Section 3: Why nothing changes · Chapter 9
Who really carries the risk
Reading time: 5 minutes 30 seconds
In this section
For decades, the default government strategy for building infrastructure has been wrapped in a comfortable euphemism called “de-risking.”
The theory sounds great: the state cannot afford to build things itself, so it must partner with private equity firms and multinational contractors. To entice them, the government offers lucrative, guaranteed contracts. But these corporations are too big to fail and immune to genuine competition.
This costs us a fortune in the long term. Firms like G4S, Capita, and the private utility monopolies own our prisons, hospitals, water and energy systems. Small and medium enterprises (SMEs) cannot compete with these corporate giants when it comes to writing multi-million-pound government bids. So monopolies capture the market, extract dividends for offshore investors, and hollow out the service.
The Private Finance Initiative (PFI) ran this way for thirty years. Private consortia borrowed the money to build our schools and hospitals. The state paid it back over decades, at a premium. One group of schools cost around 40% more than if the government had simply borrowed. A hospital, 70%.
The Treasury knew. Its own value-for-money test measured private finance against a rate higher than the government actually pays to borrow. Rig the comparison and private finance wins.
Parklands High School in Liverpool cost £24 million to build. For years it stood empty, and the council were still charged £4 million a year for the PFI contract, almost £100 million in total.1
When the system works, the private investors keep 100% of the cash. When the system collapses – when the water companies pump raw sewage into our rivers or a contractor goes bust – the state is legally and morally forced to step in and bail them out. Carillion shows what “de-risking” means in practice.
State outsourcing isn’t a market. It looks like one, but it isn’t. In a proper market, thousands of buyers and sellers keep everyone honest. Try a bad restaurant, and it closes, and a better one takes its place. State contracts don’t work like that. There’s one buyer—the government—and a handful of giant firms bidding for the work. A transaction, yes, a market, no.
Carillion collapsed in January 2018, holding 420 public sector contracts. The government didn’t rescue it. It refused a £223 million lifeline and let the company go into liquidation.
That did not spare the public the cost. The National Audit Office put the taxpayer’s bill at £148 million. The pension schemes, £2.6 billion short, landed on the Pension Protection Fund. The Cabinet Office kept paying Carillion’s staff to keep the services running.2
Here’s the part that shows who was actually captive. Carillion issued its first profit warning in July 2017, and the Cabinet Office began planning for its collapse almost immediately. Over the following six months the government handed it £1.9 billion of new work anyway, including £1.3 billion of HS2 contracts. Not because anyone was fooled. Because there was nobody else to give it to. It doesn’t stop at collapse, either. These firms often underbid to win the contract in the first place. Then they claw the money back through what’s politely called a “change control process” – premium fees for every tweak and variation, once the contract is signed and the competition’s gone. Underbid, then bill for the rest. That’s the business model.3
“A more honest term than de-risking would be privatising the profit and nationalising the loss.”
Real de-risking looks different. It means using the sovereign power of the state to provide absolute stability. Not stability for offshore financial speculators, but stability of demand, stability of supply chains, and stability of careers for our domestic workforce. Whenever I have spoken to CEOs, whether global corporates about green hydrogen or SMEs about making new kinds of modular clean rooms, they all tell me the one thing they want is to know that there will be demand in five, ten, twenty years’ time. They’re willing to risk fair market competition. What they can’t risk is a government policy killing demand for wind farms or trains.
When government policy is insulated from the panic of hedge funds selling UK gilts, the state can act as a permanent anchor for the real economy. We remove the risk by eliminating the middleman.
- National Audit Office, PFI and PF2, HC 718, 18 January 2018, unless stated otherwise. The NAO’s analysis of Department for Education data for one group of PF2 schools found cumulative cash costs around forty per cent higher than a project financed by government borrowing. The Treasury Committee’s equivalent analysis in 2011 estimated the cost of a privately financed hospital at seventy per cent higher than the public sector comparator.
On the financing premium itself: the 2010 National Infrastructure Plan gave an indicative PFI cost of capital of 2% to 3.75% above government gilts, and Infrastructure and Projects Authority data on deals since 2013 shows debt and equity investors forecast to receive between 2% and 4% above government borrowing. On the six PF2 deals agreed to date the projected post-tax return to investors was 4.5% to 5%, approximately double the cost of government borrowing at the time.
On the appraisal method: the value-for-money assessment discounts future costs at the Social Time Preference Rate, 3.5% in real terms or 6.09% with inflation, which has exceeded the government’s actual cost of borrowing for most of the period since 1992. The higher the discount rate, the lower the apparent cost of paying later — which is what private finance does. HM Treasury does not consider the government’s cost of borrowing relevant when making financing decisions on PFI and PF2 deals; Germany and the United States do make that comparison. When the NAO remodelled the assessment in 2013 using the government’s actual borrowing costs, the outcome changed in the majority of cases to show public finance as the better value option.
On Parklands High School: Liverpool City Council pays around £4 million a year for the school, stood empty for years, and will pay an estimated £47 million between 2017-18 and the contract end in 2027-28 if no changes are made. The school cost an estimated £24 million to build.
Two further findings bear on the risk transfer said to justify the premium. The NAO reports it is not aware of any operational PFI deal in which debt holders have suffered a loss once the asset was constructed and operating. And on equity returns, its analysis of a 2016-17 sale of half the M25 contract equity estimated an annual return of around thirty-one per cent over the preceding eight years.
In fairness to the model, the NAO does not reach an overall verdict on value for money, and records that departments generally regarded construction cost certainty and maintenance standards as better under PFI. Its more fundamental criticism is that the benefits which would have to offset the higher financing costs have never been quantified: HM Treasury has collected no outturn data on them, and the NAO has been unable to identify a robust evaluation of the actual performance of private finance at project or programme level.
A commonly cited figure holds that the NHS received £13 billion of investment for an £80 billion bill (IPPR, The Make-do-and-Mend Service, 2019). The capital figure is confirmed by the Treasury’s own database, which records £13.0 billion of PFI capital investment across 127 health projects. The comparison should be read with care: around half of annual PFI unitary charges relate to debt repayment and financing, with the balance covering maintenance and services the public sector would have purchased in any case, and the repayment figure is undiscounted cash spread to the 2040s, much of it index-linked. The financing premium set out above is the sounder measure of what private finance cost. ↩︎ - National Audit Office, Investigation into the government’s handling of the collapse of Carillion, June 2018. Carillion entered liquidation on 15 January 2018 holding 420 public sector contracts and employing over 18,000 people in the UK. The NAO estimated the cost to taxpayers at £148 million, subject to significant uncertainty. Carillion had requested £223 million in support in January 2018; the Cabinet Office declined, citing concerns about the company’s business plans, legal implications, open-ended funding commitments, the precedent it would set, and the likelihood of further requests. The Cabinet Office continued funding Carillion staff to maintain public services while joint venture partners took over construction contracts. The company’s thirteen defined benefit pension schemes carried liabilities of £2.6 billion, which passed to the Pension Protection Fund. ↩︎
- Same source. Carillion issued its first profit warning on 10 July 2017. The Cabinet Office began contingency planning shortly afterwards, accelerated it in October 2017, and completed it across central government by 15 January 2018. In the intervening months the government awarded Carillion £1.9 billion of new work, including two HS2 joint venture contracts worth £1.3 billion. No contracting authority believed it had grounds to disqualify Carillion under procurement rules. ↩︎