The cost premium on British infrastructure is so high that it is unlikely to be the result of a single cause. A lot needs to go wrong to generate a 65% cost premium. We identify several causes: sclerotic planning, centralisation, inefficient procurement, and a lack of sustained investment pipelines all play a role. These causes interact and compound one another in ways that make each individual problem harder to fix.

Procurement and State Capacity

Britain's approach to infrastructure procurement has hollowed out the engineering expertise that lower-cost countries retain within the state. Most European countries maintain deep in-house technical capacity. Complex projects are designed internally before being tendered, giving the state both cost control and the ability to scrutinise contractor claims. Britain, by contrast, relies heavily on private consultants for design work, and typically bundles design and construction together in contracts intended to transfer risk to the private sector.

The consequences are significant. When projects finish, teams disband and institutional knowledge is lost. Each new project effectively starts from scratch, rebuilding expertise at cost. The feast-and-famine pattern of British infrastructure investment compounds this: Crossrail was the first major rail tunnelling project since the Jubilee Line extension in 1999, which was itself the first since 1979. At present there are no plans to deploy the tunnelling expertise built up on Crossrail on any subsequent project. Rail electrification illustrates the same pattern. For most of the last thirty years Britain has electrified few or no miles of track annually, with only occasional bursts of activity. Germany, by contrast, electrified an average of 200km of railway per year for decades although the rate has slowed in recent years,27 allowing consistent accumulation of expertise and steady supply chain development. For example, the Great Western Main Line electrification cost £2.8bn (around £11.5m per mile of electrified track) while Germany's electrification of the Ulm to Lindau line cost around €370m (£2.1m per mile of electrified track).28

In 2019 the International Transport Forum found that rail electrification cost between 0.5 to 2 million Euros per single track kilometre in Europe. In the UK costs in 2018 were estimated at £3.9 million pounds per single track kilometre.29

The risk-transfer model of procurement also often backfires. Contracts designed to insulate the public sector from cost overruns tend to reduce competition and push prices up accordingly. On one section of the A9 dualling in Scotland, only a single bid was received.30 When the state lacks the technical capacity to scrutinise contractor proposals, it is poorly placed to identify gold-plating or challenge inflated cost estimates. The informational asymmetry runs in one direction, and costs reflect it.

These four factors, a planning system that imposes both direct compliance costs and large indirect costs by blocking proven cost-reduction strategies; a centralised structure that misaligns incentives against development; a procurement model that has eroded state engineering capacity; and a stop-start investment pattern that prevents expertise and supply chains from developing, are mutually reinforcing. Fixing any one in isolation would help. But Britain's infrastructure cost premium is ultimately the product of all four working together.

Case Study: A9

Scotland has generally been more successful than other parts of the UK at delivering road projects, building more roads per capita and often doing so on time and on budget. However, even here, major projects have faced significant delays and cost pressures.

The dualling of the A9 between Perth and Inverness illustrates these challenges. Originally scheduled for completion by 2025, only 11 of the planned 83 miles had been delivered by that date. The expected completion date has now slipped to 2035.31

A key issue has been delays in the planning process. Initial estimates suggested that statutory approvals would take around six years. In practice, the process has taken far longer. Fifteen years on from the announcement the project would go ahead, and one year after the project was originally due to be complete, parts of the route are still awaiting planning permission.32

Financing challenges have compounded these delays. Changes to EU and UK accounting rules meant that the original funding model, which relied on a mix of public and private finance to keep debt off the government’s balance sheet, was no longer viable. As a result, a project that was already difficult to fund on time became significantly harder to deliver on schedule.33

Procurement decisions also played a role. In an effort to limit public sector exposure, the Scottish Government transferred substantial risk to contractors. While this increased cost certainty for the government, it reduced competition. For one section of the road (Tomatin to Moy), only a single bid was received.34 This was rejected as poor value for money, forcing the procurement process to restart and delaying the project by more than two years.

Even in the part of the UK that has performed relatively well on transport delivery, long planning timelines, funding constraints, and procurement choices have combined to drive delays and increase costs.

Planning and Regulation

The most visible driver of high infrastructure costs is Britain's planning and regulatory system. Compared to peer nations, it is more complex, more time-consuming, and more expensive to navigate. In some cases, as with the Lower Thames Crossing, the cost of preparing a planning application alone exceeds what other countries spend building comparable infrastructure outright.

Case Study: Lower Thames Crossing

The Lower Thames Crossing, a planned road tunnel linking Essex and Kent, illustrates the scale of the problem. Its planning application ran to 359,070 pages, involved around 150 staff, and cost almost £300 million to write. The environmental statement was over 12,000 pages long.35

The full project, 14 miles of road including a 2.6-mile tunnel, is estimated to cost between £9 and £10.6 billion. By comparison, Norway’s Lærdal Tunnel, the longest road tunnel in the world, cost around £140 million36. Differences in geology mean the UK project was never going to be this cheap. But the comparison is still stark: Britain has spent twice as much as Norway spent building the world’s longest road tunnel simply on planning.

On top of an environmental impact assessment that stretched to over 12,000 pages37. The application includes 31 “Principal Areas of Disagreement” summaries, 378 separate “Statements of Common Ground”38 with bodies ranging from Royal Mail to NHS North East London, and even a 38-page report assessing whether consultation on the consultation process itself had been adequate.39

This is not unique to mega-projects. Smaller schemes face similar burdens. Upgrading just four miles of the Newark Bypass to a dual carriageway required 66,000 pages of planning documents.40 Laid end to end, these documents would stretch around 12 miles, three times the length of the road itself.

Multiple government bodies can require design changes, impose additional conditions, and delay approvals at any stage of a project. Consultation requirements, often backed by the threat of judicial review, add further cost and uncertainty. The result is planning applications running to hundreds of thousands of pages, sustained legal and consultancy expenditure, and projects that take years longer to approve than equivalent schemes elsewhere in Europe.

Yet the direct costs of compliance (bat tunnels, environmental surveys, bespoke mitigations) account for only a small part of the cost increases caused by regulation. Regulation also imposes large indirect costs by undermining the strategies that reliably reduce infrastructure costs elsewhere. Four mechanisms matter most.

  1. A focus on process and risk aversion lead to gold-plating. Unwieldy process and the desire to avoid legal challenges requires those building infrastructure to be maximally cautious at every stage. The fear of judicial review, or that a regulator will impose expensive design changes later, means builders often gold-plate their designs. Because builders cannot be certain what regulators and courts will ultimately require, large sums are spent on compliance that goes well beyond what would actually be demanded.
  2. A broken system inhibits learning by doing. Costs fall substantially as construction teams repeat the same design across multiple projects. Welders at Hinkley Point C's second reactor building are working at four times the pace of the first, generating an estimated 30% cost saving.41 But frequent design changes, often demanded by regulators mid-project, prevent standardisation and erode these gains. What works for the first project must be redesigned for the second, resetting the learning curve each time.
  3. Uncertainty deters investment in supply chains. Fleet construction, building multiple projects in sequence using the same workforce, equipment, and methods, is one of the most effective ways to reduce infrastructure costs. South Korea builds nuclear plants at around a sixth of British costs in part because it has maintained a continuous construction programme since 1972.42 But supply chain investment only makes sense when there is a reliable pipeline of work. Lengthy planning processes and the risk of legal challenge mean that workers may sit idle and equipment unused for years between projects. Firms respond rationally by not making the investment in the first place. One consequence is that Britain’s biggest construction firms are small by international standards and rely heavily on subcontractors – reducing economies of scale and adding costs.43
  4. Inflexible national rules block the use of proven construction methods that have worked well elsewhere. For example, in Britain tram tracks have to sit on concrete slabs of 500 to 1000mm in depth, to protect utilities underneath the tracks. However Strasbourg, France and Vienna, Austria laid tram tracks using a shallow bed that was then covered with grass. This leads to a depth of approximately 300-400mm. These shallower beds are standard in France..44 All of this extra digging and extra concrete adds cost, which is part of why Britain’s trams cost so much more than their European equivalents. Innovative local authorities such as Coventry City Council have creatively worked around these rules to reduce costs. Their Very Light Rail project uses a shallow 300mm trackform that can be laid without the expensive utility diversions that make conventional tram schemes so costly. It can be built at a cost of £7m per kilometre (making it much cheaper than many European projects), but has a lower capacity.  

Centralisation and Misaligned Incentives

Britain is one of the most centralised countries in the developed world, and this exacerbates the planning problems described above. English local authorities have limited power to approve or fund new transport infrastructure independently. Permission for trams and metros must be sought from the Department for Transport; funding is allocated by central government from national budgets. The separation between who pays, who benefits, and who approves creates a systematic misalignment of incentives. In contrast in Spain, Germany and France local politicians make the decisions on what gets built and it is mostly funded by local taxpayers. Decentralisation allows incentives to be aligned.

This centralisation causes incentives to be misaligned for projects of national significance too. In France, localised taxation means that areas hosting nuclear power stations benefit substantially. Local governments in those areas have been able to offer residents free wi-fi, free electricity, and significant tax reductions45 because they are raising so much money from the nuclear power plant. This turns local politicians and residents into active supporters of nuclear development in their area. In England, by contrast, local authorities only retain at most half of the business rates generated by new local development.46 They do not capture, at least not directly, any of the employment or corporate taxes raised as a result of the project.

The British system goes further than mere indifference: it actively incentivises obstruction. Local authorities bear none of the costs when projects go over budget, but can use the planning system, the courts, and political lobbying to extract design changes that benefit their area at the expense of the overall scheme. With no stake in cost control, demands that have poor cost-benefit ratios are entirely rational from a local perspective.

For example, Tower Hamlets Council was supportive of the Crossrail project that run through their borough in London. However, this did not stop them submitting a list of objections that ran to 96 clauses.47 After extensive negotiations, Crossrail Ltd resolved all but two by giving in. Later, when the project’s plans changed slightly, Tower Hamlets submitted yet more objections: to the use of a local park as a temporary construction site, on the size of Whitechapel Station (and whether it should encompass a nearby McDonald’s), to some vacant land being used by another rail contractor, and to night-time closures of the Blackwall Tunnel Approach Road.48 There is no evidence that anyone ever conducted even a rudimentary cost benefit analysis on these changes, all of which added to the bill. This is for an infrastructure project the local council actually supported, the councils who oppose infrastructure projects are incentivised to be much more obstructive.

France offers an instructive contrast. Benefiting local governments there contribute a large percentage of the cost of major infrastructure projects,49 giving them a direct stake in keeping costs down. Local authorities can still request design changes, but are expected to pay for them if granted. The system is far from perfect, France's national audit office has documented cases where paid-for changes contradict the original intent of a scheme,50 but the incentive structure is fundamentally healthier. In Britain, local governments see few of the benefits when schemes go well and are insulated from the costs when they go badly. Rather than acting as partners and promoters of infrastructure, local authorities too often become obstacles, using projects to extract marginal local benefits at disproportionate cost to the public purse.

Case study: Edinburgh Tram

The first phase of Edinburgh’s tram is a good example of how British infrastructure costs are driven up not just by bad luck or difficult engineering, but by poor project setup and weak governance. The project was approved in 2007 on the basis that Phase 1a could be delivered within a £545 million funding envelope and opened by summer 2011.51 In the end, the line that opened in May 2014 was shorter than planned, running only from the airport to York Place rather than Newhaven (2.9 miles short of the planned 11.5 mile route), and cost £836 million.52 The core problem was that decision-makers were given a misleading sense of certainty. Contracts were signed before the design was mature, before utilities had been properly mapped and diverted, and before key approvals were in place.53 A project that was supposed to transfer risk to the private sector instead became trapped in disputes, delays and cost escalation. At £96 million per mile it was one of the most expensive tram projects in the world.

However the second phase of Edinburgh’s tram construction was much more successful, even if still significantly more expensive than most European counterparts. It cost £35 million pounds less per mile than phase 1, and was delivered on time and on budget.54 This highlights how having a steady stream of infrastructure projects allows lessons to be learned and teams to retain their skills reducing costs.