UPI vs. M-Pesa vs. nothing — what a continental payment rail actually requires.

Every six months the conversation comes back. When does Africa get its UPI?

The implicit question is when a continental, interoperable, low-cost payment rail will appear and stitch together a billion-and-a-half people who currently transact across a forest of incompatible mobile money networks, card schemes, and bank rails. The implicit answer is usually: when one of the big mobile-money operators gets large enough to set the standard.

That answer is wrong. UPI did not happen because a private operator got large enough. UPI happened because the Reserve Bank of India and the National Payments Corporation of India decided that interoperability was a public good, set the protocol by regulation, and forced every bank in the country to implement it. M-Pesa did not happen because Kenya built a public payment rail. M-Pesa happened because one operator achieved network dominance inside one country and locked in. Those are completely different stories.

A continental payment rail requires three things. Africa has not built any of them.

The first is a protocol with a public-good owner. UPI is owned by NPCI, a non-profit consortium of public banks chartered by the central bank. There is no equivalent for the continent. The African Continental Free Trade Area has a payment system called PAPSS, which is genuinely promising, but it is currently positioned as a settlement layer between national rails, not as a citizen-facing protocol. The difference matters: UPI is what a citizen uses. PAPSS is what the bank settles on. They are not interchangeable.

The second is mandatory interoperability. UPI works because every Indian bank is required to implement it. There is no opt-out. The protocol is a tax of being a bank in India. Across Africa, mobile money operators have spent the last fifteen years building deliberately incompatible products because incompatibility is a moat. No regulator on the continent has yet been willing to legislate that moat away. The day one does, the conversation changes overnight.

The third is a cost ceiling. UPI transactions are functionally free for the user — the cost is socialised across the banking system. Mobile money fees across Africa average two to three percent per transaction. That fee is the difference between a payment rail that serves a billion people and a payment rail that excludes the bottom forty percent of the income distribution from formal commerce. A continental rail with three percent fees is not a public good. It is a tax on poverty.

I am sometimes asked why this matters when M-Pesa works fine within Kenya. The answer is that working within Kenya is the ceiling of what closed-loop mobile money can do. It cannot move money to a Ghanaian merchant. It cannot settle a Nigerian salary. It cannot underwrite a cross-border supply chain. Every African economy that wants to participate in the rest of the world needs a rail that is not nationally bounded. M-Pesa, by design, is nationally bounded.

The good news is the technical work is not the bottleneck. The technical work for a continental UPI has been done several times in academia and is sitting in white papers. The bottleneck is political will and regulatory courage. The countries that move first on this will own the rail. The ones that wait will be customers on someone else’s.

Africa does not need another payments app. It needs a payments protocol with a public chassis. The first government to ship that, properly, will reshape the continent’s economy for a generation.