The appellant was a bookmaker operating a registered VAT enterprise. From September 1991 to June-July 1996, he overpaid R1,432,038.83 in VAT. The overpayment arose because he included 'take-back' winnings (winnings received from other bookmakers when he placed covering bets with them) in his output tax calculations. This was done in accordance with VAT Guide 404 issued by SARS. However, prior to the 1996 amendment inserting s 8(13A), the Value-Added Tax Act 89 of 1991 did not actually require take-back winnings to be included in output tax calculations - this was a legislative lacuna. The appellant used a 'netting off' method of calculation (deducting take-back winnings from winnings paid to punters when calculating input tax, rather than adding them to output tax), though the result was the same as the method in the Guide. He claimed a refund of R1,417,018.99. The Commissioner limited the refund to R336,259.93 (six months' overpayment) on the basis that payment was made according to a generally prevailing practice, invoking the limitation in s 44(3).
Appeal dismissed with costs, including costs of two counsel. The Commissioner's limitation of the refund to the six-month period (R336,259.93) was upheld.
The binding legal principles are: (1) A refund under s 44(1) of the VAT Act is only available where there is an excess of input tax over output tax as contemplated in s 16(5), arising from amounts deductible under s 16(3) that exceed output tax. (2) Where tax is overpaid due to overstatement of output tax (rather than underdeduction of input tax), the refund is properly categorized under s 44(2)(a) as an overpayment, not under s 44(1). (3) The substance of tax calculations, not the particular method employed, determines the legal character of amounts for VAT purposes - where different calculation methods produce identical results, they cannot be distinguished for the purpose of determining which refund provision applies. (4) A 'practice generally prevailing' under s 44(3) is established where there is widespread industry compliance with a particular approach to tax calculation, supported by SARS guidance, even if that approach is later discovered to be based on a misunderstanding of the legislation. (5) The taxpayer bears the onus under s 37 of proving that payment was not made according to a generally prevailing practice in order to avoid the six-month limitation in s 44(3).
The court noted that it was common cause that, given the scheme of the VAT Act, one would ordinarily expect take-back winnings to be taken into account in calculating output tax, just as payment of winnings to punters is taken into account in calculating input tax. The court also observed that the absence of a provision like s 8(13A) (inserted in 1996) appeared to be a legislative lacuna 'overlooked by the Commissioner and bookmakers alike.' The court's description of the legislative gap as 'fortuitous' suggests sympathy with the appellant's position from a policy perspective, though the law as it stood required dismissal of the appeal. The court also noted in passing that whether bookmakers used the 'netting off' method or the method contemplated in the Guide and the Act 'is of little consequence; the result was the same.'
This case is significant in South African VAT law for several reasons: (1) It clarifies the distinction between refunds under s 44(1) (for excesses arising from s 16(5) due to input tax exceeding output tax) and s 44(2)(a) (for overpayments of tax). (2) It establishes that the substance of tax calculations prevails over the form - different calculation methods producing the same result do not change the legal characterization of the payment. (3) It interprets the 'practice generally prevailing' limitation in s 44(3), confirming that widespread industry compliance with SARS guidance (even if based on a legislative lacuna) constitutes such a practice and limits refund claims to six months. (4) It illustrates the application of the onus of proof in s 37 of the VAT Act in refund disputes. (5) It highlights the consequences of legislative gaps and the importance of SARS guides in establishing industry practice, even where those guides incorrectly interpret the legislation.