A government should judge an investment by what it costs and what it delivers over its life. PFI gave ministers another consideration: whether the borrowing appeared in the measure their fiscal rules were watching. I think that accounting incentive helps explain why an expensive way to finance public buildings became so attractive.
There were other reasons to use it. Departments had a poor record of maintaining assets. A long contract could make one company responsible for building a hospital and keeping it working, rather than leaving maintenance for a later budget. Private finance also allowed projects to proceed when ministers were unwilling to provide the money directly.
Those benefits deserve consideration. They don't settle whether the contracts offered value for money.
What the figures actually compare
The National Audit Office's January 2018 report recorded more than 700 operational PFI and PF2 deals, with a capital value of about £60 billion. Annual charges were £10.3 billion in 2016-17, and future charges were projected at £199 billion, continuing into the 2040s.
Those figures aren't a direct comparison between the price of buildings and an inflated repayment for the same buildings. The charges also cover financing and services over long contracts. Maintenance and other bundled work would have cost money under public provision too. The NAO didn't present the totals as a verdict on every deal.
Private borrowing nevertheless tends to cost more than government borrowing. A more expensive funding arrangement needs a convincing explanation of what the public receives in return: useful expertise, better performance or a meaningful transfer of risk.
Why the accounting mattered
For years, many PFI liabilities were kept outside headline measures of public debt. That made the model attractive under rules which constrained direct borrowing. A project could appear affordable for the purposes of a fiscal target even where public finance would have been cheaper over its life.
No conspiracy is needed to understand why ministers might prefer that option. If the rules reward moving a liability into a different category, people will have a reason to do it. That doesn't establish the motive behind each contract. It does mean the rules themselves deserve scrutiny.
The government ended PFI and PF2 for new projects in 2018. Existing contracts continued, along with their obligations. Changing the policy couldn't remove commitments already made.
The useful lesson is to compare the full alternatives. What would direct public borrowing have cost? Which risks genuinely moved to the contractor? What did the service contract buy? Keeping a liability out of a headline debt figure isn't a public benefit worth paying extra for.