The South African Revenue Service’s consultation paper, published in August 2026, proposes collecting Vat-related transaction details in near real time, and the question at the centre of the plan is not simply whether technology can lift revenue, but whether the authority can use new powers of scrutiny without punishing compliant businesses in the process.
Under the proposal, Vat-registered businesses would issue invoices in a standard digital format, with the relevant details transmitted automatically to the authority. The service could then use that stream of information to pre-fill returns for businesses to review and correct, and to identify discrepancies sooner than the current system allows, which relies on checking supporting invoices after submission, only when requested. Public comments are due by 16 October 2026.
The stakes for the state are considerable. Vat is South Africa’s second-largest source of tax revenue. In the 2024/25 fiscal year it raised R457.8 billion, roughly US$28 billion, after refunds, accounting for 24.7% of total tax revenue. Personal income tax remains the largest contributor at 39.3%. A study cited in the consultation estimated the domestic Vat reporting gap at between 35% and 41% across the 2016 to 2020 tax years, a measure of the difference between what was declared and what would have been declared under full compliance. Sectoral estimates drawn from the same period put the gap at approximately 47% in construction and 5% in gold mining, figures that can inform which sectors adopt the new system first, alongside potential revenue gains and businesses’ readiness.
The revenue service’s indicative roadmap places the pilot in the 2029/2030 fiscal year, with phased implementation starting in 2030. The consultation paper itself identifies Vat gap indicators, compliance risks and ease of adoption as considerations for how that rollout is sequenced. For an oversight institution, the pilot’s real test will be whether automated checks flag discrepancies that businesses can correct, distinguish those from claims requiring further investigation, and whether legitimate refunds reach businesses sooner rather than later.
Regional evidence suggests the answers are not automatic. Research from Uganda, Rwanda and Ethiopia, alongside South African studies, indicates that digital records can improve Vat reporting but that collecting records is only part of the task. Authorities must also compare buyers’ and sellers’ records, identify claims that do not match, and investigate possible under-reporting, otherwise businesses may find other routes to reduce what they declare.
Uganda required Vat-registered businesses to begin issuing electronic invoices from January 2021, reporting sales information to the tax authority, initially with flexibility to enter details through an online portal. In April 2022, about 5 000 larger firms were instructed to connect their accounting software directly to the system. Research on that stricter requirement found affected businesses reported 43% fewer purchases while their reported Vat liabilities rose by nearly 150%, which researchers read as evidence that tighter checks curbed exaggerated purchase claims. Yet reported sales fell by 39%, suggesting some businesses shifted toward hiding sales once inflating purchases became harder. Higher declared Vat also does not translate immediately into cash payments, because many businesses had accumulated Vat credits that could offset amounts owed.
Rwanda demonstrates why follow-up matters. A study using records from 2012 to 2019 estimated that electronic invoicing increased annual Vat declared after deductions by about US$11 000 per firm, but audits among firms using electronic invoicing raised subsequent annual declared Vat by an estimated US$33 400 per firm, suggesting digital records make audits more effective. In Ethiopia, electronic sales registers increased Vat payments among adopting businesses in Addis Ababa by at least 48% and reduced reporting inconsistencies, yet across declarations studied for 2010 to 2016, about 42% contained different sales figures for income tax and Vat, and officials did not fully use the available information to address those discrepancies. South Africa’s revenue service already plans automated risk detection, which could address precisely that gap in institutional follow-through.
Meanwhile, the second half of the accountability equation concerns refunds. Refund claims arise when deductible Vat on purchases exceeds Vat charged on sales, and delaying payment restricts money available for investment and operating costs. A South African study of refund administration found that roughly halving the audit rate on refund claims, alongside faster processing, increased investment by 31% and reported sales by 24% among affected firms, with delays particularly damaging to smaller businesses. Fewer audits and smaller adjustments did reduce additional tax recovered through those checks, and the study did not establish overall short- and long-term revenue effects, nor did it assess electronic invoicing. Still, it frames a practical test for the reform: can better records let the authority investigate questionable claims while paying legitimate refunds more quickly?
Implementation risks also fall unevenly. Software changes and staff training may create particular challenges for smaller businesses, which may need support to participate. Assessing declarations alongside cash payments, investigation workloads, refund delays and adoption costs would show how the reform works in practice. For the revenue service, success would mean detecting under-reporting while making compliance easier and paying legitimate refunds promptly. Further coverage of the proposal is available at https://www.moneyweb.co.za/mymoney/moneyweb-tax/south-africa-is-going-digital-on-vat/ as the 16 October 2026 comment deadline approaches.