Is centralisation to blame for Britain's high construction costs?

A response to Alon Levy of Transit Costs Project

Fact 1: Britain is an extremely expensive place to build just about any kind of infrastructure.

Britain is currently building both the world’s most expensive nuclear power station and the world’s most expensive high speed railway line. British tram projects cost on average twice more than the average European tramway and three and a half times more than the average German tramway. The planning application for the Lower Thames Crossing (a road tunnel between Kent and Essex) cost more to produce than it cost Norway to build the world’s longest road tunnel and the world’s deepest subsea tunnel combined.

Fact 2: Britain is an extremely centralised country. Local leaders in Britain lack the power to independently approve or fund (via local taxes) new transport infrastructure.

To obtain permission to build new transport infrastructure, English local authorities must submit a planning application (Transport Works Act Order/TWAO) to the Department for Transport. In the case of a recent one mile tram extension, it took four years from submission to approval for the West Midlands Combined Authority (WMCA) to obtain the TWAO.

To fund projects, cities and regions submit bids to the Department for Transport, who in close collaboration with the Treasury, decide whether or not to fund projects from general taxation. By contrast, Dijon (France) was able to plan, approve, fund, and build a 12 mile tram network in just 4 years. The project was funded via a local payroll tax on large employers (the Versement Transport). On a per-mile basis, the West Midlands project cost six times more than the Dijon project.

***

Is the latter (centralisation) to blame for the former (high infrastructure costs)? Alon Levy of the Transit Costs Project (TCP) doesn’t think so. In a recent blog, they argue that while there may be good reasons to devolve power, Britain’s centralisation isn’t the driver for Britain’s high infrastructure costs.

For those unaware, Alon Levy and the TCP have been extremely influential in highlighting the massive disparity in infrastructure construction costs between the Anglosphere and the rest of the world. The TCP has produced a number of country-specific deep dives into infrastructure construction. Readers can skim the Sweden, Italy and Istanbul studies to find out what works, or read the Boston and New York studies to find out what doesn’t. For those looking for even more lessons on what not to do, there will soon also be a UK study, covering Crossrail, the Northern Line Extension, and the DLR.

Levy’s key takeaway from this, as yet unpublished, UK study is that Britain’s infrastructure cost problems are not the result of excessive centralisation, but rather down to the gradual adoption of what Levy dubs the ‘globalised approach’. In essence, Britain, like other Anglosphere countries, outsources almost everything. Low-cost countries like Sweden and Italy maintain a core of engineering expertise within the state and complex projects are designed in-house then put out to tender. Britain, by contrast, is heavily reliant on private-sector consultants to design projects and public tenders bundle up design and construction. When projects finish in Britain, teams disband. Lessons learnt are lost. By contrast, the Swedish engineers who designed the Citybanan (a crossrail-style tunnel through Stockholm connecting up commuter rail lines) moved on to designing Nya Tunnelbanan (a massive expansion of the city’s metro system).

Levy points out that Italy can build cheaply (far cheaper than Britain) despite high levels of centralisation. What sets Italy apart is the deep banks of engineering knowledge within the state – knowledge that the British state can only access when it hires consultants at high cost (and at a severe informational disadvantage). It should be noted Levy concedes that Italian centralisation does differ in one key way: Britain relies on ministerial approvals while Italy lets civil servants with expertise in engineering get on with it.

***

My view is slightly different. I am persuaded that Levy is right that success is less about whether power is held centrally or locally and more about whether there is a clear programme of construction with in-house engineering expertise retained across projects, limited use of consultants and flexibility for builders to make changes . There is good evidence, for instance, that when the best engineers retire subsequent infrastructure projects are more expensive. However, I still think centralisation is an important part of the story. Let me explain.

I have argued before that while the direct costs of regulation, such as the requirement to build a £120m bat tunnel (a tenth of a percent of HS2’s total budget), are insufficient to explain why British infrastructure is so expensive to build, regulation is still a major cost-driver via an indirect route.

What matters for cost reduction is ultimately the ability to retain expertise between projects, to build supply-chains that spread fixed-costs over multiple projects, and to innovate in construction and design. The problem is regulation undermines all of these proven cost-reducers. In the case of nuclear, safety regulations force frequent design changes between projects undermining learning-by-doing. While a long drawn-out and uncertain planning process (plus the threat of judicial review) mean that investments in supply chains become incredibly risky – workers and equipment may sit idle for years between projects.

So why might centralisation be an issue? Because it exacerbates the planning and regulatory problems that drive costs increases.

In Britain, local authorities have every incentive to block new development and little incentive to say yes. New homes put pressure on public services and infrastructure. In some parts of the world, new development also brings a new cash flow. A $500,000 (£400,000) home in Houston generates around $9,000 (£7,200) in annual property tax revenue. Council tax in England is far lower, a home of equivalent value might bring in just £2,200 a year, roughly a third as much. And in Britain, councils have fewer freedoms to spend that revenue on their priorities – most of that £2,200 will go on statutory services like social care.

Britain’s most significant tax on housing is Stamp Duty, which flows directly to HM Treasury (and not to the local authority). Business Rates, which apply to commercial development, are only partially retained. Outside of special areas, local authorities only receive half of the revenue raised from business property taxes on new development. And funding formulas for local government often change for redistributive ends. Put simply, the fiscal incentives to approve development in Britain are extremely weak.

Some argue that the way to overcome local opposition to development is shifting decision making up a level. Labour’s re-imposition of housing targets and California’s Builders’ Remedy (where places that fail to build are forced to approve new homes) are two examples of this approach. The Nationally Significant Infrastructure Project system, where planning approval for major transport and energy projects is decided on by independent central experts (and not the areas affected) is another.

The problem is that nationalising planning policy doesn’t eliminate anti-development energy. It redirects it. The most effective way to stop unwanted development locally may not be to oppose it directly, but to lobby for national rules that make all types of development harder.

Want to stop new homes being built on farmland near the edge of your town? Lobby for a stronger green belt that blocks homes being built on farmland on the edge of every town. Want to stop a new motorway that might shift traffic to your area? Campaign for tougher nature protections that apply equally to motorways in Kent, trams in West Yorkshire, and nuclear power stations in Somerset. Anti-pylon activists in East Anglia are some of the strongest defenders of the National Landscapes duty, which blocked a station car park expansion in Essex, created problems for airport expansion in Bedfordshire, and delayed housebuilding in Kent.

And if you are a Government minister who wants to repeal a law that makes it easy to sue developers or a law that mandates extensive consultation, be prepared to fight a coalition of people opposed to motorways in Kent, homes in Hampshire, and nuclear power stations in Somerset who would have little in common other than opposing development near them.

Not every development is unpopular, but even popular developments have to comply with regulations and requirements designed to stop unpopular developments. There are few environmental wins more clear-cut than taking cars off the road and replacing them with trams. Trams are good for air quality, cut carbon emissions, and reduce noise. They are also popular. Poll after poll shows high levels of support for a tram in Leeds. Yet the same environmental bureaucracy built-up to delay and block more controversial types of development is also holding up popular new green infrastructure. Over 5,000 environmental surveys have been carried out for the now-delayed Leeds tram.

In other words, the incentive mismatch caused by Britain’s excessive centralisation has not only prompted the creation of tools to stop controversial development, but also created the crud that makes it hard to build almost anything.

This incentive mismatch is not the only driver of the growth of planning and environmental red-tape. People support environmental regulation in part because they sincerely want to protect the environment. Post-Grenfell fire safety rules make it much harder to build new residential towers, but they were adopted out of a genuine concern for safety. Planners genuinely believe that the planning restrictions they advocate around light, outdoor space, height, and so on, do create more ‘liveable’ places. Yet centralisation matters here too. If local authorities depend on new development for revenue, they would have a strong incentive to push back on excessive restrictions on building. There would be a political cost to regulatory growth. The problem is for many opponents and veto-players, the trade-off for greater environmental protection/fire safety (less development) is a feature not a bug.

France’s rapid nuclear buildout is sometimes understood as a triumph of the state overriding local objections. Tony Benn was famously told by a French official that ‘when you drain a swamp, you don’t consult the frogs’. Sizewell C by contrast had 7 separate consultations in 8 years. Yet there’s another side to the story. France’s tax system, which levied taxes on business structures (a bit like business rates), meant that areas which hosted nuclear power stations benefitted massively financially. Alex Chalmers notes that “French councils paid an average of €35 in subsidies … per local resident. In the 19 areas that hosted nuclear power, the average was €450.” This may be why support for France’s nuclear programme remained strong even in the wake of disasters like Chernobyl.

It is a world away from the incentives in Britain. Hinkley Point C’s community benefit package of £128m over 40 years works out to roughly £6 per Somerset resident per year. Britain’s tax on business structures provides much weaker incentives to approve development: councils retain only 50% of the rates they raise locally, face levies of up to 50% on ‘disproportionate’ growth, and see gains redistributed nationally based on ‘needs’ at regular intervals. Unlike in France, Hinkley Point C’s rates windfall will not lead to local tax cuts visible on every household bill.

There are other ways centralisation creates problems in Britain. Transport projects are typically funded out of national pots. There is a separation between who pays, who benefits and who approves. If a project is over-budget or gold-plated in various ways, locals don’t bear the burden. Londoners are not paying £4bn more in tax because Crossrail was £4bn over-budget. Instead, the costs are spread across the whole country. This can happen, as it did with HS2, but when a project has sufficient momentum behind it, it becomes a prime opportunity to get pet projects funded.

Britain’s planning system contains multiple veto points where public bodies who have, at best, weak incentives for projects to be delivered cost-effectively can cause significant delay (and in some cases, kill projects all together).

When Crossrail was being planned, Tower Hamlets Council submitted a list of 96 objections. After extensive negotiation Crossrail were able to resolve the dispute – by accepting 94 of them – though the Council came back with even more. Due to the length of Crossrail, this process was repeated for 15 other councils. In some cases, redesigns in response to objections (the Fire Brigade requested eight large emergency evacuation shafts across Central London) were themselves redesigned (read: cancelled) in response to objections from boroughs.

This isn’t Alon Levy’s first foray into the centralisation-cost debate. In response to my former colleague Ben Hopkinson’s deep dive on the ultra-cheap Madrid Metro, Levy responded by arguing that the article’s focus on decentralisation was misguided, in part, because OECD data suggests that while Spain is decentralised (and cheap), many other cheap countries are highly centralised (e.g. Italy). And some of the most decentralised countries, as measured by the OECD, are the US, Canada and Australia are also the most expensive.

The problem with Levy’s argument here, as Hopkinson points out, is that OECD fails to capture some of the most important aspects of why decentralisation is important. For example, the Madrid Metro and the West Midlands Metro both count as subnational spending under the OECD’s criteria, but Madrid’s ability to plan, control, and finance the project is hugely different. Likewise, there is a difference between devolving power to a region like the community Madrid and devolving power to a US state – a much larger (and more populous) entity.

It isn’t devolution itself that is important, but rather whether institutions are incentivised to support growth. In Britain, as it stands, there’s a mismatch of incentives and powers. Local authorities lack the power to approve and fund new infrastructure. And while in most places, an uplift in development as a result of new infrastructure is a fiscal benefit, in the UK it is often a fiscal burden. This creates hostility to development and drives a growing regulatory burden on new development whether that’s infrastructure or housing.

Levy’s attack on the ‘globalised model’ is compelling. The hollowing out of state capacity has made it harder to build. As Britain has relied more on consultants and contracts designed to shift risk elsewhere, building infrastructure has become more expensive. But, we need to look deeper. Why have we hollowed out state capacity and tried to shift risk elsewhere? The cause is a political system that generates risk, uncertainty, and delay at every turn.