Two years ago Kenyans were told that the Social Health Authority would end selling land to pay a hospital bill. One registration, three funds, no exclusions for pre-existing conditions and no age cap. On paper it seemed better than NHIF, and most of us, though skeptical, wanted (no, we needed) it to work.
Two years on, hospitals are owed more than Sh76 billion, pharmacies in level 2 and 3 facilities are padlocked, and a family in Machakos facing an emergency caesarean was asked for cash because they were told at the desk that SHA was not working. A registrar at Kenyatta National Hospital who treated patients for free was refused treatment at his own hospital because he could not pay and had no functioning cover. He was referred out, deteriorated and died in March.
The official explanation is fraud. And sure, fraud exists. The Auditor-General confirms it: 3,235 repeat or unapproved surgical cases worth 445 million, including open-heart surgery billed four times on one patient in a single day, and 148 million in multiple-delivery claims for women apparently giving birth more than once at the same moment.
But notice what that proves. The same system that paid promptly for operations that could not possibly have occurred has simultaneously not paid a dispensary in six months. Fraud did not necessarily make SHA slow. It was slow and leaky at the same time, in opposite directions, which very much likely is a design problem.
The design problem was as well knowable before launch. SHA's benefits and tariffs were priced for a country where everybody contributes. Collection was then built almost entirely around payslips. In 2024/25 the contributory fund took in 57.7 billion against 91.5 billion in claims, a utilisation ratio of 158.6 percent. For every hundred shillings collected from members, the scheme spent a hundred and fifty-nine.
Universal Health Coverage
Of that 57.7 billion, salaried workers supplied roughly 52 billion. Ninety percent. The informal sector, most of this country and the stated reason for building SHA in the first place, contributed 5.7 billion. Means testing and Lipa Pole Pole came afterwards, once it was impossible to pretend that the model matched how Kenyans actually earn.
SHA reports 32.2 million registered, which is about 68 percent of the population. Only around 5 million are paying. Registration became the measure of success because it was the only metric that would look good for a government run purely on optics. Thirty-two million enrolled with five million contributing is not exactly universal health coverage, however much you would like to pretend it is. It is very much a payroll deduction on an already strained payslip.
The SHA script is quite a spectacular failure, when you think about it. Announce an ambitious scheme. Price it for conditions that do not exist. Fund it from the narrowest base available, and when it fails to pay, you blame the beneficiaries. Then procure an expensive private fix.
And that last step is precisely why sometimes it’s hard to argue against this being deliberate. SHA's digital spine was procured at roughly Sh104.8 billion from a Safaricom-led consortium under a specially permitted procedure, not an open tender. Auditor-General Nancy Gathungu found it in neither the procurement plan nor the budget framework.
What exactly did Sh104.8 billion buy? Not the system. The consortium retains ownership and the intellectual property; the state is barred from building a competing platform, and disputes go to arbitration in London. The fees are a share of the flow: 2.5 percent of every contribution and 5 percent of every claim, projected past Sh111 billion over ten years, into escrow the Kenyan people do not control.
Meanwhile, the facilities absorb the shortfall. Controller of Budget figures show county facilities owed Sh8.3 billion in December and 26.87 billion by March, tripling in a quarter. SHA now collects 5.4 billion monthly against 8.7 billion in claims. Of 93 billion submitted since launch, roughly 50 billion has been paid.
Gathungu’s office queried 49.3 billion in total, including 26.8 billion in unsupported claims, a 3.4 billion transfer nobody could trace, and 1.3 billion paid into NHIF's account after NHIF had ceased to exist. Duale's answer is that the Sh26.8 billion is an outstanding-claims reserve, standard insurance accounting rather than money stolen. Grant him that entirely. It still does not explain the surgeries, the untraced transfer, or the payment to a dead institution.
In September, the ministry announced the provider portal would die at midnight on the 30th, with every facility required to migrate or stop serving SHA patients. Private and faith-based facilities said they were not ready, and on 1 October the deadline was extended. Through all of it, officials cite a 37 billion operating surplus. Hospitals cite closed laboratories and cancer patients whose medicine sits in the building, locked away because the system will not generate an invoice. Those two statements have never been reconciled.
So ask yourself the questions this raises. Who benefits from a scheme priced for universal contribution but funded by payroll? Who benefits from a technology contract awarded outside competitive tender, privately owned, paid as a percentage of every contribution for a decade? And who absorbs the gap when the invoice will not generate, and a family member is in desperate need of immediate medical attention?
Healthcare is one of any society’s key social services, and we are experiencing a system, probably the single most expensive one in our country’s history, succumbing to very entry-level fraud. Either the system is broken, and everyone running it is totally clueless, which is very hard to argue for, or, and this is very likely the case…there are people heavily profiting off ensuring the system doesn’t work properly. Whichever way, what a terrible situation!
