Off the Books

The case for the private finance initiative was serious, and it was made by serious people across two parties.

Public bodies are poor clients. A department that owns its own building has every incentive to defer maintenance, because the roof it does not fix this year is next year's problem and this year's saving. Private capital, tied into a contract that made one company responsible for building an asset and then running it for twenty-five or thirty years, was meant to break that habit. A firm that has to keep the boiler working for three decades has a reason to install a good boiler. Risk moved off the public balance sheet: if the project ran late or over budget, the contractor wore it, not the taxpayer. And the money was there when the Treasury's was not. A hospital could be built now and paid for across its working life, the way a household buys a house rather than saving until it can pay cash. For a country with Victorian infrastructure and a constrained budget, that was not a swindle. It was, for a while, plausibly the responsible thing to do.

Much of that held up. Buildings got built, often on time. The discipline of a long contract did concentrate minds. None of what follows requires pretending otherwise.

The number the model was really moving

The arrangement had a cost, and the auditors counted it. The National Audit Office, reporting in January 2018, found over 700 operational PFI and PF2 deals with a capital value of around £60 billion. The bill for them in 2016-17 was £10.3 billion. And even if no further deal were ever signed, the charges already committed - running on until the 2040s - came to £199 billion. Sixty billion pounds of buildings; two hundred billion pounds of payments.

A gap that size has an explanation, and part of it is ordinary. Private borrowing costs more than government borrowing, because the state can always raise money more cheaply than a company can, and across thirty years that spread compounds into real money. The service element - the maintenance and cleaning bundled into the deal - accounts for more. But the spread and the services do not, between them, explain why governments of both colours reached for the model again and again when direct borrowing would have been cheaper.

The remainder of the explanation is accounting. Money spent through the private finance initiative did not, for years, register as government borrowing. The liability sat off the recorded books. So a fiscal rule that would have stopped a minister borrowing to build a hospital directly waved the same hospital through when a company borrowed for it instead, at a higher lifetime cost. The buildings were real. The saving was presentational. The debt did not disappear; it changed which page it was written on, and the page it moved to was not the one the rules were watching.

No villain, an incentive

The easy reading is a story of ministers duping the public. It gets the mechanism wrong. A rule that measures one kind of liability and not another creates a standing reward for turning the first kind into the second, and rewards of that sort get taken without anyone conspiring to take them. Both main parties used the initiative heavily across two decades. That is not a coincidence in search of a culprit; it is what an incentive looks like when it works. The arrangement did what its design encouraged, consistently - which is the surest sign that the design, and not the people, was the thing to examine.

The confirmation came in 2018, when the government abolished the initiative for new projects, calling the model inflexible and a risk to the public finances. The existing contracts, and their bills into the 2040s, stay. An instrument is not dropped by its own authors while it is still saving money. It is dropped when the bill it deferred has begun to arrive.

What the record asks

The judgement that falls out is narrow, and on the direction it is held with some confidence. A public investment should be judged by its cost across its whole life and by an honest account of what the country owes, not by which column the liability can be made to sit in. Where private finance genuinely delivers something the state cannot - real risk transfer, real expertise, a real discipline that lasts the length of the contract - it earns its place, and the confidence on those deal-by-deal questions is lower and should stay lower. What does not earn its place is a higher lifetime cost bought only to keep a number off one particular page.

Noticing that an incentive shaped a policy does not prove every use of that policy was a mistake, and the accounts are open for anyone minded to test it contract by contract. But the aggregate the auditors recorded is not in dispute, and it points one way. Sixty billion in buildings; two hundred billion in payments; and a rule that could see the first figure and not the second.

An opinion of the house. The argument is ours; the record beneath it belongs to no one.