

But the facts in Respublica – while they also concerned student residences – were quite different. Respublica leased buildings it owned to the Tshwane University of Technology to use as student residences. It argued that, under section 10(10) of the VAT Act,13 it was only obliged to charge VAT on 60% of the consideration it received because the buildings were “commercial accommodation” as defined in the Act. The SCA held that Respublica was bound to its agreement with TUT. It could not rely on the fact that TUT had concluded rental agreements with its students – with whom it had no contractual relationship – to render its supply to TUT subject to section 10(10).

The facts here are very different and, in my view, fall under the exception contemplated in Respublica. Ponnan JA adopted the following statement of Lord Neuberger: “when assessing the VAT consequences of a particular contractual arrangement, the court should, at least normally, characterise the relationship by reference to the contracts and then consider whether that characterisation is vitiated by [any relevant] facts.”14

The facts here are that the Head Lease and the Sub-Lease are, in truth, a single commercial arrangement. Just consider the following:

[86.1] The recorded purpose of the Head Lease is to enable the Sub-Lease. The preamble to the Head-Lease records that “in terms of a separate agreement of Sub- Lease, the [Developer] is to sub-let the Improved Property back to the [Taxpayer] to use as a student residence for the period of the Sub-Lease”. Similarly, the purpose of the Sub-Lease – recorded in its preamble – depends on the Head Lease. It states that the Developer “in its capacity as Head Lessee has hired the Property from [the Taxpayer], being the Head Lessor pursuant to the Head Lease … subject to the obligation that the [Developer] erects the Residence”.

[86.2] The Head Lease defines “residence” to mean “the building, structures, infrastructure and all permanent fixtures, fittings and improvements to be constructed on the Property as part of the Project”. The improvements are, in the Head Lease, defined to mean the residence.

[86.3] The Head Lease specifies the purpose to which the Developer must put the land that it is renting. Clause 11.1 provides that the Developer “shall have the right to

‌13 VAT Act section 10(10) reads: “Where domestic goods and services are supplied at an all-inclusive charge in any enterprise supplying commercial accommodation for an unbroken period exceeding 28 days, the consideration in money is deemed to be 60 per cent of the all-inclusive charge.”

‌14 Respublica (n 11 above) at para 13, quoting Airtours Holidays Transport Limited v Commissioner for Her Majesty’s Revenue and Customs [2016] UKSC 21; [2016] 4 All ER 1 (SC) at para 47 (my emphasis).

use or occupy the Property for any lawful purpose, provided that such purpose shall primarily be that of a student residence”.

[86.4] Clause 12.1 of the Head Lease reads: “It is recorded that the main objective of the Parties entering into this Head Lease is for the [Developer] to construct the Improvements and to sub-let the Improved Property back to the [Taxpayer] in terms of the Sub-Lease.” The parties could not have made themselves any clearer.

[86.5] The two agreements are made conditional on each other. The Head Lease was subject to a suspensive condition that the Sub-Lease is signed and becomes unconditional (clause 16.1.1). Similarly, the Sub-Lease is conditional on “the Head- Lease being notarially executed and becoming unconditional”. Both are subject to a suspensive condition that the Taxpayer obtained the necessary approvals “under the Higher Education Act 101 of 1997 or any other applicable legislation, for the acquisition of the Property and construction and financing of the Residence”.

[86.6] Rentals under both agreements only became payable from 1 March 2011, a date after the Developer was required to have completed building the Residence.

Commercially, the purpose of the construction the Developer trades as consideration for possession of the land under the Head Lease is to enable it to perform under the Sub- Lease. It recoups the costs of construction only because of the rental it will receive under the Sub-Lease. The Sub-Lease rentals were set in order to ensure the Developer profited from its construction of the Residence. Without the two going together, the relationship would make no commercial sense.

These are not two separate commercial arrangements. They are a single commercial arrangement divided into two contracts. The provision of the leasehold improvements served no separate purpose. They were never intended for any purpose other than supplying a residence. The consequence is that the leasehold improvements were not acquired “for the purpose of consumption, use or supply in the course of making taxable supplies”. They were acquired for the purpose of providing student accommodation. That is an exempt supply. The VAT the Taxpayer pays to obtain those supplies is not deductible as input tax.

The case is very different from Respublica. There are only two parties in both agreements – the Taxpayer and the Developer. In Respublica the lessor sought to rely on the relationship between the TUT and the students, with whom it had no relationship. It was in that context that the SCA held Respublica to its actual agreement. The case also did not concern the question of whether the supplies were made for taxable or exempt supplies.

This is not a case of the Court improperly peering behind the terms of the parties’ contract. To the contrary, the Court is taking the parties’ contracts seriously. Those contracts, on their own terms, show that the purpose of all the Taxpayer’s expenditure under both agreements was solely to provide accommodation to its students. That is an exempt supply, for which it is not entitled to an input tax deduction.

That on its own is enough to dismiss the appeal. But I think it is appropriate to consider SARS’s next objection as well.

Inclusion would Distort the Apportionment Ratio

SARS’s argued that, even if the Head Lease otherwise meets the requirements for taxable supplies, including it would distort the apportionment ratio. Recall, the apportionment ratio under section 17(1) seeks to capture the extent to which a taxpayer that makes mixed supplies uses all the goods and services it acquires for the purpose of making taxable supplies. The Head Lease should only alter that ratio if it alters the extent to which the Taxpayer uses its annual overheads to make taxable supplies rather than exempt supplies. If it does not meaningfully increase the extent to which the Taxpayer’s overheads are used to make taxable rather than exempt supplies, then there is no basis under section 17(1) to alter the ratio.

The Taxpayer admitted, in response to a request for admissions from SARS, that its costs to manage its obligations under the Head Lease were “negligible”. Yet it sought to rely on that negligible contribution to its overhead costs to effectively double the apportionment for all of its supplies as taxable rather than exempt supplies. Mr Abba SC calculated what the effect would be of allowing the Taxpayer to claim input tax deductions for VAT charged under the Head Lease. The likely effect of adding the annual input VAT cost of the Head Lease (including it in a in the varied input formula) would be to increase the Taxpayer’s apportionment ratio to approximately 24%. That would have the consequence of upping the Taxpayer’s total input tax deduction by approximately R6.6 million per year. To justify that additional deduction, the Taxpayer would have to show it spent an additional R50.7 million in overheads as a result of the Head Lease. That would be a challenge given that the total value of the taxable supplies under the Head Lease (assuming they are not exempt supplies) is just R23.9 million. Not to mention that R50.7 million hardly qualifies as “negligible” overheads.

What is really going on here? The Taxpayer is seeking to leverage this high-value, low overhead transaction to claim that a higher percentage of all its other overheads are used for taxable supplies. But they aren’t. The Head Lease does not alter how its overheads are in fact

used. It does not increase the services or goods it needs to run its operations, and certainly not to make taxable supplies, even (assuming they are taxable) those under the Head Lease. The Taxpayer not only admitted the impact was negligible, it led no evidence to demonstrate any impact at all on its overheads.

The question, simply put, is this – why should the Taxpayer be entitled to claim double the amount of input tax deduction for every good or service it acquires because of a transaction that has no impact on how it uses those goods and services? The answer is: It shouldn’t.

As with its attack on the 12.5% cap, the Taxpayer’s claim misunderstands the purpose of section 17(1). It seeks to apply the variable input formula as if it must always produce the ratio required by section 17(1). It seeks to convert the entitlement in section 17(1), to an entitlement to the output of a formula, no matter what that output is. But the evidence shows the opposite. It shows that while the formula is a starting point, in certain circumstances it produces anomalous results so that the apportionment claimed does not “bear[ ] to the full amount … the same ratio… as the intended use of such goods or services in the course of making taxable supplies bears to the total intended use of such goods or services” as section 17(1) requires. It would allow the Taxpayer to claim input tax deductions for goods and services it in fact acquired to make exempt supplies, not taxable supplies.

For this reason too, I would dismiss the appeal.

The Capital Nature of the Input and the Time of Supply

SARS’s final defence was that the VCR excludes from the calculation “expenditure on any capital goods or services acquired unless acquired under a rental agreement/operating lease”. It argued that the leasehold improvements are not acquired “under a rental agreement”, but pursuant to a contractual obligation in the Head Lease to erect the improvements on the Taxpayer’s land. It acquires them through accession. The fact that the obligation is in a rental agreement, in which it is a lessor does not exclude the expenditure from the exception which is limited to goods supplied by a lessor to a lessee, not vice versa. The Taxpayer vehemently disagreed. It argued that it made no difference whether the goods were acquired by a lessor or a lessee. Here, the Head Lease was plainly a rental agreement and the leasehold improvements were made under it.

There was also a debate about the time of supply. SARS argued that the supply was made immediately when the Head Lease was concluded, or at the latest when the Taxpayer paid R19.22 million as a contribution to the costs of the leasehold improvements. The

Taxpayer argued that in terms of section 9(3) of the VAT Act, the supply is deemed to be made periodically over the full course of the agreement.

I prefer not to resolve these disputes. Whatever the outcome, it would not affect my basic conclusions that an input tax was for the purpose of making exempt supplies, and that the Head Lease does not increase the proportion of goods and services that the Taxpayer uses to make taxable supplies.

Conclusion and Costs
The appeal must be dismissed. The Taxpayer failed to properly plead an attack to the 12.5% cap. Even if it had the attack was bad because the cap was based on careful study and served to prevent abuse of a formula. The cap is not the law of the Medes and Persians, but a practical mechanism that permits departure on justification.

The Taxpayer’s attempt to separate the Head Lease from the Sub-Lease is artificial. They are a single, composite agreement which, from the Taxpayer’s perspective, had a single purpose – providing a residence. That is an exempt supply. Anyway, the Head Lease did not alter how the Taxpayer used the goods and services it acquired throughout the year. Even if the Head Lease could be separated from the Sub-Lease to manufacture a taxable supply, the Taxpayer is not entitled to include it in the calculation of its apportionment ratio.

Neither party sought costs. I therefore make the following order:

The appeal is dismissed.

There is no order as to costs.

MJ Bishop

CONCUR:

P. Surtees (Accounting member) TJ Ledwaba (Commercial member)