Advice to the board: do not build it alone

Do not proceed with a new build general hospital in Nigeria owned and operated as a wholly private company on the terms currently proposed.

The objection is not to the market, the country, the specialties, the ambition or to building rather than buying. Each of those survives examination, and building is the right instinct for a group whose product is demonstrated operating performance. The objection is to the combination of a wholly owned private structure, a self pay model and an opening staffed by outbidding, committed to before the two things that would make it defensible exist: a contracted academy and a measured paying population.

Instead, and in this order: buy the measurement, because it is cheap and settles four questions at once; open conversations with the partners who would change the structure, meaning other hospitals and government, because that is the avenue most likely to answer the grounds rather than mitigate them; and contract the academy independently of whether a hospital is ever built, because it is the only intervention here that is unambiguously good for the place.

If the board proceeds anyway, and it may have good reasons to, the last section sets out the minimum conditions under which the decision is defensible. They should be read as the price of overruling rather than as a compromise.

Seven grounds, three of which are treated as decisive on their own. The third field of each is the important one, because advice that cannot be overturned is a position rather than an analysis, and every ground below states the specific thing that would answer it.

The first two are ethical and legal and they are not the same weight as the others. The remaining five are commercial, and the board should notice that they point the same way. That convergence is the reason this paper exists: it is unusual for the ethical argument and the financial argument to agree, and when they do it is worth stopping to ask why.

They agree here because all of them are downstream of the same fact. The group would be committing capital before it has established the two things the entire proposition depends on, which are whether the paying population is the size the case assumes and whether the repayment mechanism for the staffing debt will exist at all.

This ground deserves separating because boards routinely hear legal risk as a matter for insurance and process, and this one is neither.

The statutory position is unambiguous: no establishment, public or private, may refuse a person emergency medical treatment for any reason whatsoever. The commercial logic of a self pay hospital is equally unambiguous: service follows payment. Those two sentences meet at the front desk, at night, when somebody arrives who cannot pay and is dying.

It cannot be designed out, because it is not a defect in the design. It is the point at which the business model and the law disagree, and the disagreement recurs indefinitely. Every other hospital in the market resolves it the same way, which is by breaking the law quietly, and the fact that the provision is unenforced is precisely what makes a policy of complying with it a commercial disadvantage rather than a baseline.

So the board is not being asked to accept a risk. It is being asked to accept a permanent conflict between its obligations and its economics, and to resolve it in advance, in writing, with money behind it. A group that has not done that before opening has decided to resolve it case by case at three in the morning, by whoever is on the desk, under pressure, with no provision. That is how the offence actually gets committed.

The provision worth noting, and it is the uncomfortable part, is that the penalty is small enough to be irrelevant to a hospital budget. Nothing about this ground is driven by the risk of prosecution. It is driven by the fact that the group would be operating a model whose ordinary functioning requires it to break a criminal law about dying people.

Advice against something is worth very little without a use for the money, so here are four, and they are not mutually exclusive.

One option is deliberately absent and it is worth saying why, because it is the obvious suggestion and it fails on the group’s own test. Buying an existing hospital would convert a design problem into an improvement problem, which is cheaper and faster and lower risk, and it would supply a measured current state instead of an assumed one. It would also mean inheriting an entitlement somebody else set. The walking distances, the plate depth, the lift core and the expansion capacity would all be fixed before the group arrived, and the method page establishes that no amount of improvement work reaches past them.

For most buyers that is a tolerable trade. For this one it is not, because the product is demonstrated, transferable operating performance and the exit condition is best in class. A group whose entire proposition rests on proving what it can achieve cannot afford to have its ceiling set by a building it did not design, and would spend the asset’s life explaining that its numbers are limited by decisions taken before it owned the place. Acquisition is therefore ruled out, not because it is a poor deal but because it is incompatible with what this company sells.

There is one target where the temptation is much stronger than that, and it is worth naming because it is the case that would tempt a board. The teaching hospital this study uses as its local comparator has the two assets the rest of this work says are hardest to obtain. It sits on acreage of a size that the land page shows cannot be assembled in the dense core at any price a hospital can pay, and it comes with a standing pool of trainees, accredited programmes, placement capacity and faculty, which is the academy problem already solved rather than contracted for.

Put those together and it looks like a shortcut through three of the grounds at once: land, the repayment mechanism for the staffing debt, and an institutional reputation that trust otherwise takes years to build.

It is also near impossible, and the reason is ownership. A federal teaching institution is not an asset that trades. There is no seller with authority to sell, no process through which an offer would be considered, and a change of control would need political decisions at a level no hospital group can procure. So the question is largely academic before any of the rest of it arrives, and a board should hear that first rather than spend a year discovering it.

It is just as well, because the same facts that make it attractive are the ones that would be destroyed by acquiring it. Both assets exist because the institution is public. The acreage was assembled by a state that could assemble it, and the trainee pipeline exists because the hospital carries a public teaching mandate. A transfer into private hands would not move those assets, it would dissolve them, and the extraction charge this paper treats as decisive at the level of individual consultants would arrive at the scale of a whole institution. The patients who cannot pay would lose not a surgeon but the hospital.

Which is the argument for the partnership chapter rather than against the observation. Everything genuinely attractive there, the trainees, the faculty, the accreditation, the placement capacity and the adjacency, is available through partnership without any of it changing hands. The hospital a board would most want to buy is the hospital it should most want to partner with, and that is the same conclusion the next chapter reaches from a different direction.

Which leaves the first row as the one to press hardest, and it is examined in its own chapter below.

The second tests whether anybody meant it. If the academy is genuinely the repayment for the staffing debt, it has value independent of the hospital, and funding it alone is the strongest thing this group could do for the place it says it wants to serve. If it only makes sense inside a package that also creates the debt, it was a justification rather than an obligation, and the board should establish which before relying on it.

The fourth should be taken whatever else is decided. Nothing has been optioned, so the decision is fully reversible today, and the cost of establishing the one number that moves everything is trivial against the capital at stake.

This is the avenue that would change the advice, and it is flagged here rather than worked through, because what it needs is a negotiation and a governance review rather than another chapter.

Read the seven grounds again and notice how few of them are properties of building a hospital in Nigeria. They are properties of one structure: a wholly owned, for profit, self pay operator that hires its own staff, grades its own commitments and sells the asset at maturity. Change what stands behind the hospital and several of those stop arising rather than being mitigated.

Other hospitals are the first partner worth approaching, and the mechanism is joint clinical appointments. If consultants hold sessions across both institutions instead of moving from one to the other, the extraction charge does not need repaying, because the extraction does not happen. The clinician is shared rather than taken, the public patients keep their surgeon, and the group still gets the seniority it cannot open without. That is the same partner relationship the academy already requires, doing two jobs instead of one, and it is the single most promising idea in this paper.

Government is the second, and it reaches a different ground. A public partner can bring land, a licence route, regulatory alignment and, most importantly, patients who are not paying out of their own pocket. That converts the free care commitment from charity, which a board can quietly reduce in a difficult year, into a contract with a counterparty who will notice. It answers the exclusion charge structurally rather than philanthropically, which is the only durable way it can be answered.

And a governance structure that puts the commitments outside the people who benefit from relaxing them, whether through a foundation holding a stake, an entrenched provision or an independent trustee with consent rights, answers the objection the ethics paper concedes has no other solution, which is that the group currently grades itself on all of it.

The cost is real and should be stated rather than discovered. Partners want control, they slow decisions down, they take margin, and a partnered asset is harder to sell and to a narrower field of buyers, which cuts directly across the model this group is built on. Those are exactly the things the board would be trading, and they are worth trading only if the grounds above are taken seriously.

None of this is a recommendation here, because none of it has been tested against the capital, the jurisdiction or the appetite of any actual partner, and a structure proposed without that work is a wish. What this paper says is narrower and firmer. The advice is against building it alone on these terms. Whether it should be built at all is a question that a partnership review may well answer the other way, and that review is cheap, quick and has not been started.

A board may properly disagree with all of the above. Several of the grounds rest on judgements about a market nobody has measured, and a group with capital and conviction has beaten a paper like this before.

So these are the conditions under which the decision is defensible if taken against this advice. They are written in the same form as the group’s values, because the test is the same: a commitment that cannot state what it costs and how you would know it was broken is decoration.

The last condition is the one the board will like least and it is the one that matters most over the life of the asset. This decision will be examined again, by a buyer, by a regulator, possibly by a journalist, and certainly by whoever is running the hospital in twenty years. What protects the board is not having been right. It is a record showing what was known at the time, what the argument against was, and why it was rejected. A decision minuted as though no case against existed is the one that looks worst later, and it looks worst precisely when things have gone wrong.

Five things, and any two of them together would probably reverse the recommendation.

A measured paying population at or above the level the demand case assumes. That single figure answers the third ground outright and materially weakens two others, and it is the reason the survey is the first thing to buy.

A signed academy partnership with placement capacity and a start date. That converts the first ground from an unanswered charge into a priced debt with a repayment schedule, which is a completely different proposition ethically and commercially.

Evidence that a buyer of this asset class would pay a premium rather than a discount for the published commitments. That removes the sixth ground and would make the ethical position an asset rather than a cost, which is a result this study would very much like to be true and currently has no evidence for.

And the one most likely to do it on its own: a partnership review that finds credible counterparties on acceptable terms. Joint appointments with another hospital would answer the first ground outright rather than repaying it. A government partner bringing non self paying patients would answer the exclusion charge structurally. A governance lock would answer the enforcement objection. That review has not been started, it is inexpensive, and of everything on this list it is the thing that would most change the answer.

This document argues one side on purpose, and it should be weighed as what it is.

It does not restate the case for the project, which is set out at length elsewhere on this site and is genuinely strong in places: an observed demand signal rather than a forecast, a specialty concentration named by where money already goes, a form and a method that survive examination, and a staffing sequence that is honest about its own debt. A board reading only this paper would be as badly served as one reading only the recommendation.

It also relies on the same unmeasured term as everything it criticises. If the paying population is much larger than assumed, several grounds weaken at once, and this paper would have been wrong in the same direction and for the same reason as the case it opposes.

What it is for is narrow. The instruction excluded not proceeding from the scoring and said so. An exclusion that is never argued is a decision taken by omission, and this is the argument, written down, so that the board rejects it deliberately if it rejects it at all.