Why a natural monopoly is regulated rather than broken up: the falling LRAC diagram, the loss at P = MC, average cost pricing, price caps and who should own water and rail.
Water in England and Wales was privatised in 1989. In December 2024 the regulator, Ofwat, set the price limits for 2025 to 2030: the average combined water and sewerage bill rises from £440 in 2024-25 to £597 by 2029-30, before inflation. In May 2025 Ofwat fined Thames Water a record £122.7 million for failures at its sewage works and for paying dividends that did not reflect its performance. In July 2025 the government announced that Ofwat itself would be abolished and replaced by a single new water regulator.
If water companies are private firms, why does a regulator decide their prices? And should they be private at all?
Sources: CCW (December 2024); Water Magazine (May 2025); Civil Service World (July 2025).
Privatisation is the transfer of a firm or industry from the public sector to the private sector, usually by selling shares: telecoms, gas, water, electricity and rail were all privatised between 1984 and the late 1990s. Nationalisation is the reverse. Deregulation is different again: removing legal barriers to entry, such as licences, so that new firms can compete, as when coach services and later bus services outside London were opened up in the 1980s.
Many of these industries are natural monopolies: the fixed costs of the network (pipes, cables, track) are so high that long-run average cost is still falling when the firm supplies the whole market. One network is cheaper than two, so competition would waste resources. Privatising a natural monopoly creates a private monopoly, which is why each one has a regulator: Ofwat for water, Ofgem for energy, the ORR for rail and Ofcom for telecoms.
A regulator orders a water company to charge a price equal to its marginal cost.
The company has huge fixed costs in pipes and treatment works, and a low cost of supplying each extra litre.
What happens to the company?
With high fixed costs and a low marginal cost, average cost is above MC at every output. A price equal to MC does not cover the fixed costs, so a private firm would make a loss and leave unless it was subsidised. That is why regulators usually aim for a price nearer average cost.
The water company is the only supplier in its area, so it faces the whole market demand curve. MR starts with AR and falls twice as steeply.
Now the costs. Laying a network is a huge fixed cost; supplying one more customer costs very little. So MC is low and roughly constant, and average cost, the fixed cost spread over more and more units, falls at every output. Draw LRAC falling all the way across, always above MC.
Where demand meets LRAC, average cost is still falling. That is the test for a natural monopoly: one firm can supply the whole market at a lower average cost than two or more could. Draw this falling LRAC, not the standard U-shaped monopoly diagram: examiners mark down the wrong one.
Check: Why would two competing water networks in one town be inefficient?
With two networks, the fixed cost is paid twice and each firm sits further up its falling LRAC. One firm supplying everyone has the lowest average cost.
Left alone, a private monopolist maximises profit where MC = MR, at Qm, and charges Pm. Price is far above average cost: abnormal profit, protected by the highest barrier to entry there is, the network itself. Price is also far above MC, so output is well below the allocatively efficient level.
Could the regulator force the efficient outcome? Allocative efficiency needs P = MC, where demand meets MC, at Qe. But there AC is above price: the firm loses the whole of its fixed cost. No private firm would stay unless the government paid a subsidy, which is one argument for public ownership.
Check: If the regulator sets price equal to MC, the firm:
AC = MC + fixed cost per unit, so AC is above MC everywhere. At P = MC, the price covers the cost of each extra unit but none of the fixed cost.
The usual compromise is average cost pricing: a price where AR = AC, at Pr. The firm covers all its costs and earns normal profit. Compared with Pm, price is lower and output higher, so most of the welfare loss disappears. Price is still above MC, so it is not fully allocatively efficient.
Teacher tip: Say both halves: no abnormal profit, but P is still above MC.
In practice a regulator does not know the firm’s true costs: the firm knows far more than the regulator (asymmetric information). So regulators set a price cap, classically RPI − X: prices may rise by inflation minus X%. If the firm cuts its costs faster than X, it keeps the extra profit until the next review, so the cap gives it a reason to cut X-inefficiency.
Check: Inflation (RPI) is 3% and X is set at 2%. Under an RPI − X cap, by how much can the firm raise its prices this year?
RPI − X is 3% − 2% = 1%, so prices can rise by at most 1%: a real-terms cut of 2%. Most regulators now use CPI rather than RPI, but the sum works the same way.
The UK now runs water in three ways. England’s water companies are private, with shareholders. Since 2001 Welsh Water has been owned by Glas Cymru, a company with no shareholders, which keeps any surplus for customers. NI Water is owned by the government. On rail, six train operators have returned to public ownership since May 2025, starting with South Western Railway.
Check: Which is the strongest argument for privatising a state-owned firm?
Shareholders want profit, so managers are pushed to remove X-inefficiency. The others are false: a privatised natural monopoly is still a monopoly and still needs a regulator.
Now weigh the case for and against privatisation.
Check: On this analysis, how well privatisation works in an industry like water depends most on:
A privatised natural monopoly is still a monopoly, so the outcome turns on whether a regulator can control its prices, investment and quality.
They put it where demand meets MC. With high fixed costs, AC is above MC there, so the firm would make a loss, not normal profit. Normal profit is where AR = AC.
Explain why a natural monopoly is usually regulated rather than broken up, and why a regulator might set price equal to average cost rather than marginal cost.
A natural monopoly exists when the fixed costs of a network are so high that long-run average cost is still falling where it meets market demand. One firm can then supply the whole market at a lower average cost than two or more firms could, so breaking it up would mean duplicating pipes or track and raising average costs.
Left unregulated, the firm would maximise profit where MC = MR, restricting output and charging a price well above both average and marginal cost. Consumers would pay more, and output would be below the allocatively efficient level.
The regulator could aim for allocative efficiency by setting price equal to MC. But because fixed costs are high, average cost is above marginal cost at every output, so at P = MC the firm would make a loss and would need a subsidy to keep supplying. Average cost pricing, where price equals AC, is the compromise: the firm covers its costs and earns normal profit, price is lower and output higher than under monopoly, and no subsidy is needed.
**Evaluation:** the regulator rarely knows the firm’s true costs, so in practice it sets a price cap such as RPI − X rather than a price equal to AC. How well that works depends on the information the regulator has and on whether quality and investment are protected, as the Thames Water fine shows.
Sources: CCW, how much will my water bill increase by 2030 (Ofwat final determinations, 19 December 2024); Water Magazine, Ofwat fines Thames Water (28 May 2025); Civil Service World, Ofwat to be scrapped (22 July 2025); Dŵr Cymru Welsh Water, Glas Cymru; NI Water, who we are; GOV.UK, South Western Railway services transferred to public ownership (May 2025). Diagrams drawn with Rough.js and perfect-freehand, both MIT licensed. Your answers are saved only on this device.