Nearly £200 billion — the bill Britain locked in
When the UK's independent spending watchdog finally tallied the Private Finance Initiative and its successor PF2, the number wasn't a forecast of what might go wrong. It was the contracted, already-owed cost of what had already been signed.
That is the same trick you've read on every page here, just at national scale: a cost kept off today's books and turned into a multi-decade stream of payments the public cannot walk away from. Britain didn't get a different deal from Australia. It got the same deal, more of it, and kept the receipts.
Sources: New Statesman — NAO puts the UK's PFI bill at nearly £200bn (Jan 2018, reporting the National Audit Office review); The Conversation — "PFI has been a failure and Carillion is the tip of the iceberg".
40% and 70% — read these numbers carefully
The cost premium is where PFI critics get sloppy and defenders get to change the subject. So here is the careful version. The National Audit Office did not declare that all PFI costs a fixed percentage more. It gave illustrative comparators — specific examples of how much dearer particular projects can be than plain government borrowing.
The lazy version (don't say this)
"PFI is 40% more expensive than public procurement." A blanket, all-projects claim the NAO did not make — and one that lets defenders dismiss the whole critique on a technicality.
What the NAO actually indicated
As illustrative examples: a group of schools could cost around 40% more under PFI than if funded by government borrowing; and some hospitals as much as roughly 70% more. Examples, not a universal rate.
The distinction matters because it's the honest one — and because it's worse for the model, not better. You don't need a blanket 40% surcharge to make the case. You only need the fact that a country's own auditor, looking at real projects, found premiums running from tens of per cent to around 70% and "little evidence" the model delivered value for the money. When the illustrative case is that bad and the systemic verdict is that damning, the argument is settled without exaggeration. verified 40% / 70% are NAO illustrative comparators, not a blanket rate
Why we quote this precisely This whole site lives or dies on being fact-checkable. So: the 40% (schools) and ~70% (hospitals) figures are the NAO's illustrative comparators against government borrowing, not a claim that every PFI project costs that much. Cite them exactly that way and they can't be waved away.
Sources: New Statesman on the NAO PFI/PF2 review; The Conversation — PFI failure & Carillion. The 40% (schools) and ~70% (hospitals) figures are the NAO's illustrative comparators against government borrowing, not a universal cost premium.
Carillion: when the private partner falls, the public catches it
If PFI's price was the quiet scandal, Carillion was the loud one — the moment the model's central promise, "the private side carries the risk," was tested in public and failed.
Carillion was one of the UK's largest construction and outsourcing firms, running a vast portfolio of public contracts — building and servicing hospitals, schools, prisons and roads under PFI and outsourcing deals. On Monday 15 January 2018 it went into liquidation: one of the biggest corporate collapses of its kind in British history. Overnight, essential public projects were left mid-flight and thousands of subcontractors and workers were left exposed. verified
And then the risk did exactly what it always does. Because you cannot let a half-built hospital stay half-built, or let a running prison go unstaffed, government had to step in and keep the essential services going. The "risk transfer" that justified the premium evaporated the instant it was needed. This is the same shape you've already met here: the private operator of an Australian toll road goes bust but the road keeps charging tolls; a PPP hospital operator hits trouble but the hospital cannot be allowed to close. The upside was private. The failure was socialised.
When the toll road goes bust →
Australia's own private operators went into receivership on fantasy traffic forecasts — and the public monopoly kept billing regardless. The exact Carillion pattern, on asphalt.
The PPP hospitals →
When the private partner running an essential hospital stumbles, government cannot let it close. The "risk" was never really theirs to carry.
Source: National Audit Office — Investigation into the government's handling of the collapse of Carillion (June 2018). Liquidation commenced 15 January 2018.
The country that ran it hardest stopped signing new ones
Here is the fact Australian governments never bring up when they announce the next "partnership."
Even the UK — the model's world champion — pulled the plug on new deals Weeks of Carillion fallout and a damning value-for-money verdict later, the UK Chancellor used the October 2018 Budget to announce that the government would abolish new PFI and PF2 deals. The country that pioneered and exported this model, and signed more of it than anyone, decided it would stop signing new ones. verified Oct 2018 Budget announcement to end new PFI/PF2
Note what that is and isn't. It didn't cancel the ~£200 billion already owed — those contracts run for decades yet, which is precisely the point of the model. But as a forward decision it is a verdict: the government with the most experience of PFI, the best data on it, and the most reason to defend it, looked at the numbers and refused to sign more. Australia, with the UK's entire case file sitting there in plain sight, is still doing it — still moving costs off the books, still calling sales "leases," still promising the private side carries a risk that keeps flowing back to us.
Source: the UK Chancellor's October 2018 Budget announcement to end new PFI/PF2 contracts. Existing PFI/PF2 liabilities — around £200bn per the NAO — remained in force and continue to be paid.
The lesson Australia keeps refusing to learn
Read the UK story back-to-back with this site and the resemblance stops looking like coincidence. The off-book accounting, the cost-of-capital gap, the "risk transfer" that reverses at the first real test, the collapse that lands on the public — these aren't British quirks or Australian accidents. They are what the model does. Britain ran the experiment to the end, wrote up the result, and quit ordering more. The only open question is how much of the same bill Australia signs before it does the same.
Cross-references: How the deal works · When the toll road goes bust · The PPP hospitals. Citations on Sources.