Why We Can't Build Big Things Anymore

Britain can build a tunnel beneath a working city. What it struggles with is deciding what should emerge at the other end, how much it ought to cost and whether the decision will survive the next person appointed to announce it.

This distinction matters. The familiar complaint that we cannot build anything anymore suggests a country whose engineers have forgotten how concrete works. Yet the engineering continues. Beneath the argument about national decline, people are excavating, assembling and testing complicated things.

The more troubling possibility is that we retain much of the technical competence while repeatedly organising projects in ways that make that competence difficult to use.

First, acknowledge the railway

The Elizabeth line is an inconvenient fact for anyone proposing that Britain has become physically incapable of major construction. Its central section opened in 2022, and the Office of Rail and Road recorded 257.4 million journeys across the line in 2025/26.

That is evidence of substantial use, not a retrospective pardon for every decision made during construction. Crossrail arrived late and over budget. In its 2019 investigation, the National Audit Office identified a compressed schedule, contractual problems and the absence of a realistic plan.

Both things can be true. A project can be badly managed in important respects and still become valuable infrastructure. Pretending that eventual usefulness excuses everything is complacent. Pretending that overruns make the finished railway worthless is equally unhelpful.

The question is why delivering something useful so often requires an extended public demonstration of how not to organise it.

The first number wins

An early estimate occupies a strange position in public life. It may be based on incomplete design and unresolved assumptions, but once announced it acquires the authority of a promise.

Later information then looks like failure. Discovering what the project actually requires becomes “cost escalation”, even when part of the problem was that the original figure never adequately described it.

This does not excuse poor estimating. It explains why honesty at the beginning matters. A range with explicit uncertainties is less satisfying at a lectern than a single confident number, but the ground has never shown much interest in the requirements of a press conference.

The NAO’s review of lessons from major programmes identifies recurring problems with scope, planning, interdependencies and oversight. These are not exotic engineering failures. They concern establishing what is being attempted and maintaining a credible account of how it will happen.

A project cannot become affordable through the force of its initial announcement. It can merely become embarrassing later.

We keep changing the thing

There are legitimate reasons to alter a project. Demand changes, evidence improves and an early design may turn out to be wrong. Refusing all revision would replace one form of incompetence with another.

But changes have consequences beyond the meeting in which they are approved. Designs must be reconsidered. Contracts, construction sequences and connections to other work may need to change. A decision that looks small in a briefing can travel a considerable distance through a programme.

Euston provides a documented example. In March 2023, the NAO reported that the latest estimate for HS2’s ten-platform station design was £4.8 billion, £2.2 billion above its budget. It found that the department and HS2 Ltd had not developed an affordable scope integrated with the other activity at the site.

Those are historical figures for that design, not today’s complete HS2 bill. The lesson is about the difficulty of agreeing a buildable proposition where several projects and organisations meet.

The passenger eventually wants a station. Before that can happen, the institutions involved must agree which station they are collectively trying to produce.

Permission and competence are different problems

Planning is an obvious target because its delays are visible and its language invites parody. A document explaining the preliminary requirements for consulting on an assessment can test the commitment of the most patient democrat.

But consultation has a purpose. Big projects impose costs on people who may receive few of the benefits. Homes, businesses and landscapes are not irrelevant obstacles because somebody has drawn a compelling line across a map.

The practical issue is whether the process reaches decisions reliably, with clear requirements and enough competent people to assess the evidence. The government’s 2023 infrastructure-planning reform plan identified insufficient capacity and capability as barriers to faster, better decisions.

Removing a requirement and employing someone capable of assessing it are different interventions. Sometimes the rule is unnecessary. Sometimes the queue exists because the necessary person is missing.

An understaffed institution does not become efficient merely because a minister has become impatient with it.

The handover is part of the machine

A major project is not simply a collection of structures. It is an arrangement of things that must function together.

A station may look finished while the work needed to operate it safely remains incomplete. Tunnels, trains, signalling and staff procedures must form a railway. Completing the visible object is not the same as completing the service.

The NAO’s 2021 Crossrail update found that the revised schedule agreed in 2019 had been unachievable because the programme was further from completion than understood.

That is a particularly expensive kind of misunderstanding. It reveals how progress can be reported convincingly while the dependencies that determine readiness remain unresolved.

The same problem exists in software, hospitals and almost any complicated system. Everybody can finish their assigned part while the whole remains unusable. Responsibility has been distributed successfully; completion has not.

The past had costs too

Nostalgia supplies an easy answer: previous generations simply got on with it. We remember the bridge that survived and the engineer whose portrait accompanied it. The disruption, arguments and discarded schemes are less likely to appear on the commemorative mug.

There is no need to pretend every modern requirement is sensible. Nor should we mistake the ability to disregard people for a transferable engineering method. Faster construction is desirable; making affected communities powerless is not a sufficient delivery strategy.

The useful comparison asks what earlier arrangements enabled that ours obstruct. Was authority clearer? Could teams retain expertise between projects? Was the design more repeatable? Which costs were counted, and which were imposed elsewhere?

Without those questions, “the Victorians would have built it” is mainly a way of recruiting dead people to support a procurement opinion.

Ambition needs continuity

The answer cannot be another declaration that we should become more ambitious. We have considerable experience of those. Ambition must survive contact with funding, design and the person responsible for an awkward connection beneath a road.

A serious approach would spend enough effort defining the project before treating its price and date as settled. It would make changes visible, including their consequences. It would give capable delivery teams authority while requiring them to report uncertainty rather than conceal it.

It would also distinguish a necessary cancellation from a convenient postponement. Stopping a bad project can be responsible. Repeatedly interrupting a worthwhile one is not automatically a saving because less money leaves the account this year.

We can still build big things. The harder task is maintaining a sufficiently stable intention long enough for people to finish them.

At the opening ceremony, someone will praise the vision that made it possible. They might also thank the people who preserved the drawings, resolved the interfaces and continued answering difficult questions after the vision changed hands. There may not be room for all of them on the platform. Fortunately, they will understand why.