Mbugua Njihia arrived at the Nairobi Serena Hotel on a boda boda. He said so from the stage, and he said it to make a point: it was the most seamless way to get across Nairobi on a Tuesday morning, and the measure of success for ride hailing regulation, as far as the person paying the fare is concerned, is whether he can be in a meeting in Upper Hill, another in Westlands fifteen minutes after it ends, and a third somewhere else before lunch. Whether you are rich or poor, he said, you have the same twenty-four hours.

That was one of five answers to the first question I put to the panel I moderated at the Institute of Economic Affairs’ Future of Mobility Public Forum on the 6th of October 2026. The others came from a driver who has been on the apps since 2016, the general managers of Uber and Bolt in East Africa, and the Director of Road and Rail Transport at the State Department for Transport. The question was deliberately simple. What should ride hailing regulation in Kenya be trying to achieve, and how would we know in two years that it had worked? Two hours later, I came away with a clearer view of this sector than I have had in a decade of using it, writing about it and, this year, moderating conversations about it. The most important thing I learned is that the argument Kenya has been having in public is not the argument the people inside the sector are actually having.

Why IEA convened this forum, and why now

Some context before the panel. The Institute of Economic Affairs (IEA) is a membership think tank, and it convened this forum independently with a specific purpose: to put evidence in front of the stakeholders before the next round of rules is written. The timing matters. The 2022 Transport Network Companies regulations have been contested since the day they were gazetted, and a High Court decision in early September, which may still be subject to appeal, has put the question of how this sector should be regulated back on the table. John Mutua, IEA’s Programmes Coordinator, defined the morning around a single question: how can Kenya develop a regulatory framework that simultaneously supports sustainable driver earnings, affordable mobility, consumer safety, fair competition, innovation and continued investment?

His presentation was the economics lesson the sector has needed for years. Ride hailing is a two-sided market. Riders on one side want affordable, safe, reliable mobility. Drivers on the other want sustainable earnings and fair treatment. The platform in the middle sets or influences fares, commissions and incentives. Change one variable and the effects ripple through the whole system. A driver’s net earnings are not the main fare. They are the fare minus platform charges, minus fuel, financing, insurance, depreciation and waiting time. Regulate the fare alone and you have not touched most of what determines what a driver takes home. His closing argument was that regulation exists to correct market failure, but it can just as easily create regulatory failure, and that smart regulation is outcome-based, technology-neutral, predictable, proportionate, evidence-led and subject to review. He also noted that the court’s finding on public participation, together with a National Assembly committee report calling for a regulatory impact assessment, has opened a policy window that the sector should use.

Maggie Ireri of TIFA Research then presented the consumer evidence. TIFA interviewed 733 Nairobi residents face to face in July, a sample she defended as robust given that Nairobi accounts for the bulk of Kenya’s ride-hailing trips. The findings are stark. 81% of Nairobi households are already under financial pressure from the cost of living, and 92% say they are concerned about rising prices. The average respondent takes 6.8 ride-hailing trips a month, and 72% of those trips are for work, business, shopping, errands or emergencies rather than leisure. 73% had never heard of the proposed minimum fare, which raises an uncomfortable question about public participation. 63% believe fares should be set by competition rather than by Government, 59% think a higher minimum fare is the wrong policy, and 60% say they would switch to matatus or other alternatives if fares rose significantly. That was the evidence base the panel walked onto the stage with.

Five people, five different definitions of success

I asked the opening question in the order I had planned, and the answers set up everything that followed.

Dimmy Kanyankole, Bolt’s Senior General Manager for East Africa, put safety first: vetted drivers, tracked trips, everyone getting home. On welfare, he made the distinction that would run through the whole morning. Driver welfare is not a mandated price. It is the full set of costs that sit between what a driver grosses and what a driver takes home, from vehicle financing to fuel, insurance and maintenance, and the regulator’s most useful role may be in policies outside the fare altogether, such as electrification, which he said could cut a driver’s operating costs by more than half.

Daniel Manga, Chairman of the E-hailers Transport Operators Umbrella Association, surprised the room. I had expected the driver representative to argue for higher fares. He did not. He described the early days of Uber in Kenya as “a blessing”, when drivers knew who they were picking up and a trip was a connection rather than a gamble, and then he listed what drivers actually want: transparency in the pricing mechanism, protection against arbitrary deactivation, a clean appeals process, social protection for drivers who are injured or sick and cannot go online, and access to financing. His estimate was that around 70% of drivers now treat this as their full-time job. “It’s not that we want the price to increase,” he said. “We are looking at so many things.”

Imran Manji, Uber’s General Manager and Head of East Africa, gave his perspective of the demand side. Ride hailing, he said, is an earning opportunity, not employment, and there is a widespread expectation that it can absorb an unlimited number of drivers every time the formal economy sheds jobs. It cannot. He also argued that any measure of regulatory success must count what the sector has changed in Kenyan life: the nurse who can get home safely after a night shift, the drop in drink driving, the way the city moves. These, he said, are benefits nobody is capturing in the current debate.

Paul Kingori, speaking for the Ministry of Roads and Transport, described Government’s role as the neutral arbiter between drivers and operators who were, in his words, “hearing each other but not listening”. He was candid that the 2022 regulations were written amid a great deal of noise and that he was new in the department at the time. Government’s interests, he said, are fair competition between platforms, safety for drivers and riders, reliability, and affordability for commuters, and the Ministry is benchmarking other countries as it prepares to rebuild the framework.

The Tanzania data point that should give every policymaker pause

The most consequential piece of evidence on the morning did not come from a research deck. It came from Dimmy, who ran Bolt’s business in Tanzania before taking on East Africa, and who had been in Dar es Salaam the day before, reviewing the effects of the regulator’s latest fare decision.

When fuel prices spiked in April, he said, the Tanzanian regulator responded with a new pricing band that lifted blended fares by around 25%. Five months later, demand across the industry is down by around 17%, and for four-wheel cars specifically it is down by 38%. Drivers who celebrated a higher fare per trip are now doing far fewer trips. “Ride-hailing is not a must-have,” he said. “You hike the price, people walk.” He put a number on the sensitivity: in East Africa, a 1% rise in price reduces demand by roughly 0.65%, so a 40% increase in fares would cut demand by close to 30%. That is an operator’s figure rather than an independent one, and it deserves scrutiny, but it is consistent with what TIFA found among Nairobi riders.

He also drew a lesson Kenya should take seriously. Kenya, he said, does stakeholder engagement well before a rule is gazetted. Tanzania does something Kenya does not: it brings stakeholders back three or four months after implementation to review whether the rule is working. Smart regulation needs both.

Imran added a second Tanzanian example. When a regulator in Dar es Salaam asked him why Uber Comfort, the premium category Kenyans know, had never launched there, his answer was that fare ceilings made it impossible. A driver who invests in a better car cannot earn more from it under a capped fare, so the product does not exist. The unintended consequence of price regulation was a market with less choice and less reason for drivers to invest.

The driver’s grievance, in the driver’s own words

The exchange I will remember longest came late in the panel, after Paul Kingori had set out Government’s current thinking and Imran had responded to it at length. I had been watching Daniel write furiously through both, and I gave him the floor.

He did not argue for a higher fare. He argued about the trips nobody pays for. A base fare of around KSh 220, he said, comes to about KSh 180 after commission, and a driver may be sent five or six kilometres to collect a passenger who then travels one. Nobody pays for the dead miles. He described sitting at a petrol station on Haile Selassie Avenue at three in the morning, accepting an airport trip, driving most of the way across the city to reach the passenger, and watching the request cancel as he approached. “Who pays for that loss?” he asked. “We are not in the streets because we want higher prices. We want what we call transparency in the pricing mechanism.”

Then he explained why drivers accept trips they know are unprofitable. On at least one platform, an acceptance score below 70% triggers deactivation, the decision is automated, the appeal runs through a chatbot that gives the same answer three times and then closes the case, and the shortest suspension is fourteen days. A driver financing a vehicle cannot afford fourteen days offline. “If your child scores 70% in KCSE, that is a B plus,” he said. “And you want to tell the driver that below 70% you are gone.” He called it forced labour. He also made a safety point that the operators would later concede: a request from Kibera to Dandora at two in the morning is a trip a driver may have good reason to decline, and should be able to decline without penalty.

What he asked for was not complicated. Let drivers decline trips without punishment. Pay for the distance to the pick-up. Recognize organised driver groups so that there is a table at which these things can be discussed, rather than a street on which they are protested. And, from Government, a legal structure, modelled on the matatu saccos, that would let drivers organise as transport cooperatives, something he said drivers have been asking for since 2018.

Mbugua pushed back on one part of this, and it was a useful challenge. Daniel had put the number of active drivers at around 57,000. That, Mbugua said, is immense collective purchasing power that drivers are not using. Most of the cars on the apps are the same few models. A matatu operator he once worked with bought tyres and filters by the thousand for a single chassis type and cut his costs accordingly. Drivers could do the same with tyres, parts and insurance tomorrow, without waiting for a regulator. “What is in your hand?” he asked. Daniel’s reply was that without a recognized legal entity, 57,000 scattered individuals cannot negotiate with anyone, which is precisely why the cooperative structure matters. Paul Kingori confirmed that the Ministry is considering a sacco or company model for ride-hailing drivers, similar to the one that organised the matatu sector, as part of the review of the 2022 regulations.

I will note in passing that Daniel’s figure of 57,000 drivers is a fraction of the 300,000 to 350,000 that appears in some desk research on the sector. Nobody on the stage could say with confidence how many people actually drive on these platforms. That gap is itself an argument for the shared data that came up later.

What Government is actually proposing, and the numbers behind it

Paul Kingori gave the room more detail on the Ministry’s plans than I have seen reported anywhere, and it is worth setting out carefully, with the caveat he repeated several times that the figures are subject to approval by the Cabinet Secretary.

First, the process. The Ministry has procured a consultant, with World Bank funding, to deliver three things within about five and a half months: a National Taxi Policy covering both conventional and digital taxis, a review of the 2022 TNC regulations, and a model that the regulator can use to advise the industry. He compared that model to the one the Energy and Petroleum Regulatory Authority uses to set fuel prices each month, which prompted the most popular question on the Slido running during the forum: will Government now calculate ride-hailing prices the way it calculates fuel prices? His answer was an unambiguous no. The process, he said, will involve exhaustive stakeholder and public participation, and the court’s judgment requires a regulatory impact assessment before anything is gazetted.

Second, the substance. Government is no longer talking about a minimum fare. It is proposing what it calls driver minimum compensation: a floor on what the driver keeps from each trip, not subject to commission or tax, to cover the cost of being available for the next ride. The current reality, he said, is a minimum fare of around KSh 200 on most platforms, from which the 18% commission and VAT on that commission leave the driver about KSh 174. The industry’s own proposal had been a floor of KSh 220 for small cars and KSh 500 for larger ones. Government rejected both and is proposing KSh 219 for the smallest category and KSh 272 for the larger one, with the platform’s commission and taxes to be added on top. He cited Dubai and the United States as markets where the meter does not start at zero, and he raised the question of whether the 18% commission cap would still be needed once a compensation floor was in place.

That is a signficantly more nuanced position than the one reported in July, when the public debate was about doubling fares to KSh 500. It is also, as Imran pointed out immediately, a position with numbers attached.

Uber’s rebuttal: five years of knee-jerk decisions and the numbers behind a Ksh 219 floor

Imran’s response was the sharpest ten minutes of the morning, and the operators’ case deserves to be heard in full because it is more interesting than “platforms oppose regulation”.

He began with a history. In 2022, drivers protested and the commission was capped at 18%. In 2023, drivers were still protesting, this time about booking fees, and the booking fee was removed. In 2024, drivers were still protesting, and NTSA asked the platforms to raise prices. In 2025, the protests were about receipts. In 2026, here we are. “Five years of making a change every year because of the pressure, and drivers are still protesting.” He was sympathetic to the Ministry, which he said faces drivers arriving with coffins, but his conclusion was that knee-jerk decisions made without assessing their impact leave the sector in the same place or a worse one a few months later.

Then the numbers. A KSh 219 floor for the driver, plus the platform’s service fee, plus the VAT that Government requires platforms to charge on that fee and which drivers cannot recover, produces a minimum fare of around KSh 270. That is a 35% increase on the current minimum. His analogy was a landlord raising rent by 35%: you pay the first month because you have to, and then you start changing your behaviour. You skip the third meeting. You walk.

The second-order effect is the one I had not considered. If a driver is guaranteed KSh 219 for any trip, a two-kilometre trip pays roughly KSh 110 per kilometre and a five-kilometre trip pays about KSh 40 per kilometre. The AA mileage rate for a private car, he noted, is around KSh 33 per kilometre. The rational driver accepts only the shortest trips and declines everything over three or four kilometres, while the passenger paying KSh 270 for two kilometres decides to walk. You end up with an industry in which passengers only want the medium trips that drivers will not take, and drivers only want the short trips that passengers will not request. “It’s not even unforeseeable,” he said. “We can show it in the data.”

He also attacked the “spin of the wheel” model, where the meter starts the moment the car moves, as forty-year-old technology. His argument was that Daniel’s own examples proved the point: a ten-kilometre trip to Kitengela at seven in the evening, with no return fare, is not worth the same as a ten-kilometre trip to the airport, where a tourist and a tip are waiting at the other end. Drivers want to be paid more for trips that are less desirable. A flat per-kilometre floor does the opposite.

On commissions, he made a claim that I have not been able to verify independently and record here as his: that of the roughly 10,000 cities Uber operates in, only Portugal and Tanzania, both at 25%, and Kenya at 18%, cap platform commissions. His point was that Uber’s largest cost in Kenya is promotions to riders, that promotions are what create the demand drivers earn from, and that capping the commission caps the money available to reinvest in that demand. His closing ask was regulatory certainty: a five or ten-year licence rather than an annual one, so that investors can decide whether Kenya works for them and plan accordingly.

Bolt’s numbers on what a driver actually pays to be on the road

Dimmy added the cost structure that neither the driver nor the Ministry had fully laid out. Around 70% of drivers, he said, do not own the vehicle they drive, and some fleet owners charge as much as KSh 4,000 a day. At an average economy fare of around KSh 500, that is eight trips just to cover the car before fuel, insurance or food. Raise prices by 40% and cut demand by close to 30%, and the driver who needed eight trips and now gets seven is worse off than before. Insurance is a fixed cost whether you do one trip or twenty. “Net earnings,” he said, “is not gross minus commission. It is gross minus commission minus every operating cost, including the fixed ones.”

He was also disarmingly honest about who benefits from a price rise. “If you were to increase the price today and maintain the same demand, the first beneficiary is us, the platforms,” he said, since commission is a percentage of the fare. The reason Bolt does not want the increase is that demand will not hold. And on the 18% cap, he said something that will surprise drivers: on some trips, Bolt invests more than 18% of the fare in promotions to generate the request in the first place, so the trip runs at a loss. His parting shot was the most compact summary of the operators’ position I heard all day: fix the marginal cost, and do not pass laws that hurt demand, because demand is the heart of the service.

Both operators committed, on stage, to data. Dimmy said Bolt already shares trip counts, completed rides, average fares and safety incidents with the regulator on request and is open to a real-time integration. Imran said Uber has offered Government unlimited aggregated, anonymized data since the start of the process. I will be watching to see whether those commitments survive contact with a consultant’s data request.

The consumer’s ideas were the most original in the room

Mbugua’s contributions kept pulling the conversation towards what regulation could enable rather than what it should prohibit, and three of his ideas deserve a wider audience.

The first was a cap on driver hours enforced through shared data. Most drivers run every app on their phone. Nobody wants to be driven by someone who has been on the road for ten hours, but no single platform can see a driver’s total hours across apps. A shared data layer could make a driver unavailable for new requests once a shift limit is reached, whichever app the next request comes from.

The second was the Nairobi Expressway. It is seen as an elitist product, he said, and the toll is passed to the rider on every platform. What if a shared ride carrying three passengers in the same direction used the expressway for free? The data exists to verify it. Riders save money, drivers complete more trips, platforms see more demand and the expressway sees more traffic.

The third was aimed at Government as a buyer of services rather than a regulator. Declare Tuesdays and Thursdays no-car days for civil servants and have them share ride-hailing trips into the business district. Government functions are concentrated, civil servants live in the same neighbourhoods, and the demand it would create would do more for driver earnings than any floor. He also wanted a trackable service level agreement for safety incidents escalated beyond the platforms, because that is where, in his words, issues go to die.

On electric vehicles, he proposed that the levies and tolls already collected from the sector fund zero-interest EV loans to drivers in good standing, with repayment drawn from platform earnings data. Paul Kingori’s response was a reminder that Government has already moved further on this than most people realise. He cited VAT zero-rating and excise relief for EVs, the presidential directive waiving import duty on the first tranche of electric vehicles, a credit guarantee scheme for EV purchases that is about to be rolled out, a Japanese financing facility that he put at around KSh 15 billion at a 6% interest rate, and a World Bank-funded financing framework. For the record, the directive as reported in May covers the first 100,000 vehicles, and the Japanese Samurai facility announced in June was reported at KSh 22 billion, of which about KSh 13 billion is earmarked for automotive localisation. The direction of travel is the same whichever figures one uses.

What IEA took away, and what I took away

Jackline Kagume closed the forum with a summary that was, for a convenor, remarkably accurate. The evidence shows demand is highly price-sensitive. Higher fares do not necessarily mean higher net earnings. No single variable can be regulated in isolation. Efficiency of outcomes is the best marker of successful regulation. Drivers can use organisation and economies of scale to cut their own costs. And there is a significant difference between regulating an industry and controlling its prices: regulation of the sector is necessary, pricing is the job of the market, and pricing mechanisms must be transparent. IEA promised a media brief the same day and a detailed stakeholder report the following day.

My own takeaways are four:

1. The fare was never the real grievance: The driver representative said it more clearly than anyone: drivers are not on the streets over the price of a trip. They are on the streets over dead miles, automated deactivation, chatbots that close cases, fourteen-day suspensions and the absence of any table at which to raise these things. Every one of those is fixable without touching the fare, and most are fixable by the platforms without waiting for a regulator. If the operators want to take the fare off the table, that is where to start.
2. Government has moved, and the sector has not noticed: The Ministry’s position is no longer a KSh 500 minimum fare. It is a KSh 219 compensation floor for the driver, a consultant-led review with a regulatory impact assessment, an open question over the commission cap, and a willingness to organize drivers into cooperatives. That is a different conversation from the one the headlines have been having, and it is one the operators can engage with on the numbers.
3. The real regulatory problem is information, not price: John Mutua named it in his opening: the platforms know how fares are set, what the algorithm rewards and what drivers earn, and nobody else does. We could not agree on the stage how many drivers there are. Until there is a shared, verified evidence base, every rule will be contested by the party whose data it ignores, and every forum will end with competing surveys. Both operators committed publicly to providing that data. The consultant should hold them to it, and so should the drivers.
4. The next five and a half months are the whole game: A consultant, a policy, a regulatory review, a model and an impact assessment, all with public participation, all before anything is gazetted, and all with a court decision in the background. Paul Kingori’s closing offer was an open door at the Ministry for anyone with a document and an idea. The sector should walk through it with evidence rather than placards.

Imran said something near the end that I think every stakeholder in that room agreed with, whatever they thought of the rest of his argument. Kenya is the envy of a great many countries when it comes to its technology sector. Regulation that locks in a forty-year-old pricing model, or that blocks the next category or the next vehicle type, would be a self-inflicted wound. Smart regulation, done the way this forum described it, is how Kenya avoids one. The evidence is on the table. The process has a timetable. For once, the people who need to be in the room were all in the room. Now we find out whether they stay there.