Theme: spectrum_digital_infrastructure
Responsible: ICASA / Department of Communications and Digital Technologies
Implemented: announced at MTBPS November 2025, effective immediately. SARB's MPC now formally targets 3% CPI (±1pp). First test is the 2026 interest rate cycle. Market reaction post-announcement: long bond yields compressed modestly, consistent with improved credibility. Political risk: low — SARB constitutional independence insulates the target from political interference. Directly linked to S&P sovereign outlook upgrade and fiscal consolidation narrative.
The 2024 ICASA spectrum auction allocated high-demand spectrum across five frequency bands (700 MHz, 800 MHz, 2.6 GHz, 3.5 GHz, and 26 GHz) to South Africa's mobile operators, raising approximately R14.4 billion in licence fees — a landmark revenue event and the culmination of a decade-long process delayed by litigation, regulatory disputes, and COVID-19. The allocation provides the spectral backbone needed for 5G rollout in urban centres and affordable LTE coverage in underserved rural areas. Operation Vulindlela identified spectrum release as a critical digital economy enabler: spectrum undersupply had kept mobile data costs among the highest in Africa. Licence conditions include roll-out obligations requiring 30% rural coverage within five years. ICASA must now monitor compliance, operationalise secondary spectrum trading mechanisms to allow efficient reallocation as technology evolves, and conduct a follow-on auction for the remaining digital-dividend spectrum freed by the analogue switch-off. As of early 2026, operators have received their licences and are deploying 5G infrastructure in major metropolitan areas.
The reform is structurally complete: announced at MTBPS November 2025 and effective from the 2025/26 MPC cycle. Implementation now focuses on communication strategy (anchoring expectations), MPC forward guidance, and monitoring the pass-through to bond yields and the neutral repo rate. The 3% point target with ±1pp tolerance band aligns SA with best-practice emerging market frameworks and is projected to save R15–20 billion/year in debt service costs as the inflation risk premium compresses.
Formal MPC announcement of the new 3% point target (±1pp) with a supporting technical note on inflation expectations, credibility transition path, and implications for the neutral real interest rate
SARB publish Monetary Policy Review incorporating the new target: modelling of structural disinflation path, expected 10-year bond yield impact, and updated neutral rate estimate
National Treasury publish framework note on fiscal implications: projected debt service cost savings, updated borrowing cost assumptions in the Medium-Term Expenditure Framework (MTEF)
Communication campaign targeting wage negotiators, pension funds, and bond market participants to re-anchor inflation expectations toward 3% breakeven; update National Development Plan growth model inflation assumptions
India built a national digital identity and payments stack: Aadhaar biometric ID covers 1.36 billion people; UPI processed 83 billion transactions worth USD 1.8 trillion in FY2023 — surpassing Visa and Mastercard combined. The stack is public infrastructure (UIDAI and NPCI, non-profit) with private apps competing on top. Financial inclusion rose from 35% to 80% of adults in 10 years. SA's Smart ID rollout and SASSA payment infrastructure face the same design choice between open-stack public goods and proprietary concessions.
NPCI launched UPI in 2016 as an open-architecture, interoperable, real-time payment rail built as a public good: zero merchant discount rates, open APIs, no proprietary lock-in. Digital payments grew from USD 8 billion (2016) to USD 1.5 trillion (2023). SMME access to merchant credit expanded to 70 million previously unbanked businesses. The India Stack DPI model has been replicated in 50+ countries. SA has emerging open-banking frameworks but lacks a national interoperable payment rail; government-owned open infrastructure generates adoption far faster than market-led alternatives.
Approach
NPCI (National Payments Corporation of India) launched UPI in 2016 as an open-architecture, interoperable, real-time payment rail. It was built as a public good — zero merchant discount rates, open APIs, no proprietary lock-in. Aadhaar biometric ID and DigiLocker (digital document store) created the identity layer. The government mandated UPI acceptance for government payments, driving adoption. By 2024, UPI processed over 100 million transactions daily.
Electricity Regulation Amendment Act — Competitive Electricity Market
Integrated Resource Plan (IRP) 2024 Update — Revised Electricity Mix
Energy Bounce-Back and Industrial Energy Self-Generation
National Transmission Company Capitalisation and Grid Expansion
Eskom Restructuring — Generation, Transmission, and Distribution Unbundling
How to cite
Wilse-Samson, L. (2026). IMT Spectrum Auction Completion (ICASA). SA Policy Space. NYU Wagner School of Public Policy. Retrieved 11 May 2026, from https://sa-policy-space.vercel.app/ideas/imt-spectrum-auction-completion-icasa?snapshot=2026-05-11
Data as of 2026-05-11 · latest PMG meeting 2026-05-08
12-month pass-through review: monitor BER inflation expectations survey, 10-year bond yields vs. inflation breakeven, and MPC repo rate trajectory relative to the new neutral rate estimate
Parliamentary briefing to the Standing Committee on Finance (SCOF) and Portfolio Committee on Finance on the framework change and its fiscal dividend implications
Announced November 2025; 12-month expectation anchoring period through 2026; full debt service saving benefit realised over 2–3 years as bond market adjusts
No direct fiscal cost; estimated debt service saving of R15–20 billion/year over the medium term as the long-run inflation risk premium compresses (MTBPS 2025 estimate based on R5.4 trillion debt stock)
None required. The SARB inflation target is set administratively by the Minister of Finance under Section 224(2) of the Constitution (coordinated monetary-fiscal objective). The South African Reserve Bank Act 90 of 1989 provides the operational independence framework. An amendment to the MPC Charter (SARB internal document) formalises the ±1pp tolerance band interpretation.
Strong support across the fiscal policy establishment and GNU economic cluster. Labour-aligned economists (NUMSA, COSATU research) argue the lower target may sustain higher real interest rates in the near term — a transitional credibility risk. DA and ANC economic clusters both support the reform for its fiscal dividend. EFF opposes SARB independence on political-economy grounds but lacks blocking power.
Chile adopted 3% ±1pp in 2000 — now the benchmark framework for emerging market central banks. Czech Republic, Peru, and South Korea operate similar frameworks with strong credibility. IMF 2023 Article IV review of SA monetary policy explicitly recommended a lower target to reduce the inflation risk premium. New Zealand (2% midpoint) and Australia (2–3% band) demonstrate tight bands with operational independence do not impair long-run growth.
Timeline: 2 years to critical mass (2016–2018); full ecosystem by 2022
Lessons for South Africa
SA has PayFast, SnapScan, and emerging open-banking frameworks but lacks a national interoperable payment rail. The SARB's Vision 2025 National Payment System roadmap is the closest analogue but moves slowly. India's model shows that government-owned, open infrastructure combined with mandated acceptance generates adoption far faster than market-led approaches. The SA fintech sector has repeatedly called for interoperability mandates — India's experience strongly supports this.
Kenya invested in the national fibre optic backbone (NOFBI) from 2009, combined with submarine cable landings (TEAMS, SEACOM, EASSy), with an open-access mandate. Internet penetration rose from under 4% (2008) to 85%+ (2022). Nairobi became Africa's leading tech hub ('Silicon Savannah') hosting 400+ startups and USD 1.1 billion in startup investment by 2021. Digital services contribute ~8% of GDP. SA Connect Phase 2 has the right open-access architecture but has been hampered by execution delays and underfunding.
Approach
Kenya invested in a national fibre optic cable network (National Optic Fibre Backbone Infrastructure — NOFBI) beginning 2009, combined with the landing of submarine cables (TEAMS, SEACOM, EASSy). The government mandated open access on the backbone and set aggressive rural connectivity targets. The Konza Technopolis was designated a special tech zone with simplified licensing and tax incentives.
Timeline: 5–8 years for infrastructure maturity; ecosystem formation takes 10+ years
Lessons for South Africa
SA Connect Phase 2 has the right architecture (open access backbone) but has been hampered by execution delays and underfunding. Kenya's model suggests that government-anchored backbone investment, combined with open access mandates and SEZ-style tech zones, creates the platform for private ecosystem formation. SA's spectrum release bottleneck (5G licensing) is a direct analogue to Kenya's 2010 WiMax licensing controversy — resolved by a regulator that prioritised access over incumbents.
Vietnam's 2020 National Digital Economy Strategy targeted 20% of GDP from digital economy by 2025. The government mandated Make-in-Vietnam software targets for public procurement, allocated 5G spectrum (world's 7th commercial 5G deployment, 2022), and established dedicated digital industrial parks. Digital economy reached 14.3% of GDP in 2022 (vs 8.2% in 2018). Vietnam exported USD 16 billion in software and IT services (2022), becoming a top-10 global tech outsourcing destination. SA's 5G spectrum allocation was significantly delayed by Telkom legal challenges; Vietnam prioritised access over incumbent protection.
Approach
Vietnam's 2020 National Digital Economy Strategy targeted 20% of GDP from digital economy by 2025. The government mandated Make-in-Vietnam software development targets for public procurement, invested in 5G spectrum allocation (world's 7th country to deploy 5G commercially, 2022), and established dedicated digital industrial parks (Vietnam Software and IT Park, Ho Chi Minh City). University partnerships with Samsung, Intel, and LG created a tech talent pipeline.
Timeline: 3 years for 5G deployment; 5 years for ecosystem maturity
Lessons for South Africa
SA's 5G spectrum allocation was significantly delayed (Telkom legal challenges) compared to Vietnam's streamlined approach. Vietnam's Make-in-Vietnam software procurement preference has a SA analogue in the B-BBEE software procurement targets, but SA's version is not enforced consistently. The tech park model (Vietnam's SHTP) is directly relevant to SA's plans for Tshwane Innovation Hub and Bandwidth Barn, which need dedicated SEZ status and 5G infrastructure guarantees.
Freight Rail Third-Party Access and Transnet Separation