The Pink Umbrella
What a walk down William Street tells you about who mobile money was actually built for
Stand at the corner of William Street and Ben Kiwanuka around ten in the morning. Count the pink umbrellas. Count the red ones. Under every one of them there is a wooden stand, a printed sign listing MTN and Airtel rates, a small metal cash box, and a person with two phones. That is the entire retail face of a rail that carried 2.37 billion transactions in Uganda in the first quarter of 2026 alone.
Watch for ten minutes. Notice a thing that no press release mentions. Nearly every customer who steps up to the stand is there to withdraw. Cash-in is rare. Cash-out is the whole business. Whatever went into the wallet up-country, whatever a boss sent, whatever a customer sent, is being converted back into paper shillings on William Street within an hour of arriving.
That single observation, repeated at every pink umbrella from Nakasero to Kabalagala, contains the whole argument I want to make.
The rail is not a rail. It is a delivery pipe with a hole cut at one end. Money enters, travels the digital distance, and then leaks back to cash the moment it lands. Almost nothing circulates on it. The transactions that look like flow are mostly one-way trips ending at a plastic table under an umbrella.
Now walk two hundred metres to the katogo stall in the alley behind Owino market. Ask the mama whether you can pay her with MoMo. She will say no. If you press her, she will give you her personal number and quietly add UGX 500 to the bill. She is not being difficult. She is doing arithmetic. Her margin on the plate is about UGX 1,500. If she takes MoMo she has to walk to the pink umbrella in an hour to convert it into shillings, because her supplier at Kikuubo does not accept mobile money either, and the withdrawal will cost her a UGX 330 provider fee, a fifteen percent excise on that fee, and a half percent government excise on the whole amount. She loses somewhere between five and ten percent of that plate to the return trip. She is being asked to underwrite the state’s revenue on every UGX 5,000 she earns.
She cannot afford to.
The bank on the other side of the street
Now walk to Kampala Road. Enter Stanbic, or Centenary, or DFCU. Withdraw a million shillings across the counter. The fee is around UGX 315. There is no government excise line on that receipt. The teller counts the notes and slides them to you.
Withdraw the same million shillings from an MTN agent under a pink umbrella. Depending on how you split the amount, once every fee and excise is counted, you have paid up to UGX 6,630.
The same one million shillings. The same shopping trip after it. Twenty-one times the cost to move it through the technology that was pitched, in every launch speech since UTL’s first Mobile Money campaign, as her bank.
This is the sentence I want to sit with, because it exposes something that most conversations about digital payments in Uganda skate past. Mobile money in Uganda is now more expensive than a bank for the class of transaction it was invented to serve.
Not slightly more expensive. An order of magnitude more expensive.
So the question a reader has to hold, before deciding whether mobile money is a lie or a wonder, is a specific one. What kind of infrastructure requires the poor to pay twenty times what the rich pay to use the equivalent private service?
Is it a lie
No.
The technology is not a lie. The rail exists. The 2.37 billion transactions in Q1 2026 happened. My mother, who lives in Kasilo, can receive her pension inside ninety seconds. When my nephew’s school fees are due, I do not queue at a bank. I press *165#, key in the school code, and it clears in the time it takes to boil a kettle. Whatever else is true, that is a genuine thing that did not exist in Uganda in 2005.
But the covenant around the technology has been broken, and it is the covenant, not the technology, that people are calling a lie when they say cash is better.
The covenant was this. The rail will be built cheaply. The rail will be priced below the alternatives, especially for the population the older alternatives never reached. Poor Ugandans will move their small commerce onto it. Their small commerce, seen for the first time, will grow.
The technology kept its side. The phone in the mama’s hand is capable of carrying the transaction. The USSD flow works. The rail’s uptime, network shutdowns aside, is high.
The state did not keep its side. It taxed the exit door. Half a percent of every withdrawal, from the mama’s UGX 5,000 to a corporate treasurer’s UGX 500,000, goes to URA. Fifteen percent excise sits on top of every provider fee. Ten percent withholding tax reduces the commission that keeps the pink umbrellas open. The rail is not just being maintained. It is being milked, and the milking is priced at a level that has changed the behaviour of the people it was supposed to serve.
The mama at the katogo stall is not rejecting mobile money because she does not understand it. She is rejecting it because she has understood it exactly.
Should you revert to the bank
For some things, yes. And it is important to say this plainly, because Ugandan financial commentary is full of vague loyalty to mobile money that does not survive contact with the actual fee schedule.
If you are moving five hundred thousand shillings or more once a month, a bank transfer through the national settlement rail, EFT or RTGS, is now cheaper than the mobile money equivalent for most banks and most amounts. If you are holding savings you do not intend to touch for weeks, a bank pays interest and mobile money mostly does not. For salary receipts, standing orders, and any transaction where the counter-party has a bank account too, the numbers favour the bank.
But this is where the honesty ends.
The bank is not present in Kasilo. The bank counter is closed at 6pm when the mama shuts her stove. The bank does not accept a UGX 3,000 payment for a plate of matooke. The bank does not receive a small remittance from a boda rider in Bwaise at midnight. The bank cannot replace mobile money for the transactions mobile money was invented for, because the bank was never trying to be there.
So the sensible reading of the bank comparison is not that mobile money is obsolete. It is that Ugandan mobile money has been degraded, by pricing, to the point where on many mid-size transactions the century-old alternative outperforms it.
That is a specific policy failure, not a technology failure. And it can be reversed.
Why URA and BoU are being short-sighted
The revenue argument for the withdrawal excise is real. Mobile money is easy to tax. Two big operators, digital records, no field inspectors required. In a country where the informal cash economy is enormous and unmeasured, taxing the small share of it that has become visible on digital rails looks, to a Ministry of Finance under revenue pressure, like the responsible choice.
That reading is short-sighted in a particular way.
The tax is being charged on a base that is smaller than it would otherwise be, precisely because the tax exists. Every trader who chose cash because MoMo was too expensive is a merchant who does not pay merchant fees, does not build a transaction history, does not become bankable, does not enter the formal economy, does not eventually pay VAT on visible sales, does not eventually pay corporate income tax on visible profit. The withdrawal excise collects a shilling today by preventing the formation of the base that would have paid ten shillings in five years.
You can see the counter-example ninety minutes east by plane. Kenya, whose Central Bank chose not to introduce a withdrawal excise, has a mobile money system that carries a share of national economic activity Uganda has never approached on a per-capita basis. Kenya has 2.1 million active informal merchants receiving payments through a product called Pochi la Biashara, activated by dialling *334# on a Safaricom line. No paperwork. Free to register. Free for the customer on payments up to KSh 200. The Central Bank did not tax that rail. It negotiated with Safaricom, most recently on 7 August 2026, to cut merchant fees further.
The visible Kenyan revenue from mobile money withdrawal tax is zero. The invisible tax base Kenya has built, through business income on 2.1 million visible informal merchants who would otherwise be invisible, is very large.
This is the pattern the phrase gestures at when I say that taxing infrastructure trying to grow shrinks the tax base over the medium term. It is not a paradox. It is not clever. It is arithmetic, and it unfolds slowly enough that in any single budget cycle the visible collection looks like prudence and the foregone base looks like nothing at all.
The Ministry of Finance proposed cutting the withdrawal excise from 0.5 to 0.25 percent in the 2026/27 budget. The proposal was shelved in April 2026. The Permanent Secretary confirmed the reduction would not be part of the coming financial year. The stated reason was revenue protection. The unstated cost, borne by every mama at every katogo stall in downtown Kampala, does not appear on any spreadsheet URA maintains.
The lesson the Kenya story actually teaches
Kenyan mobile money worked not because Safaricom was virtuous, and not because Kenyans are more entrepreneurial, and not because Nairobi is more modern than Kampala. It worked because the Kenyan state, from 2009 onwards, made a policy choice to treat M-Pesa as infrastructure rather than as a tax base.
The Central Bank of Kenya set the tone. When Safaricom’s dominance produced pricing concerns, the regulator pushed for lower fees rather than higher taxes. When merchant fees looked like a barrier, the regulator pushed Safaricom to eliminate them for customers in 2018. When smaller informal traders were still not participating, Safaricom launched Pochi la Biashara in 2021 to remove the last friction, and its adoption in the years since has outrun the older Lipa Na M-Pesa till number product. Each of these was a deliberate design choice made against short-term revenue interests, in favour of long-term ecosystem growth.
The Ugandan state made every mirror-image choice. It introduced the withdrawal excise in 2018 at one percent, halved it under pressure, then let it stay. It layered fifteen percent excise on top of provider fees. It required identity documents to register even the smallest informal merchant. It shelved a modest reduction proposal in 2026 that would have been the first real signal of a policy change. And it did all of this while its own financial inclusion strategy documents continued to describe mobile money as central to bringing the unbanked into the formal economy.
The gap between the strategy documents and the tariff is the story.
What a walk back up William Street looks like
Turn around and walk back the way you came. Notice the pink umbrella again. Notice the notebook the agent keeps beside her cash box, in which she still writes every transaction number by hand, because she does not trust that a query on her phone will produce the record when she needs it. Notice that the customer withdrawing money in front of her did not sit on the balance for a single afternoon. Notice that she will not accept a UGX 2,000 tip if you try to send it by MoMo, because the fee would eat it.
Every one of these small refusals is a piece of information the treasury does not currently read.
The technology in her hand is not the lie. The launch speeches were not entirely lies either. Something was actually built, and it does work, and it can be seen working every day. What has been broken is the price at which it is offered to the people it was built for, and that price is a policy choice that can be unmade.
The mama at the katogo stall does not need financial literacy training. She needs a country whose revenue authority is willing to lose a small visible line today to grow a large invisible base tomorrow. She needs a regulator willing to negotiate with two operators rather than tax them and their customers by default. She needs a merchant product she can activate by dialling a short code, not by producing an identity card and waiting for a printed sticker.
None of these things require new technology. They require a decision.
The pink umbrella is still up on William Street tonight. It was up ten years ago. It will be up ten years from now. What changes, or does not change, is whether the woman under it is presiding over an exit door or over a beginning.
Right now it is an exit door. That is not the technology’s fault. That is our fault, and it is fixable, and we are running out of years in which we can pretend not to have noticed.