By Caroline Mbugua

Africa does not have to convince anyone to invest in its mobile networks. Operators are projected to invest $76 billion in African network infrastructure between 2025 and 2030. In 2025 the sector added $240 billion to the continent’s economy, about 7.8% of GDP, supported 13 million jobs and paid $45 billion in taxes and fees. The forecast for 2030 is $290 billion.

The harder question is whether that investment is sustainable, in the plain financial sense. Does the capital that arrives this year earn enough, safely enough, to come back in five years and go further than it did the first time?

At the moment the answer varies widely across the continent, and the reason is risk. Only 9% of Africans live outside mobile broadband coverage. Close to a billion people are covered and still offline. An asset that is built and not used struggles to pay for its own expansion, and every factor that keeps people offline is a risk on the balance sheet of the next investment decision.

Most of those risks are known and most can be reduced. They sit in four places: finance ministries, regulators, operators and the institutions that finance all three. Investors can price risk they can measure. What raises the cost of capital is uncertainty, and much of that can be removed at very little cost.

Finance ministries: Demand risk and currency risk

A large share of what decides telecommunications investment is settled in the finance ministry rather than the ICT ministry. Having both at the table does more for investor confidence than most incentive schemes.

Device taxation is the first place to look, because it goes directly to demand. Of the $45 billion the sector paid in 2025, $20 billion came from VAT, sales taxes, excise and customs duties on handsets.

“An entry-level internet-enabled phone costs around 24% of monthly income on average in low- and middle-income countries, and as much as 80% at the lower end. The handset decides whether a network earns a return, and it is the most heavily taxed part of the chain.”

The industry is not asking to pay less tax. It is asking governments to look at where the tax falls. South Africa removed a 9% excise duty on entry-level smartphones in April 2025. In the eleven months that followed, entry-level smartphone sales rose by 80%, feature phone sales fell by 87%, and around 1.1 million additional smartphones were sold. More users on an existing network is the lowest-risk growth there is.

That is why the Digital Africa Summit in Lilongwe in August was encouraging. Finance and ICT were in the same room, with regulators and operators, working through the same numbers.

Two further measures cost a treasury nothing and remove currency risk almost entirely:

● Reliable access to foreign exchange is one. Operators earn in local currency and buy equipment in dollars; if they cannot convert, a profitable business cannot reinvest, and no investor commits a second round into a market where the first is trapped.

● Designating telecommunications as critical national infrastructure is the other. It gives networks the same forex priority, grid access and legal protection that power and water already have. It is a classification, not a spending line.

Regulators: Regulatory risk and asset-life risk

Across more than a thousand spectrum assignments in 102 countries, final prices in developing markets were over three times higher than in developed markets once income is accounted for. The consequence is visible in coverage. In the highest-priced markets the average operator’s 4G network would reach 7.5 per cent more people had spectrum been bought at the median price. Where spectrum was released two years earlier, 4G coverage was 11 to 16 percentage points higher. Those are millions of people per country.

The fixes are specific. Licence terms of twenty years or more, so the licence outlasts the payback period. Technology-neutral authorisation, so a band can move from 4G to 5G without a fresh application. A published spectrum roadmap, so operators can plan capital over a decade rather than a licensing cycle. Fees set in local currency, so a devaluation does not raise costs at the moment customers have less to spend. None of this reduces spectrum revenue over time. It makes it predictable.

The same logic applies to rights of way. By October 2025, eleven Nigerian states had waived right-of-way fees for fibre and seventeen more had capped them. The Nigerian Communications Commission linked those reforms directly to more than $1 billion in additional broadband rollout commitments. Where a regulatory cost falls, capital follows.

Operators: Execution risk

The industry’s own obligations mirror what it is asking for. A twenty-year licence should be matched by a twenty-year commitment, including to districts where payback is slow. Critical infrastructure status only holds if operators share towers and fibre instead of building duplicate sites a few hundred metres apart. Shared infrastructure, separate competition, is the model. And if governments cut duties on handsets, operators should have financing ready so the cheaper device actually reaches new users.

The commercial case is clear. Every new user improves the return on infrastructure already paid for. That is the cheapest, lowest-risk growth available to the industry, and it is what makes the next round bankable.

Financiers: Cost-of-capital risk

Development banks, commercial banks and African pension funds lend to roads, ports and power plants over twenty years at infrastructure rates. When they lend to towers and fibre at all, it tends to be over shorter periods at technology-sector rates. A tower estate outlasts most roads. Treating digital assets as infrastructure would lower the cost of capital, and expensive short-term debt is a main reason second-round investment stalls.

Device financing is the larger gap on the demand side. A duty cut helps the household that can afford a phone at the new price. It does nothing for the household that cannot pay in one instalment, which, where a phone costs up to 80 per cent of monthly income, is a large share of them. That needs banks and development partners structuring credit with operators rather than around them.

There is also money already collected. GSMA studies of universal service funds across Africa have repeatedly found large sums levied from operators and left undisbursed. Deploying that money is faster than designing a new instrument.

The test of a de-risked market

The clearest sign that a market has been de-risked is whether investment happens twice. Did the operator that committed capital three years ago come back and commit more, in the same country, to harder districts than before?

By that measure, several African markets are doing considerably better than their headlines suggest. Others have announcements and little second-round money. The test is harder than counting pledges, but it is the one that matches what people actually experience: whether the network reaches them, and whether they can afford to use it. Sustainable investment and connected people are the same outcome, reached by the same decisions.

Caroline Mbugua, HSC Senior Director, Public Policy & Communications, GSMA