Start with the case for selling the water companies, because it was a real case, made by serious people, and it deserves better than the sneer it usually gets now.
It went like this. The pipes were Victorian, the state was broke, and public ownership had demonstrably failed to invest - the Treasury will always raid the maintenance budget before it touches a headline programme, because voters see hospitals and do not see aquifers. Privatise, and investment could be financed from private capital markets instead of taxes. Shareholders demanding efficiency would discipline costs in a way no minister ever had. Pricing could be set by an independent regulator on engineering evidence rather than by politicians buying elections with artificially cheap water. The taxpayer would shed the risk. None of this was mad; much of it, in 1989, was plausibly the responsible position.
The condition that was never met
A market disciplines its participants through one mechanism only: the customer who can leave. Every efficiency argument for private ownership routes, eventually, through that exit. And water has no exit. You cannot switch supplier, you cannot decline the product, and the state cannot let the supplier fail, because a city without water is not a bargaining position, it is an emergency.
Remove the exit and everything downstream changes its nature. The "customer" becomes a captive revenue stream. The "market discipline" becomes whatever the regulator can enforce - which makes the regulator, not the market, the real counterparty, and regulators are outnumbered, out-resourced and courted by the firms they police. Most importantly, the risk transfer runs backwards: profits can be extracted in good years, but catastrophic failure must always land on the public, because the public cannot walk away from water. Private profit, socialised risk - not as a slogan, but as the literal contractual geometry of the arrangement.
Under that geometry, the rational strategy for an owner is not to build a century of quiet engineering. It is to load the company with debt, pay the proceeds out, sweat the assets, and rely on the fact that the state must ultimately stand behind the taps. Whether any particular company behaved that way is a question for its accounts - they are public, and worth reading - but the point here is structural: the arrangement rewards exactly that behaviour, and arrangements that reward a behaviour tend, over decades, to get it.
What the other side of our own argument says
Honesty requires the concession: public ownership has failure modes of its own, and they are the ones the privatisers correctly named. Treasury raids. Political pricing. Producer capture by workforces rather than shareholders. Anyone promising that nationalisation is a costless unlock has not read the history of nationalised industries. The choice is not between a broken model and a perfect one; it is between two imperfect models, judged by which one's failures are survivable and correctable.
And that is where the decision falls out. A publicly owned utility that under-invests can be forced to invest by the voters who own it - the mechanism is slow and clumsy, but it exists and it points the right way. A private monopoly that extracts cannot be disciplined by customers who cannot leave; the only lever is a regulator the arrangement itself tends to wear down. One model's failure has a democratic remedy. The other's failure has a bailout.
Our decision
So: natural monopolies - water first among them, for nothing is less optional - belong in public or non-profit hands. Not because markets fail, but because this was never a market; a market requires a choice, and there is none. Where choice exists, we would defend private provision against most comers. Where it cannot exist, calling the arrangement a market is a category error with a shareholder register.
This is an opinion, held with the usual confidence and the usual openness to being argued out of it. The bills, the debt, and the investment record are all on the public file for anyone who wants to check whether the structure has produced the behaviour it rewards.
An opinion of the house. The argument is ours; the record beneath it belongs to no one.
How this piece was made
House opinion piece. The privatisation case is stated at full strength first (Treasury under-investment was real; private capital and depoliticised pricing were serious aims), then answered structurally (no exit means no market; risk transfer runs backwards). Public ownership's own failure modes conceded explicitly. Decision made: natural monopolies in public or non-profit hands. Deliberately argues from structure with no figures; company accounts and Ofwat determinations named as the checkable record for a follow-up piece with receipts.