The government wants to keep the water companies off the public balance sheet, and Sir Jon Cunliffe wants greater public supervision. Together, they want: public control; private finance; with current and especially future customers paying. The Labour backbenchers and most of the public want nationalisation. None supports the original privatisation model, with economic incentives and arm’s-length, light-touch regulation.
Thames Water, the unacceptable face of water privatisation, displays just why the status quo is so awful: extremely high costs (and fees) for new short-term private finance; control by and in the interests of specialist distressed-debt funds; and sharply rising bills. These bills come with a demand to lay off environmental and economic regulatory fines for breaches of the law and the water companies’ licences. And for the “turnaround” failures, Thames Water has demanded greater bonuses and “retention payments” for its “greatest asset”: its top managers and board members. Given the chasm between performance and salaries and bonuses to date, the rewards can’t have much to do with performance so far. At each twist and turn, it looks like it can’t get much worse, and then – at least in the case of Thames Water – it does.
For the rest of the water companies, the report card is mixed, and for the country as a whole, the existing model leaves England and Wales with a potential shortfall in water supplies in the near future, as well as heavily polluted rivers, partly the result of an underinvested sewerage system. Not all this is the fault of the companies. Regulators have pushed hard to keep bills down, and politicians have, as ever, wanted a version of cakeism: lower bills, more investment and more capital maintenance. Regulators have made some pretty catastrophic mistakes, nowhere more obvious than the allowing of the great exercise in financial engineering. And then there are the farmers, who want to be paid not to pollute.
Thames Water is the biggest and arguably the worst of all the water companies in England and Wales, and if this is what the current ownership and regulatory model produces, it is understandable that the public and the politicians they elect are far from content to just carry on as usual. The critique bit of the Interim Report from the Cunliffe Commission and the earlier National Audit Office (NAO) report set out, in clear and unambiguous terms, the scale of the failures and the completely justified reasons why the public has, as Cunliffe puts it, lost trust in the current model.[1] It is possible to have a debate about how big the failures have been and who should bear what proportion of the blame, but the overall failure of the current model is beyond reasonable doubt.
Public control and private finance
The question is: can we have public control and private finance? The answer is a qualified “yes”, not least because that is how the Victorians approached their great utility investments – notably in Manchester and Birmingham – through the public corporation. This approach had its controversies, but governments of various political persuasions fell into this model after experiencing what pure private water supply actually meant in practice.
This is not a version of the current model. Trying to exert public control in the current context is the best way to maximise the cost of capital and hence bills, as the unfolding drama at Thames Water demonstrates. Public control is not a magic wand of discretion as and when politicians and regulators get excited and when the media gets in on the act. It is very hard to marry control and finance. Even KKR, having extracted sole-bidder status from the distressed-debt specialists who have gained control, and presumably, having given the nod to them to carry on with the high-interest, short-term £3 billion loan, could not make the numbers add up. Not even a nod and a wink from the government through direct calls with No. 10 and Peter Mandelson could persuade them otherwise. Deals, fixes and exemptions from fines could not (and should not) be stitched together. No wonder KKR walked.
If public control and private finance are to marry up, it is imperative that the rules and terms under which the public control is exercised are clearly specified within a clear legal framework. That is something the British are very bad at. The British state and its institutions are based on discretion – good chaps (and it has been largely men) have exercised wide discretion to interpret the public interest since the Civil Service reforms of the nineteenth century, on a case-by-case basis, quite contrary to European and US practice. It is why almost all attempts at private finance of government-controlled activities come with a very heavy price. Think the numerous Private Finance Initiative (PFI) schemes; think the disaster of the public–private partnership (PPP) for the London Underground; and think of all the interference in HS2. Discretionary political and regulatory risk is a core feature of UK utility regulation and it is outside management control, and hence is always going to increase the cost of equity. British experience points to just how high this cost can be: a cost of capital premium overriding any actual efficiency gains that the claimed incentives have brought.
Control is hard to separate from ownership, though it can be done. Supermarkets do not usually own farms but they exercise great control over their food supply chains and the terms paid to farmers. Marks & Spencer’s does not produce clothes, but it did garner a great reputation for both the quality and the price of its garments. Yet these are special examples, combining intense retail competition in final markets with market power in the (im)balance between the supermarkets and the farmers and retail outlets for clothes and the clothing manufacturers.
The reason supermarkets can exercise control is because the relationship is contractual, specifying quality and price, and with the very credible risk of loss of contract for supplier failure (unlike the government’s determination not to impose a special administrator in the Thames Water case and hence generally). As we shall see, this carries over to the water companies: the government could directly contract new assets and is proposing to do so in respect of new reservoirs. But it would need to specify these contracts, rule out ex-post discretion and stick to the spirit and letter of these contracts.
Cunliffe’s supervisory model
Cunliffe proposes that public control should be achieved through regulation on a company-by-company supervision model, borrowing a model from financial services regulation. The idea is deceptively simple: OFWAT should set up specialist teams to understand in detail the decisions of each water company and intervene in the day-to-day business of the company, as and when it sees fit. It should get away from using industry-wide econometric data and instead focus specifically on each company. Crucially, it should have even wider discretion. The “good chaps” should take de facto control.
The supervisory model is probably the worst option that could be advanced. It maximises the scope for capture, maximises the regulatory interference and leaves each company focused on how to please the OFWAT team assigned to it. In principle, all the main decisions of the board are at the discretion of OFWAT: the capital budget, the decisions on new investments, the maintenance decisions, the dividend decisions, the capital structure decisions, and eventually the appointments too.
If this is what Cunliffe advises the government to do, then it would be better to just take control. The Cunliffe Commission has been told not to get into nationalisation, and it has pussyfooted around the ownership structures as a result. This is a mistake. If the Commission had been able to consider the conventional nationalisation model, it would have had the opportunity to spell out just how bad this might turn out to be. A nationalised industry, reporting to the relevant secretary of state, and reliant on Treasury finance and political approval for customer bills, and hence funding, would be a practical disaster. Witness the recent Comprehensive Spending Review and the “headline” spending focused on social housing, SMRs, Sizewell C, and the NHS. All of these may (or may not) be wise spending commitments, but what matters is what is not on the list. Prioritising constituencies vulnerable to the Reform Party in a nationalised water industry might lead to prioritising sewers in the North East and North West of England, for example, but not in London, and the South East. This is no way to allocate scarce capital to meet the requirements for investment and capital maintenance in the water industry.
Demolishing the case for renationalisation (had Cunliffe been allowed to do so) would have opened up a lot of space to consider other ownership models and institutional structures. This is left to the final report – no room for interim considerations and consultation on this territory.
The ownership options of water companies are many, and include the plc privatisation model, the private equity model, the infrastructure fund model, the not-for-dividend model, the not-for-profit model, the public corporation, and the nationalised industry.
The public corporation model
The only model in this list that matches the government’s preference for public control and private finance is the public corporation. It is a model that was deployed extensively by municipalities throughout Europe and in the UK up until the Second World War, after which the nationalised industry model took over.
Water and electricity are the two stand-out examples of past UK public corporations, and in particular for the water supply arrangements to upgrade supplies to Manchester from the Lake District. The companies were set up with arm’s-length management, subject to strategic direction by their owners, typically local authorities, and they mostly had specific purposes. They raised private finance – various varieties of debt – which attracted interest and capital repayments. This was, in effect, similar to the state raising debt finance and then dispensing the monies to the nationalised industries, but without the intermediary.
There are many attractions to this model, but there are also downsides. The primary one is the status and capabilities of local government in the UK, which lacks the necessary skills after the great centralisations of the electricity and gas industries following the Second World War (and much else, as local government was emaciated) and consolidation of the water industry into large regional nationalised utilities in the 1970s. Unlike many of the European cities and municipalities, the skillset to manage production activities has long gone. It is a stretch for local government to run even local transport services, and typically these are contracted out.
To move to a public corporation model would require a huge political and institutional step. Contrast this with the Cunliffe Commission’s route to political control through regulatory supervision whilst leaving ownership to the private corporate format created at privatisation in 1989. Whereas the public corporation model in the nineteenth and early twentieth centuries had a high degree of autonomy, the supervisory model maxes out on discretion. Because the supervised companies continue to hold equity (they are privately owned), the finance includes not just the cost of debt, but also the cost of equity. They become very high-risk businesses, because they have lost important elements of risk management. The board is responsible for, but not in control of, many of its core decisions.
The big difference between finance for the privatised companies and for public corporations is that the former have equity risk and the latter should not. It is not that equity risk goes away. It doesn’t. It is just that it sits with customers. The public corporations lend themselves to rate of return and pay-as-you-go regulation, not incentive-based regulation, although in the past many municipalities sought to make a profit out of their utility businesses to cross-subsidise other public goods. These profits were really a form of disguised taxation, levying high bills on customers over and above the costs. They are optional forms of local tax, whereas profits are of the essence of the private model, and indeed excess returns to drive the incentives.
Customer funding
These considerations link back to the ultimate bedrock of any of these models: us, the customers. If the government does not want to make us pay through taxation, then it must make the customers remunerate the operating costs and provide the basis for private finance. The customers must pay for the services, pay the interest on the debt, repay the capital debt, and pay the dividends on equity.
It is here that the government’s third objective comes into play: funding from current and future customers. This is why there will always be a cap on investment. For political reasons, governments try to hold down bills because we, the voters, do not want to pay. It passes governments by in announcing all sorts of large-scale capital investment schemes across energy, water, transport and communications, that all the private debt and private equity have to be paid for. Each sector argues that its specific extra capital expenditure can be managed within the household budgets (though not energy during the recent gas price spike), but they rarely add them all up. It is obvious – but typically unnoticed – that household budgets must be able to absorb all the multiple bills simultaneously. What’s more, the greater the political and regulatory risk and uncertainty, the higher the cost of capital, and hence the more likely that the affordability constraint is breached.
How to better match public control with private finance
The public corporation model is an ownership solution, and one that otherwise leaves the monopoly utility in place. The badges on the corporate offices replace the word “private” with “publicly owned” (even though private listed companies are called “public” in the confusing nomenclature). But that is largely it.
What the public corporation lacks is a separation of those things best driven by competition, and those best internalised in the public corporation. Again, confusing the names, the recently created NESO (National Energy System Operator) in energy is a public corporation, but one that does not produce anything. Yet it does contain the germ of an idea for resolving the ownership and regulation issues in water.
What the government should control is the “plan”, and therefore define what the overarching outputs should be. This strategic direction is rightly emphasised by Cunliffe. In the private water model, the periodic review process leaves the defining of the investment programme and the necessary strands of capital maintenance to the companies in their business plans that form the basis of their submissions to their periodic reviews.
Public control takes these functions away from the private companies and places them in a system operator/system planner/system regulator. NESO carries out many of these functions in electricity, notably now coming up with a spatial plan and running the auctions for some of the new capacity.
In the water case, the system is the catchment, and the analogy to the NESO would be new Catchment Water System Operators (CWSOs), in effect planners who are also auctioneers and regulators. These new bodies receive the overarching objectives and strategic direction from government (or it could be from local authorities in the catchments), provide digital maps of the catchment, simulate options to meet the objectives against the digital maps, and then invite bids to deliver the outputs. This maximises transparency (the digital map is publicly available and can be played with by all and any interested parties) and brings in the many players in the catchments, including but not limited to water companies. Notably, farmers, NGOs (non-governmental organisations) and trusts, flood managers, and other users of the catchments all have roles to play.
This is not the place to repeat the main features of the catchment regulatory model (for more on this, see Helm 2019.[2] But it is the place to set it against the privatised current model and the public corporation and nationalisation models.
The catchment model achieves public control, and it is transparent and inclusive. It brings all the parties to the table. It challenges the incumbent companies with competition. It is neutral as between natural capital and hard concrete solutions. These are big gains. Instead of automatically defaulting to investing in multiple large tanks to store storm water run-off, for example, it could invite alternative bids from natural capital approaches, and bring farmers and housebuilders and highways into play, encouraging them to create porous surfaces and water storage to stem the flows.
This is public control and private delivery, with competition. The bidders (as the water companies could become) are mostly private through trusts, and public bodies can bid too. Strictly, it is ownership-neutral. New assets are financed from private investors, and operations of the assets are subject to bidding for the contracts too, whether private or public.
The catchment model is likely to be very significantly cheaper than the current model or the public corporation because it allows for natural as well as concrete options, and brings multiple parties to the table with influence over catchment flows and catchment pollution. Competition is applied where it should have the greatest effect on costs and maximises efficiency incentives.
Most importantly, the catchment players do not need company-specific supervision, as proposed by Cunliffe.
A better and more enduring way forward
Instead of adding yet more regulation on top of what is already in place, which is what the Cunliffe “supervision model” would do, a better approach would be to sort out the fundamentals that lie behind the regulatory and customer failures. “Supervision” means much more intrusion by OFWAT in the running of the businesses, and it would require a big expansion of OFWAT’s staff, skills and hence costs, and a duplication and overlap with the proper functioning of the company boards. With all the challenges to boardroom remuneration and threats of jail and media personal attacks, it is not at all clear why anyone would want to be a director of a water company in the Cunliffe model (in stark contrast to the banking sector Cunliffe draws upon).
One way forward is to look to the ownership of the industry. The monopoly route is to turn the companies into public corporations, going back to the nineteenth-century approach. This has attractions, but it would also require a really major overhaul of local government, not only in its functions and skillsets, but also its legal powers. There would need to be a major new local government Act. This is not likely to happen.
The alternative is to set up the catchment model, and maximise the opportunities for all parties within the catchment to participate in shaping the options and to bid to do the works. This maximises competition. Doing the works (what the bidders would be in business to do) is naturally a private sector activity. There would be some residual monopoly (especially initially). But, even here, the incumbents would feel the pressure to justify end-of-pipe concrete solutions and they might also feel the pressure of competition from each other. There would need to be a transitioning continuation of regulation of these monopoly residuals, and instead of the current merry-go-round of the periodic reviews, there would be negotiated contracts, with an agreement about rates of return.
Could a public corporation model and a catchment model work together? In theory, the answer is “yes”, as the public corporation would be able to bid and compete. There is no need to prescribe the ownership model in the catchment model. But, in practice, the government would still need to resolve its desire for control with private finance. These public corporations would effectively be debt-financed, close to gilt rates, and in essence an off-balance-sheet exercise for HMT.
Recall that, in the catchment model, there are the CWSOs. As with the NESO, these should be public and not private. In the NESO case, it is actually a public corporation, but this ownership format was chosen not because it is optimal for what is, in effect, a public agency, but rather to get out of the Civil Service salary restrictions. It resulted in the ridiculous consequence that, as a result of its corporate status, NESO is regulated by OFGEM. The water sector does not need OFWAT + the Environment Agency + the Drinking Water Inspectorate + CWSOs (catchment regulators). In the catchment model, OFWAT should be abolished, with the contracts negotiated with the catchment regulator.
Back to Thames Water and how the catchment approach resolves the Thames problem
All discussions of ownership and regulation in the water industry come back to Thames Water. It is by far the largest water company, and the scale of its failures dominate discussions (though others have failures too). Thames drives the discussion about nationalisation, about special administration, and about whose interests this once great company is pursuing. The current mess looks like the worst of all possible worlds: it has management side-tracked away from the day job, trying to manage a business that is too big and where the short-term interests of the A-class bondholders are driving the short-term agenda of the company. But don’t underestimate the capacity of the main actors to make it even worse.
It would be better to face up to the most likely outcome: it will end up in special administration. Every day that passes is a day in which the fundamentals are not addressed, and the performance deteriorates relative to whatever is projected in the latest “turnaround plan”. What needs to happen in the case of Thames Water is that it should be placed as soon as possible in special administration, bids should be invited (and there will be many) for the parts – a split between London and the rest, and between water and sewerage would be highly beneficial– and the bids should be judged against the public interest. The bondholders can then be left to scrap amongst themselves over the monies raised from the sale. This is the exact opposite of the current situation. None of the A-class bondholders has much experience of running a long-term turnaround programme. They include world-class specialists in exploiting short-term distressed-debt opportunities. Good though they are in extracting economic value for their investors, theirs is not the public interest.
[1] Independent Water Commission (2025), ‘Independent Water Commission: Interim Report’, 3rd June. National Audit Office (2025), ‘Regulating for investment and outcomes in the water sector’, 25th April.
[2] See Helm D.R. (2015), ‘Water catchment management, abstraction and flooding: the case for a catchment system operator and coordinated competition’, May 7; (2017), ‘Regulatory Reform and The System Operator Model’; and (2019), ‘The Systems Regulation Model’, 12 February.
