In this section
More than a million British homes have solar panels on the roof. If yours is one of them, and it exports to the grid on a bright afternoon, you are a private electricity generator. Nobody is coming for your roof.1
Nationalisation, public ownership or common ownership are used as blanket terms, and carry a lot of assumptions.
In October 2008 the state took majority ownership of the Royal Bank of Scotland and held it for more than a decade. The shareholder register changed. How the bank operated did not.
The National Coal Board was a nationalised industry throughout the miners’ strike. The most bitter industrial conflict in post-war Britain was at a nationalised industry, between the government directing it and its own workforce.
The Post Office was wholly owned by the state for every year of the Horizon scandal. It prosecuted innocent sub-postmasters. Public ownership neither prevented that nor brought it to light.
And yet
“In 2025 privatised water companies in England discharged untreated sewage 291,492 times.”
2
Not one major new reservoir has been completed in this country in over thirty years.3 That is an industry figure, not a campaign group. So in three decades a structure designed to attract private capital into water has not attracted private capital to build reservoirs.
And in a dry summer, farmers are told how much they may draw from rivers and boreholes, and have to decide which crops to water and which to let die. Some of that is the weather. The reason the weather bites this hard is that the storage was never built.
The sixteen water monopolies paid out £78bn in dividends in the 32 years from privatisation in 1991 to March 2023. That’s nearly half the £190bn spent on infrastructure over the same period. They accumulated £64 billion of debt, despite being sold with none. All Financial Times figures, inflation adjusted.4
Public ownership is not a magic bullet. Change the ownership and leave everything else the same, and you have bought a letterhead. But public ownership allows you to solve problems that are intractable structural faults with privately owned monopolies.
So what are we trying to achieve?
Chapter 10 drew the line that matters. Markets work where entry is possible and exit is real. If the pizza shop serves you cardboard with burnt pepperoni on it, you walk down the road for a curry or fish and chips. Leave that market to take care of itself.
But there is a difference between market mechanisms that allocate resources, and unregulated free markets. That pizza shop is registered, inspected, rated and can be shut down by an environmental health officer on a Tuesday afternoon. Nobody calls that ideology. Nobody wants salmonella.
Everybody accepts regulation everywhere. But what if regulation is not enough? What if the customer cannot leave because there is no second pipe, no alternative grid, no other railway line. That is where regulation is being asked to do the entire job by itself.
We should ask what we need public ownership to do.
The questions
Here is what any proposal for public ownership should be able to answer. Not all of them, in every case. But if it cannot answer several, it is a slogan.
1. Does it lower the cost of capital on what still has to be built?
Not on the assets that already exist, but on building new capacity we collectively need.
2. Does it let us publish an order book?
Will public ownership unlock a demand signal like a twenty-year commitment that a manufacturer can build a factory against?
3. Does it address a coordination failure that regulation cannot?
Connection queues, whole-system sequencing, building ahead of demand. A regulator can penalise a network for what it did; it struggles to instruct one to build something for which there is not yet a customer.
4. Does the mandate change, or only the shareholder register?
The RBS test. If it still operates to maximise return on capital, ownership has changed but behaviour has not.
5. Is there a route for the people it can harm to be heard?
Good governance does not arrive with the share certificate. It has to be designed in. This is the test the Post Office and the NCB failed.
6. Does it rebuild capability, or just transfer a contract?
Workforce planning is national resilience. Ownership that retains the same outsourced delivery model does not help.
7. Does ownership capture consequences that currently land somewhere else?
Would public ownership save the state money by preventing costs the state pays to fix?
Not every case is the same
Water answers 1, 3, 4 and 5. There is a forward capital programme of a size nobody disputes, it is financed at a private cost of capital on a regulated asset base, running up debt that bill payers foot. The business model rewards maximising existing assets, not building new capacity – a new reservoir won’t get it any more customers. The government does not dispute the diagnosis. It commissioned a review under Jon Cunliffe, a former Deputy Governor of the Bank of England, which reported in 2025 and recommended abolishing Ofwat. Ministers accepted it, and a Bill replacing four regulators with one was announced in the King’s Speech. But reorganising the regulator changes nothing about who owns the assets, what the capital costs, or whether a reservoir gets built. Their answer to a structural problem has been a new body.
Energy answers 1, 2 and 3 most strongly. The connection queue is the clearest case in British infrastructure of a coordination problem that regulation cannot solve, because the instruction required is to build ahead of demand.
Public transport answers 2, 3 and 7. A bus route is a monopoly on that route. And as Chapter 25 sets out, the cost of not running it does not land on the operator. It lands on the people who cannot get to work.
Council housing answers 1, 2, 3 and particularly 7. A private developer is not paid for the homelessness it prevents. It cannot be. So it does not build for it. The state is on the other side of that bill – temporary accommodation, health, children’s services, the school a child changes four times in two years. It can benefit from positive externalities that no developer can. That the rents also anchor the private market, as Chapter 24 sets out, is a bonus.
Burger vans, barbers and pizza shops answer none of them, which is the point. Inspect them. Rate them. Close the bad ones. Then leave them alone.
GB Energy
In October 2024 Juergen Maier, the chair of Great British Energy, gave evidence to the Public Bill Committee on the Bill that created it. Asked when bills would come down, he said the way to lower them is to get more renewable generation onto the grid, and that every gigawatt added helps. But how those savings reach the consumer is a matter of policy, and “that is not the scope of Great British Energy.”5
He was right. The pricing rule set out in the last chapter is not his to change.
That speaks to what its mandate is. GB Energy pushes: it invests in supply chains, co-invests alongside private capital and takes stakes, in the expectation that supply will call forth the rest. The alternative is to pull — to publish twenty years of demand and let the supply chain assemble itself against a guaranteed market. Everyone knows you cannot push on a chain.
The difference matters because a firm handed capital still faces the same question it faced before: will anyone buy this? It now has money it did not have and no more confidence than it had yesterday. And it has watched what happened to the firms that geared up for the zero carbon homes standard, the feed-in tariff, the Green Homes Grant and the carbon capture competition.6
None of that is an argument against public equity or blended finance, which are respectable instruments with real uses. It is an argument about sequence. De-risking supply is second-order while demand is uncertain.
Then there is the size. The headline figure is £8.3bn over this Parliament. Around £2.5bn of it goes to Great British Energy – Nuclear, a separate company, leaving £5.8bn for GB Energy itself. The department rejects the suggestion that the budget was raided, saying a nuclear role was always intended.7 Take either figure. Set it against a grid that needs rebuilding, and neither matches the scale of the task.
On the day GB Energy’s last pound is spent, the following are unchanged. Gas still sets the price of electricity. The hurdle rates in the last chapter still apply to every project it co-invests in. The returns on the generation it does not own still leave the country. The connection queue is still there. And no manufacturer has been shown an order book.
This is not a criticism of the people running it. It is a description of what resources they were given and notes they were not asked to fix the broken structures.
- MCS, the certification body for small-scale renewables, recorded a cumulative total of 1.85 million certified solar PV installations in the UK by the end of 2025, following a record 257,397 installations in that year alone — the highest annual figure on record, exceeding the previous peak set in 2011 during the Feed-in Tariff boom. DESNZ deployment statistics put total UK solar capacity at 22.8 GW at the end of June 2026 across 2,076,000 installations of all sizes, of which the large majority by number are domestic rooftops (DESNZ, Solar photovoltaics deployment, table ET 6.3, published 30 July 2026). The figure of a million in the text is deliberately conservative: MCS certification is not mandatory for rooftop solar, so unregistered systems are not captured, and the MCS count includes some non-domestic installations. Households with a certified installation are eligible for the Smart Export Guarantee, under which licensed suppliers pay for surplus electricity exported to the grid. New build is an increasing share: installations on new build properties accounted for 35 per cent of the 2025 total, and the Future Homes Standard is expected to make solar effectively standard on new homes. ↩︎
- Environment Agency, Event Duration Monitoring — Storm Overflows — Annual Returns, published 26 March 2026 for the 2025 return year. 291,492 spill events were recorded across storm overflows in England in 2025, totalling 1,873,751 monitored hours. Three qualifications are material. The 2025 figure represents a fall of approximately 35 per cent against 2024, with total duration down 48 per cent and average spills per overflow down from 31.8 to 20.5; the Environment Agency attributes much of the reduction to a drier than average year, since storm overflows operate in response to rainfall. Storm overflows are permitted to discharge during rainfall and snowmelt, and a discharge is therefore not in itself unlawful; the Environment Agency takes enforcement action where discharges fall outside permit conditions. And what is discharged is untreated wastewater diluted with rainwater, not undiluted sewage. The figure is given here as a level in a single, comparatively good year rather than as evidence of a trend, because monitor coverage rose from fewer than 1,000 devices in 2016 to every storm overflow in England by the end of 2023, which means year-on-year comparisons across that period are dominated by changes in coverage rather than in behaviour. Comparable annual totals are 301,091 spills in 2022 and 464,056 in 2023, the latter exceeding 3.6 million hours in England’s sixth-wettest year on record. Volume is not measured anywhere in the national dataset: event duration monitors detect the level at which an overflow begins to spill and record start and end times only. The water industry’s own open data hub confirms the data carries no information on discharge volume or on resulting water quality, and Water UK’s assessment, cited by Northern Ireland Water, is that the cost of upgrading monitoring to capture volumes would be better spent addressing the underlying problem than on improving its measurement. ↩︎
- Water UK, the industry body, states that no new reservoir has been built in more than thirty years, and attributes this to Ofwat and previous governments blocking construction. Nine new reservoirs have now been approved, in addition to Havant Thicket, which is already under construction: built by Portsmouth Water under a bulk supply agreement with Southern Water, it broke ground in October 2024, will hold 8.7 billion litres, is due to open in 2029, and is described by the Institution of Civil Engineers as the first major new UK reservoir in over thirty years. The last completed was Carsington in Derbyshire, begun in 1979 and finished by Severn Trent in 1992, three years after privatisation. The industry’s explanation is recorded here in full because it does not weaken the point in the text but sharpens it: a regulatory settlement under which the regulator and successive governments could prevent construction for three decades, and under which no company had sufficient commercial reason to overcome that, is a description of the structure rather than a defence of it. On leakage, the regulator’s figures are more favourable to the industry than the industry’s own: Water UK states that companies have reduced leakage by around 40 per cent since privatisation, while Ofwat’s Water Company Performance Report 2024-25 puts the reduction at 43 per cent, from a benchmark of 5,000 megalitres a day at privatisation to 2,869 megalitres a day in 2024-25, the lowest on record. The further 17 per cent reduction by 2030 is a target set by Ofwat rather than an industry undertaking. Both claims are accepted. ↩︎
- Financial Times analysis based on regulatory data and adjusted for inflation, covering the 32 years from privatisation in 1991 to March 2023. The sixteen regulated water monopolies in England and Wales paid £78bn in dividends over the period, against £190bn of infrastructure spending, and accumulated more than £64bn of net debt despite being sold with none. At privatisation the existing debts of the water authorities were written off and public money was injected alongside the sale. Ofwat’s own “Historic dividends since privatisation” dataset gives a lower figure — approximately £52.7bn in statutory dividends for 1992/93 to 2023/24 — because it is expressed in nominal terms and records dividends declared by the regulated companies rather than flows through the wider holding structures above them. Other figures in circulation reflect different endpoints and methods: £57bn against £123bn of capital expenditure to 2018 in 2018 prices, from research at the University of Greenwich; £65.9bn cited in the House of Lords in February 2024; £72bn cited in the House of Commons in June 2023. The Financial Times figure is used here because it is inflation-adjusted and set against capital expenditure on the same basis. The direction of the argument does not depend on which is preferred. ↩︎
- Official Report, Great British Energy Public Bill Committee, 8 October 2024, column 6, Q5, in response to a question from Stephen Flynn MP. Juergen Maier is chair of Great British Energy; Dan McGrail is chief executive. The answer is quoted in the sense in which it was given: Maier’s position was that additional renewable generation is what lowers prices and that every gigawatt added helps, but that the mechanism by which any saving reaches the consumer is a matter of policy rather than one for the company. That is an accurate account of the company’s remit, and it is the remit rather than the answer that is under discussion here. ↩︎
- See the endnote to Chapter 22 on withdrawn climate policy, which sets out the zero carbon homes commitment and the Code for Sustainable Homes, the Feed-in Tariff, the Green Homes Grant and the £1 billion carbon capture and storage commercialisation competition, together with sources and dates. The examples span successive governments of different parties, which is the point: the pattern is structural rather than partisan. In each case an announcement created an investment signal, firms responded to it, and the signal was then withdrawn — so the cost falls not only on the emissions foregone but on the supply chain that geared up and then stood down. ↩︎
- The Spending Review of June 2025 allocated £8.3bn over this Parliament to Great British Energy and Great British Energy – Nuclear together. Around £2.5bn is for the latter, a separate government-owned company established in July 2023 as Great British Nuclear and renamed in June 2025, whose principal current commitment is to small modular reactors. That leaves approximately £5.8bn for Great British Energy itself, of which £1bn is the Supply Chain Fund. The Department for Energy Security and Net Zero rejects the characterisation that Great British Energy’s budget was raided, stating that a nuclear role was envisaged from the outset. Great British Energy is not an energy supplier and does not sell electricity or gas to households or businesses; its functions are investment, project development, supply chain support and local power. Its Strategic Plan, published in December 2025, commits it to owning and operating at least 15 GW of clean generation and storage by 2030, to mobilising £15 billion of private finance, to a £1 billion supply chain programme, and to supporting over 1,000 community and local energy projects. None of those commitments takes the form of a forward order book shown to a manufacturer, which is the distinction drawn in the text. ↩︎