A typical PFI project involves the private sector paying up front for the creation of an asset such as a school, a hospital or a road. The state then pays a private company for the use of the asset and also, especially in the case of hospitals and roads, pays the company for the provision of services over a long period of time.
In total, there are £60bn worth of assets used for government services financed by the PFI scheme with expected total future payments by the government running to about £200bn by the 2040s.
In the best-case scenario, the government is able to pay for the services as it uses them over the life of the asset and the private sector takes the risk of things going wrong. Private sector risk management techniques can be used to reduce cost over-runs.
It is true, of course, that the private sector has to borrow at higher interest rates than the government to finance the up-front spending. But, then the taxpayer does not have to bear the risks. We only have to think of projects such as the Humber Bridge, HS2 and the nuclear power programme to realise how the government can easily get its fingers burned when it tries to manage long-term project risk itself.
So, what can go wrong? Apart from governments losing control of spending by pushing costs on to the next generation, problems have come from the complexity of PFI contracts. The costs of broking a PFI are huge: it’s good work for lawyers. And then, when costs over-run, the private sector has sometimes dumped the losses back onto government. If a PFI provider goes bust, there may be no effective way of enforcing the contract.
The collapse of Carillion, the ending of PFI contracts on the London Underground and a number of other problems with PFI projects led to popular calls for PFI to be abandoned. But does this involve abandoning the market economy and the use of private capital for public services — policies which some people argued naively had become the accepted ideology of both main political parties in the 1990s?
This would be to misunderstand PFI. When the government uses PFI for funding the provision of services such as education and healthcare, government control is maintained. The government, like any private sector organisation, has to decide to what extent it contracts-out services or lays off financial risk. Airlines have to decide whether to lease or buy aircraft. And banks have to decide whether to contract out the catering and the cleaning or use in-house services. If Easyjet leases 20 planes financed by JP Morgan, people do not think that they are travelling with JP Morgan Air. Similarly, if the government uses PFI for a school, it is still a government school.
As the Nobel Prize winner Ronald Coase taught us, how companies should organise things like this depends crucially on transactions costs. The transactions costs of PFI are huge and governments are not well placed to handle them. One Birmingham City Council roads maintenance contract that was 5,000 pages long with 200 pages of definitions still went to court for interpretation.
Governments should privatise functions directly. There should be more free schools instead of PFIs. We should have more genuinely private roads charging tolls. And the government should send NHS patients to private hospitals and clinics, or promote alternative forms of funding healthcare altogether.
The failure of PFI is an indictment of government provision, not of markets.
This article first appeared in CapX.