No Ms. Orr, we really have already done this experiment

I have to admit to never really having been a great fan of the PFI idea. I can recall conversations in the early 90s as it first arrived and thinking, no, come on, it's not going to work out that way. I thought that the combination of people who write contracts for a living interacting with civil servants would lead to some, umm, negotiating inequality. Further, allowing politicians to get now something that the next lot woud have to pay for after the election was going to end in tears.

However, that PFI wasn't that great doesn't mean that Ms. Orr isn't spouting nonsense here:

Essentially, PFI is a way of introducing an artificial profit motive into public projects, to entice the private sector to get involved. Where does this profit come from? It is provided by taxpayers. PFI can only "work" for the taxpayer if the profit handed to the private sector is equal to the saving made by involving them in the first place.

No, the idea of PFI, or privatisation, of private provision, of contracting out, of any and every method that isn't direct supply of goods or services by government itself, is that the efficiencies to be made from organisations subject to competitive pressure will be greater than the profits they will earn.

This might sound like an odd idea but we've actually had an experiment which tested this thesis very nicely. Called the 20th century.

Efficiency is, over time, improvements in total factor productivity. That's just the fancy name for either getting more output with the same inputs or getting the same output with fewer inputs. In that gloriously planned economy that was the Soviet Union, according to Bob Solow at least, it appears that there was no increase in tfp at all. Nada, zip, from 1917 to 1991. Whereas 80% of the growth in the market economies over that same short century came from tfp improvements.

I would call that evidence that Ms. Orr's contention is provably wrong.