In 2014 Oxfam published a report on a new hospital in Maseru, Lesotho. Queen ‘Mamohato Memorial Hospital had been built and run under a public-private partnership designed with advice from the International Finance Corporation, the World Bank’s private-sector arm, and was pitched as a model for Africa. Oxfam’s analysis found it was consuming more than half of the Ministry of Health’s budget, under an 18-year contract, in a country where three-quarters of people live in rural areas served by primary care.
Five years earlier, a very different partnership had started. In 2009 a group of donors including the Gates Foundation, Canada, Italy, Norway, Russia and the UK backed an Advance Market Commitment through Gavi to bring pneumococcal vaccines to poor countries far sooner than the usual 10 to 15 year lag behind rich ones. Gavi reports that children in 60 eligible countries have since been immunised against the leading cause of pneumonia.
Both are called public-private partnerships. They have almost nothing in common. And the difference between them is the most useful thing a health ministry can learn before signing a PPP contract.
Two meanings of one phrase
The label “PPP” covers two separate ideas.
One uses private capital to build or run public infrastructure: hospitals, labs, equipment. The state pays the private partner over many years through availability charges or service fees. The attraction is that construction happens now and the cost is spread out.
The other uses public or philanthropic money to change a market: to make it worth a company’s while to produce something that would not otherwise be made at a price poor countries can afford. The pneumococcal commitment is the classic case. Donors promised to top up the price for a guaranteed volume if manufacturers built capacity and committed to a long-term capped price.
My argument is that the second type has a much better record in global health, and the first is often a way of borrowing that does not look like borrowing.
Why infrastructure PPPs so often disappoint
The British experience is the most documented. The UK’s National Audit Office reported in January 2018 that there were more than 700 operational PFI and PF2 deals across government, with a capital value of around £60 billion, annual charges of £10.3 billion in 2016-17, and future charges running into the 2040s totalling £199 billion. Hospitals were a significant part of that portfolio. Later that year the government announced it would not use PFI for new projects.
The core problem is not that private companies are greedy. It is that the contracts turn a flexible public budget into a rigid one. A ministry can cut a hospital’s budget in a bad year. It cannot easily cut an availability payment written into a 25-year contract with lenders behind it. When revenues fall or priorities change, the fixed payment stays and everything else gets squeezed.
That is what happened in Lesotho. The hospital may have been good. But a contract that locks in a large, rising share of a small health budget crowds out the clinics that most people actually use.
There is also an accounting temptation. Private finance can keep the upfront cost of a hospital off the public balance sheet, depending on the rules in force. A minister under a debt ceiling can open a building now and leave the bill to successors. The partnership is, in effect, off-book borrowing at a higher interest rate.
Why market-shaping deals work better
The vaccine commitment worked for a different reason. It paid for an output that was clear and measurable: doses of a specified vaccine, at a capped price, delivered to eligible countries. The private partner took on the risk it was best placed to manage, which was building manufacturing capacity. The public side took on the risk it was best placed to manage, which was demand.
Gavi reports that a 2020 supply agreement brought the price of one pneumococcal vaccine to US$2 per dose. That kind of result is only possible when the payer is aggregating demand across dozens of countries and using that scale as leverage.
Market-shaping partnerships also end naturally. Once a market is working, the subsidy can taper. An infrastructure contract, by contrast, often outlives the government that signed it by a decade or more.
The mechanism: the price of transferring risk
Why do infrastructure PPPs so often cost more than expected? The core reason is that private finance is more expensive than government borrowing, and the extra cost has to be justified by something.
A government can usually borrow more cheaply than a private company, because lenders see it as less likely to default. When a private consortium builds a hospital, it raises money at a higher cost and passes that cost into the payments the government makes over the contract. In principle, the premium buys something real: the transfer of risks such as construction overruns, delays and poor maintenance from the public sector to the private partner, which is supposed to manage them better.
The value of a PPP therefore depends on whether the risks transferred are worth the extra cost of finance. When they are, the deal can be good value. When the government ends up bearing the important risks anyway, because the contract is badly written, because the private partner can walk away, or because a hospital is too important to be allowed to fail, the government pays the premium without getting the protection.
Health care makes this especially tricky. The biggest risks in running a hospital are clinical and demographic: changes in what care is needed, how it is delivered and how many people need it. Private partners in infrastructure PPPs rarely take on those risks. They take on the building and, sometimes, facilities management. The risks that matter most stay with the state, while the contract makes the state less able to adapt.
A scenario: the hospital that could not change
Imagine a hospital built under a 25-year PPP contract, designed around the model of care that was standard when it was planned: large inpatient wards, a big emergency department, extensive on-site diagnostics.
Ten years in, the health system’s strategy shifts. More care is to be delivered in the community, lengths of stay are falling, and some services are to be consolidated in a regional centre. In a publicly owned hospital, managers could convert wards, lease space to community services or close a wing. Under the PPP, every change to the building or its use requires negotiation with the private partner, whose revenue is tied to the contract as signed. Changes are possible, but they are expensive, and the partner has a strong negotiating position.
So the hospital keeps operating in a way the strategy says it should not. The availability payments continue regardless. The health authority, meanwhile, has less flexibility to fund the community services it now needs.
This is the hidden cost of infrastructure PPPs in health: they lock in not just a payment, but a model of care, for longer than most health strategies last.
The strongest counterargument
The best defence of infrastructure PPPs is that public projects often do worse. Publicly procured hospitals can run late and over budget, and, once built, are often poorly maintained, because maintenance budgets are among the first to be cut. A PPP contract that makes the private partner responsible for keeping the building in good condition for 25 years can prevent that decay. The whole-life costing forces the question of maintenance to be addressed at the start.
This argument has substance. Deferred maintenance is a genuine problem in many public health systems, and buildings allowed to deteriorate eventually cost far more to repair. If a PPP guarantees maintenance, that has real value.
But the argument proves less than its supporters think. The benefit comes from committing to maintenance, not from private finance. A government could create the same discipline through ring-fenced maintenance budgets or through design-build-maintain contracts financed with public borrowing. The question is whether the higher cost of private finance is the cheapest way to buy that discipline. Often, it is not.
There is also a defence of market-shaping partnerships that deserves scrutiny. Critics argue that advance commitments can overpay established manufacturers for products that would have come to market anyway. That concern has been raised about the pneumococcal commitment. Even the good model has trade-offs, and its design, including price levels and which suppliers benefit, matters as much as the concept.
Second-order effects
PPPs shape the health system beyond the individual contract.
They can concentrate spending in hospitals. Because PPPs work best for large capital projects, they bias investment toward big buildings rather than the dispersed, smaller investments that primary care needs. Lesotho is the extreme case, but the bias is general.
They create fiscal risk that is hard to see. Obligations under PPP contracts may not appear as debt in public accounts, depending on accounting rules. That can make a country’s fiscal position look stronger than it is, and can lead to more commitments than the budget can bear.
They create a renegotiation market. Long contracts almost always need changes, and each change is a negotiation between a government under pressure and a partner with expert lawyers. Over time, renegotiations can shift value toward the private partner.
What this means for each actor
- For health ministries: before choosing a PPP, compare it honestly with public borrowing plus a maintenance commitment. If the PPP is chosen mainly because it keeps debt off the books, that is a fiscal decision, not a health one.
- For finance ministries: count PPP obligations as the long-term liabilities they are, and cap the share of any sector’s budget that can be committed to them.
- For development finance institutions: judge health PPPs on their effect on the whole health budget, not only on whether the building was delivered on time.
- For hospital leaders: if you are running a PPP facility, invest in contract management expertise. The partner will have it.
What to watch next
Watch whether development banks continue to promote hospital PPPs in low-income countries, and whether they publish full contract terms. Watch for new market-shaping deals in areas such as diagnostics and new vaccines, and whether they learn from criticisms of past commitments. And watch whether governments start disclosing total PPP liabilities in health budgets.
A short test for any PPP proposal
Before signing, a health ministry could ask four questions:
- What exactly is being bought? A countable output, like doses or tests, or a vague bundle of “services” that is hard to price and harder to renegotiate?
- Who holds which risk? The private partner should carry risks it can control. If the state still carries demand risk, cost overruns and political risk, private finance is just expensive borrowing.
- How much of the budget is locked in, and for how long? If the answer is a double-digit share for 20 years, the partnership is setting health priorities for future governments.
- What happens if the partner fails? The NAO report was completed just before the collapse of Carillion, a major PFI contractor. Failure is not hypothetical.
The takeaway
Public-private partnerships are not good or bad as a category. The ones that worked in global health mostly used public money to change what private firms found worth doing, then got out of the way. The ones that failed mostly used private money to let governments build now and pay later, and then discovered that “later” arrives on a fixed schedule.
If a PPP’s main selling point is that the cost does not show up in this year’s budget, that is not a benefit. That is the warning.
Have you seen an infrastructure PPP in health that genuinely delivered value for money? What made it different?
