Britain’s Most Expensive Mistakes: The Projects Too Costly To Quit
Ranked: Britain’s Biggest Sunk Cost Fallacies And The Bill That Keeps Growing
1. HS2: The Railway That Became A National Psychology Test
HS2 is now the purest British sunk cost fallacy because the argument for continuing has shifted so visibly. It was once sold as a transformative national railway, a high-speed spine linking London, Birmingham, Manchester and Leeds. Today, the remaining project is narrower, later, more expensive and politically radioactive, yet it is still defended partly because stopping could be almost as costly as finishing.
That is the brutal logic of the sunk cost trap. The government’s March 2026 update said £43.6 billion had already been spent on the programme by the end of February 2026. Fresh government admissions now place the possible cost at above £100 billion, with first services between Old Oak Common and Birmingham potentially not arriving until 2036–2039, and full central London completion possibly stretching into the 2040s.
The danger is not simply that HS2 became expensive. Major infrastructure is expensive by nature. The deeper issue is that the country kept making decisions inside a collapsing frame: altered routes, cancelled legs, disputed estimates, shifting station designs, delayed delivery and repeated resets. The original political promise and the final operational reality have drifted so far apart that Britain is no longer judging the same project it originally approved.
That is why HS2 ranks first. It has become a live demonstration of how public investment can mutate into institutional self-defence. Britain may still receive useful rail capacity from the surviving line, but the central question is now unavoidable: if this project were proposed today, in its current form, at its current likely cost, with its current timetable, would anyone dare approve it?
2. Brexit: The Constitutional Bet Too Big To Reprice
Brexit is not a normal project, which is exactly why it belongs near the top. It was not a tunnel, a hospital computer system or a procurement programme. It was a national strategic turn. That made the sunk cost effect stronger, because reversing or softening the decision was never just administrative. It became emotional, tribal and identity-based.
The economic debate remains politically explosive, but the fiscal warning is no longer abstract. The Office for Budget Responsibility has continued to use an assumption that new trading barriers associated with leaving the EU will reduce UK trade intensity by 15% and leave GDP around 4% lower than otherwise in the long run. A parliamentary answer in 2025 repeated that OBR view and said about two-fifths of the effect had already occurred by the time the UK-EU Trade and Cooperation Agreement came into force.
That does not mean every problem in Britain is “because of Brexit". It does mean the country made a huge economic and constitutional move, then spent years arguing over whether the visible costs were real, temporary, exaggerated or worth paying. That is classic sunk cost behaviour: when the original promise becomes harder to defend, the debate moves from measurable outcomes to loyalty tests.
The deeper cost is strategic. Britain is now repeatedly debating partial realignment, trade easements, defence cooperation and sector-by-sector fixes while still refusing the political language of reversal. That is why Brexit sits alongside Britain’s wider struggle to redefine its place in the world: it is not only an economic event but also a long-running argument over national pride, sovereignty and reality.
3. Hinkley Point C: The Energy Bet That Locked In The Future
Hinkley Point C is a different kind of sunk cost fallacy because the case for nuclear power is not absurd. Britain needs reliable low-carbon energy. It needs baseload capacity. It needs energy security. The trap is that those valid strategic goals can become a shield for almost any cost escalation once a project is politically too important to fail.
The National Audit Office warned in 2017 that the Hinkley Point C deal had locked consumers into a risky and expensive project with uncertain strategic and economic benefits. The contract included a strike price of £92.50 per megawatt hour in 2012 prices for the first 35 years of production, with a possible reduction if Sizewell C proceeded.
Since then, the project has faced repeated delays and cost pressure. Publicly available project summaries and recent updates show the expected first power date moving toward 2030, with construction cost estimates far above the original approval-era figure. The argument for continuing is powerful: Britain still needs the electricity, the project is well advanced, and abandoning nuclear capacity would create its own strategic hole.
But that is exactly why Hinkley matters. It shows how a country can make a decision under one set of assumptions, then remain bound to it after the economic surroundings change. The question is not a simple "yes or no" regarding nuclear. The sharper question is whether Britain has learnt how to build nuclear power without creating another machine that becomes too large, too slow and too politically protected to challenge.
4. Pandemic PPE: The Emergency Spend That Became A Waste Mountain
Pandemic PPE procurement is a more morally complicated case because the original pressure was real. In early 2020, governments were scrambling for protective equipment in a global emergency. Hospitals, care homes and frontline workers needed supplies fast. In that context, speed mattered.
But emergency conditions do not erase the duty to control risk. The National Audit Office investigated PPE procurement and contract management after the initial crisis, while the Public Accounts Committee later highlighted huge volumes of unusable or excess equipment. In 2022, the Committee announced that it would burn £4 billion of unusable PPE bought in the first year of the pandemic to generate power.
The sunk cost here was not simply buying too much in a panic. The system involved the following actions: storing, defending, writing down, repurposing, disposing, and litigating the consequences. Once billions had been committed, the public sector inherited a second problem: how to explain the waste without admitting that parts of the procurement machine had failed under pressure.
This approach ranks high because the emotional defence is so strong. Ministers can say they were trying to save lives. That is true in broad terms. But a crisis can justify rapid buying without justifying weak controls forever. The real lesson is that emergency government must be fast and disciplined, not fast instead of disciplined.
5. Test And Trace: The System That Promised Control But Could Not Buy Trust
Test and Trace was built around a simple national promise: identify infection, break chains of transmission and help avoid further lockdowns. That promise was huge. So was the budget. The Public Accounts Committee said NHS Test and Trace had been given an “eye-watering" £37 billion budget over two years and had not achieved its main objective of helping break chains of transmission and enabling a return toward normal life.
The important distinction is that the full allocated amount was not the same as the final amount spent. The National Audit Office later noted that £13.5 billion had been spent in 2020–21, and subsequent analysis has put total spending below the most viral £37 billion claim.
Even with that correction, the sunk cost pattern is obvious. The more expensive the system became, the more politically necessary it was to present it as central to national control. Yet public trust, compliance, speed, local integration and practical effectiveness mattered as much as scale. A giant system that people do not fully trust can become an expensive symbol of action rather than a decisive instrument of control.
This sits beneath the wider question of whether Britain’s public systems reward efficiency or simply defend monopoly-scale bureaucracy. In a crisis, size can help. But size without feedback can become a very expensive way of discovering what local systems already knew.
6. The NHS National Programme For IT: The Digital Dream That Collapsed Under Its Own Weight
The NHS National Programme for IT was supposed to drag health records into a modern digital age. The ambition was defensible. The execution became one of the most notorious public-sector technology failures in British history.
The National Audit Office’s 2011 work showed projected total costs of £11.4 billion across central and local elements, with billions already spent and major care records' delivery still unresolved. The Public Accounts Committee later described the dismantled programme as carrying uncertain full costs, with the department's reported forecast at £9.8 billion excluding significant future contract-related costs.
The sunk cost fallacy here was technological arrogance. Once the state had committed to a huge centralised digital transformation, each delay became an argument for another reset, another supplier negotiation, another partial salvage attempt. The original mission — better information for better care — remained attractive enough to keep justifying a structure that was not delivering at the required pace.
This is one of Britain’s most important warnings for the AI era. Digitisation is not automatically modernisation. A giant technology programme can still fail if it ignores users, local complexity, data quality, incentives and implementation reality. The bigger the promise, the easier it becomes to confuse abandoning a flawed design with abandoning the goal itself.
7. Universal Credit: The Welfare Reform That Could Not Admit How Hard It Was
Universal Credit was created to simplify the welfare system, improve incentives and merge multiple benefits into one payment. Again, the strategic aim was not ridiculous. Britain’s old benefits architecture was complex, fragmented and difficult to navigate. But the delivery became a long-running lesson in how reform can become harder to challenge once it has become the flagship.
The National Audit Office said in 2014 that the Department for Work and Pensions had reset Universal Credit on a sounder basis, but at significant cost, by extending implementation and choosing a more expensive approach. Later NAO work in 2018 said the department planned to invest £2 billion in developing Universal Credit services.
The sunk cost was political as much as financial. Once Universal Credit became the defining welfare reform of the era, every operational difficulty could be framed as teething trouble rather than evidence that parts of the design needed more fundamental reconsideration. Claimants experienced the consequences in real life: payment waits, administrative complexity, hardship risks and uneven implementation.
Universal Credit ranks below HS2 and Brexit because it eventually did become embedded as the main system. But that may make the lesson sharper. A project can survive and still carry the marks of its sunk-cost phase. Success is not just survival. Success is whether the final system delivers enough benefit to justify the pain, cost and institutional defensiveness used to get there.
8. Crossrail: The Rare Case Where The Sunk Cost Finally Paid Back
Crossrail deserves a more nuanced ranking because the Elizabeth line is now widely used and strategically valuable. This is the uncomfortable twist: not every over-budget project is a mistake. Some expensive projects are badly managed and still become useful. That does not erase the sunk cost problem. It makes it harder to judge.
The National Audit Office said in 2021 that Crossrail’s latest cost estimate had reached £18.9 billion, exceeding the original £14.8 billion budget set in 2010 by 28%. The Public Accounts Committee had already warned in 2019 that taxpayers had been let down by delay and climbing costs, with the project then expected to be two years late and £2.8 billion over budget.
Crossrail therefore belongs on the list, but not as a simple failure. It is a warning that Britain sometimes gets the asset only after paying a chaos premium. Delayed openings, underestimated complexity and weak transparency still matter, even when the final railway is successful.
That distinction matters for HS2. Crossrail gives defenders of expensive infrastructure a powerful argument: endure the pain, and the country may eventually be grateful. But it also gives critics a sharper counterargument: if Britain normalises uncontrolled delivery because some projects eventually work, it will keep paying far more than it should for things it genuinely needs.
9. The Emergency Services Network: The Replacement That Became A Dependency Trap
The Emergency Services Network was designed to replace Airwave, the communications system used by police, fire and ambulance services. The goal was modern, high-speed and operationally important. The risk was that replacing critical national infrastructure is never just a technology purchase. It is a safety system.
The National Audit Office reported in 2019 that delays meant introducing the Emergency Services Network was forecast to cost £3.1 billion more than planned and represented poor value for money. In 2023, it said the Home Office had spent almost £2 billion on the programme since 2015 but remained far from having a working network to replace Airwave.
The sunk cost fallacy here is brutally practical. The old system cannot simply be switched off until the replacement is safe. That gives the existing provider leverage, extends legacy costs and traps the government between two expensive positions: keep paying for the old system or keep funding the delayed new one.
This is the quiet nightmare version of public-sector failure. It does not always produce dramatic headlines. It produces years of dual running, renegotiation, risk management and hidden costs. When the service is critical enough, the state loses the option of walking away cleanly.
10: The Pattern Britain Still Refuses To Name
The common thread across these cases is not that Britain should never attempt big things. That would be the wrong lesson. A serious country needs railways, energy, digital health records, welfare reform, emergency communications and crisis capacity. The problem is not ambition. The problem is uncorrected ambition.
Sunk cost fallacies thrive when leaders confuse cancellation with humiliation. They also thrive when public debate becomes too binary: support the project or betray the nation, finish the scheme or waste the money, defend the reform or admit defeat. That framing is emotionally powerful and financially dangerous.
The smarter test is colder: if the project were proposed today, with today’s facts, today’s cost, today’s timetable and today’s alternatives, would it still be approved? If the answer is no, the burden of proof should move to those demanding continuation.
Britain’s most expensive failures did not become expensive overnight. They became expensive, one defended decision at a time. The first mistake hurt. The refusal to stop, reset or redesign hurt more.
