Binding constraint
5 reform ideas in the database address this constraint.
In HRV terms, digital infrastructure binds South African growth because its shadow price on economic activity is demonstrably high: firms and households face mobile data costs that, until the 2022 spectrum auction, ranked among the most expensive in middle-income peers, and backhaul and spectrum scarcity directly throttled the productivity of every downstream sector — finance, logistics, BPO, township retail — that now runs on connectivity. The binding character shows up in the classic diagnostic signature: private actors were willing to pay (MTN, Vodacom and rivals committed roughly R14.4 billion at the ICASA high-demand spectrum auction), returns to relieving the bottleneck were visibly high, yet the constraint persisted for over a decade because of regulatory and coordination failure rather than a shortage of capital or demand. This is the textbook "low appropriability via government failure" node on the HRV decision tree.
The database reflects a portfolio that moves from acute unblocking to structural deepening. The two quick wins — IMT Spectrum Auction Completion (ICASA) and Digital TV Migration and Analogue Switch-Off — are already implemented and between them release the sub-1 GHz digital dividend that makes rural 4G/5G economically viable; they are the reason this constraint is loosening rather than tightening. The forward-looking ideas address the layer above the pipes: Digital Platforms Competition Framework (Buyer Power Regulations) tackles the appropriability problem once connectivity is solved (who captures the rents on the platforms that ride the spectrum?), while the CSIR 10-Year Foundational Digital Capabilities Research Programme and satellite commercialisation target long-horizon sovereign capability. The common thread is sequencing: ICASA and DTT were DCDT-led regulatory unlocks; what remains is competition policy (Competition Commission) and industrial-capability investment (DSI/CSIR).
Watch three indicators over the next year: whether Vodacom and MTN's post-auction capex translates into measurable declines in the Competition Commission's data price index, particularly for sub-1 GB prepaid bundles; whether the Buyer Power Regulations survive consultation with teeth intact; and whether SA Connect's fibre rollout numbers finally converge with its repeatedly-missed targets. If two of three move, the constraint is genuinely slackening.
Synthesis drafted by Claude from the 5 ideas under this constraint on 2026-04-23, then human-reviewed. Reassessed as the database grows.
How to cite
Wilse-Samson, L. (2026). Digital Infrastructure — binding constraint. SA Policy Space. NYU Wagner School of Public Policy. Retrieved 11 May 2026, from https://sa-policy-space.vercel.app/themes/digital_infrastructure?snapshot=2026-05-11
Data as of 2026-05-11 · latest PMG meeting 2026-05-08