Build Now, Bleed Later: PFI Is Back From The Dead

In October 2018, Philip Hammond stood up at the dispatch box and announced the government would no longer use PF2, the latest version of the Private Finance Initiative. The language was final. PFI had become a byword for expensive hospitals, inflexible contracts, off-balance-sheet accounting, and the broader New Labour habit of buying things the country could not afford by pretending they were free. The model was dead. The Chancellor said so. Parliament moved on.

Seven years later, in June 2025, the government published its UK Infrastructure: A 10 Year Strategy. It promised at least £725 billion of public investment over the coming decade. It created yet another useless new body, the National Infrastructure and Service Transformation Authority (consolidating two other useless quangos, the National Infrastructure Commission and the Infrastructure and Projects Authority), to oversee delivery. And tucked inside the strategy, expressed in careful bureaucratic prose, was a familiar proposition: the state would explore new public-private partnership models to mobilise private capital for public infrastructure.

Not PFI. Never PFI. The strategy was explicit on this point. There would be no return to PFI or PF2. But the underlying machinery being assembled, quietly, across Whitehall and the City, is designed to do precisely what PFI once did: build now, hide the cost, bind future governments, and let private finance collect long-term contracted returns from things citizens cannot refuse.

PFI did not die because the logic was wrong. The label became unusable. And bad policy, once it loses its name, does not disappear. It gets a rebrand.

What PFI Actually Was

For anyone who missed the original scandal, or who has been told PFI was merely a technical procurement method, the explanation is simple. And the simplicity is the point. The legendary magazine Private Eye did the most comprehensive demolition of it anywhere in print.

1. A public body needed a hospital, a school, a road, a prison.
2. Instead of the government borrowing money and building it directly, a private consortium would finance the project, build the asset, maintain it, and sometimes operate parts of it.
3. The public body then paid annual charges, called unitary payments, over 25 to 40 years.

Put simply: private companies built things up front in exchange for politicians guaranteeing taxpayers would rent it back from them. It didn't need to be recorded as public spending.

The official justification ran like this: private firms would absorb construction risk, deliver on time, embed whole-life maintenance, and bring commercial discipline. Government borrowing would fall, or at least appear to fall, because the debt sat with the private consortium rather than on the Treasury's books.

Some of this was real. PFI projects were often delivered on time and on budget, a claim the National Audit Office has broadly upheld. Maintenance was locked in contractually, unlike publicly owned buildings where maintenance budgets are routinely raided to cover revenue shortfalls.

But the political magic of PFI was never really about construction discipline. It was about accounting.

The Public Accounts Committee, in its 2025 report on private finance for infrastructure, finally said this plainly. HM Treasury's focus on ensuring investment in public assets was not recorded on the government balance sheet, the committee found, gave the illusion of lower levels of debt. This, it said, was a key driver of private financing decisions and detracted from value for money.

The parliamentary committee responsible for scrutinising public spending concluded PFI's primary motivation was cosmetic. The state wanted buildings without visible borrowing. Private capital wanted predictable long-term yield. PFI married the two, and future taxpayers paid for the wedding.

How £60 Billion Became £199 Billion

By the time the NAO conducted its landmark 2018 review, the PFI estate looked like this:

Those numbers are still live. Every one of those contracts continues. The 2018 abolition applied only to new deals. Existing obligations run until the 2040s and, in some cases, the 2050s. PFI may have been cancelled, but ordinary taxpayers are paying for it for at least another twenty years.

A £60 billion building programme generated almost £200 billion of future payments.

Some of this reflects maintenance, lifecycle costs, and services bundled into PFI contracts, not pure capital repayment.

But the ratio tells you something important about how the model actually worked: the state got infrastructure at the point of decision for a price it could politically tolerate, then locked in decades of payments at a cost it could not easily escape.

In the NHS alone, PFI repayments hit £2 billion a year. Some trusts spend a fifth of their budget servicing PFI debt. At Barts Health NHS Trust in London, a £1.1 billion rebuild generated a total contractual cost exceeding £7 billion, with annual repayments of over £140 million running until 2049. The trust has been in and out of financial distress for years.

PFI did not just build hospitals. It created annuities. And the annuity holders were not the patients, or the nurses, or the local health economies. They were the special purpose vehicles, the banks, the equity investors, and the facilities management companies contracted to supply cleaning, catering, portering, and building maintenance at prices fixed decades in advance.

Liverpool provides the starkest example of PFI's mechanical absurdity. The city council pays £4 million a year for Parklands High School, a building constructed under PFI and then closed because pupil numbers fell. The school is shut. The payments continue. The contractor, Equity Solutions, still posts annual profits from the deal. Roughly £42 million in repayments remain outstanding on a building nobody uses.

This is what happens when public infrastructure is financed as a private income stream: the income stream survives the public need.

It is also the plot of an episode of "Yes, Minister."

Why It failed On Its Own Terms

The conventional critique of PFI is moral. Public services should not generate private profit. Hospitals are not investment products. Schools belong to communities, not to special purpose vehicles registered in Jersey.

All of this is true. But PFI also failed technically, on the criteria its own advocates set out. The failures were not random. They were (bingo moment)... structural.

Procurement was ruinously slow and expensive

PFI bids required years of negotiation between public bodies, construction firms, banks, equity investors, lawyers, and consultants. Each deal was bespoke. The transaction costs alone could run into millions before a single brick was laid.

Contracts were rigid over timescales no one could predict

A 30-year contract for a hospital signed in 2000 could not anticipate changes in clinical practice, demographic shifts, NHS reorganisation, or a global pandemic. Modifications were possible but expensive, because the contract was also a financial instrument. Changing the building meant repricing the debt.

Risk transfer was partly fictional

Government believed, or claimed to believe, risk had been transferred to the private sector. In practice, the state remained the backstop. When Carillion collapsed in January 2018, taking PFI contracts and public service obligations with it, the government had to step in to keep hospitals, schools, and prisons running. The PAC now notes this directly: risks should sit with whoever can manage them, but inappropriate transfer becomes disproportionately expensive, and government may still need to intervene when a private partner fails.

Risk, in other words, was not sold. It was rented back at a higher price.

The public sector was outmatched at the negotiating table

Departments and local authorities were routinely up against better-resourced, more experienced commercial teams from the private sector. The PAC identified poor risk pricing, poorly drafted contracts, and inadequate public-sector commercial capability as persistent problems. The state was buying complex financial products without the skills to understand what it was signing.

And the accounting distortion corrupted decision-making

If the primary attraction of PFI was keeping debt off balance sheet, then decisions about how to build hospitals and schools were being driven not by what was best for patients or pupils, but by what made the fiscal position look least alarming. This is not a conspiracy theory. It is what the Public Accounts Committee concluded in its own report.

A Country With Short Memory And A Long List

All of this was known, documented, and debated when PFI was abolished. So why is the same logic returning? There is no mystery. There is no degree needed to establish the simple mechanics and mathematics.

Because the pressures have not changed. If anything, they have intensified.

Britain faces a maintenance backlog across its public estate estimated at £49 billion.

• The New Hospitals Programme, originally promised as 40 new hospitals by 2030, has been beset by delays, with some not expected to begin construction until 2039.
• School buildings are crumbling.
• Courts are falling apart.
• Prisons are overcrowded.
• The energy system needs rebuilding.
• The transport network needs expansion.
• Housing targets require entirely new settlements.

The fiscal space is tight. The new fiscal rule, based on public sector net financial liabilities, gives marginally more room for capital investment, but not enough to cover everything. The political incentive is obvious: find a way to build things without the borrowing showing up in the headline numbers.

This is the environment into which the 2025 infrastructure strategy arrives, along with the worst defence crisis in British living memory and a national debt about to reach 3TN. And the strategy is not subtle about where it is heading.

The government has created at least five public financial institutions, including the National Wealth Fund (by renaming the UK Infrastructure Bank), to deliver large-scale financial transactions. It has established NISTA to develop and oversee a pipeline of infrastructure projects. It has published a digital infrastructure pipeline, initially listing 780 schemes worth £530 billion, since updated to 734 projects worth £718 billion. And it has committed to exploring public-private partnerships for specific types of social infrastructure, including community health facilities and public estate decarbonisation projects.

It has written reports. It has collected statistics. It has measured things. It has renamed something. It has merged quangos. It has plans and promises.

What it has not done is actually build anything.

The strategy says a decision on new PPP models will follow the Autumn Budget. It says any model will reflect lessons learned. It insists there will be no return to PFI or PF2.

But look at the architecture being assembled.

A long-term pipeline giving investors visibility over future projects. A new authority developing what it calls "investable models." Public financial institutions providing guarantees, equity, and risk mitigation. A fiscal framework designed to accommodate financial transactions without breaching borrowing rules.

This is the infrastructure of a new PFI, built without the name.

Investable Models For Captive Populations

The most revealing phrase in the entire policy apparatus comes from NISTA itself. Giving evidence to the Public Accounts Committee, NISTA said the challenge was not so much "working up the finance" but developing "investable models." Its goal, it explained, was to develop an investable set of projects capable of attracting domestic and international capital.

Let's pause on the stupid abstract jargon here.

A hospital is not, to a patient, an "investable model." A school is not, to a parent, a component of a capital pipeline. A road is not, to a commuter, a risk-adjusted revenue stream. But to the institutions and funds being courted by NISTA, this is precisely what they are.

The astute reader will have spotted the problem by now: quangocrats with no business acumen trying to build the plane while flying it.

Yet again, the civil service is up to its neck in something it is not qualified to do, and has failed at, catastrophically, over and over, repeatedly.

The infrastructure pipeline, updated in March 2026 to include new project metrics on potential investment opportunities, now provides detail on the type of investment sought, the amounts involved, and the business models to be used. NISTA's CEO, Becky Wood, described the pipeline as giving investors:

more granular information about forthcoming investable opportunities they told us they need for their strategic planning.

The public sees hospitals, schools, and roads. The market sees cashflows.

Wood's background is exactly what you'd expect. Entirely appropriate for a quango; entirely inappropriate for common sense and reality. She is the person in HR elevated to national incompetence.

she was a partner at the consultancy firm EY, and prior to that was a Commercial Advisor at the Infrastructure and Projects Authority. For ten years Becky oversaw major infrastructure developments at the Department for Transport, serving as the Senior Responsible Officer for the Crossrail, Thameslink and Intercity Express programmes. 

• Crosslink cost £18.9bn and was 2.5 years late.
• Thameslink cost £7.5bn with infrastructure rising about 9.4%.
• Intercity Express cost £7.65bn, was 2.5 years late, and the planning was useless.

There is nothing inherently wrong with private investment in infrastructure. Canada, Australia, and several European countries use PPP models with varying degrees of success. Private capital can bring speed, discipline, and long-term maintenance commitment. The question is not whether private money should ever touch public infrastructure. It is whether Britain, in its current fiscal and institutional condition, is capable of using private finance without repeating the same mistakes.

The PAC does not think the evidence is reassuring. Its 2025 report found:

• The Treasury had not yet set out principles for matching private financing models to suitable infrastructure projects.
• No comprehensive evaluation of the costs and benefits of different financing approaches. No central monitoring of private finance contracts outside PFI.

In other words: the government is building the machinery before proving the model.

Again.

Same Pressures, Cleaner Vocabulary

The vocabulary has changed. The incentives have not. The continuities between old PFI and the new PPP environment are a cunning set to re-frame language while the state continues the same foolish practices.

The structural danger is identical. When the state uses private finance not because it is genuinely better value, but because it allows politicians to open buildings without admitting the full cost, the result is the same: present-day ribbon-cutting, future-day billing.

The PAC saw this clearly. Its report warned against a repeat of past mistakes: inefficient procurement, poor contract drafting, poor risk pricing, inadequate capability, and above all, the use of financing structures chosen to manage the perception of debt rather than to deliver value.

Honest Borrowing Vs Convenient Fiction

The scandal is not, and has never been, the involvement of private capital in public infrastructure. Capital is capital. It does not care whether it builds a wind farm or a warehouse. What capital requires is return, and the question for any society is whether the return being offered is justified by real risk, genuine efficiency, and transparent cost.

The scandal is the opposite: a state so fiscally constrained, so politically frightened of honest borrowing, and so administratively hollowed out it repeatedly converts public necessity into guaranteed private income while pretending the bill has vanished.

Britain does need infrastructure. It needs hospitals, schools, housing, energy systems, transport, digital networks, and courts. The maintenance backlog alone is a national disgrace. And public capital budgets, while significant, are not sufficient to cover everything.

But financing method matters. If private finance is chosen because it genuinely delivers better whole-life value, with risk properly priced, contracts properly drafted, and the public sector properly equipped to negotiate, then the argument is defensible.

If it is chosen because it reduces visible borrowing while increasing future obligations under a different label, it is an accounting trick. And Britain has tried the accounting trick before. The bill is £199 billion and still running.

A serious country borrows honestly, builds competently, maintains what it owns, and accounts for its obligations in plain sight. A weak country brokers deals, hides liabilities, outsources capability, and calls the result "partnership."

PFI is not returning as a policy. It is returning as a habit. The language is cleaner. The institutions are newer. The pipeline is digital. But the essential proposition remains: build now, pay later, keep the cost out of the headlines, and let someone else worry about the invoice.

The country has seen this film before. It is still paying for the tickets.

Tomorrow, in Part II: how the state discovered it could hide taxation inside utility bills, and why regulated asset base models are converting citizens into captive revenue streams.