How centralisation makes infrastructure more expensive

Or how to get politicians to care about building cheaply

In the run up to the 1995 regional elections in Madrid, the centre-right People’s Party promised to build 30 miles of new metro by the end of their four year term. In the end they actually delivered 35 miles within the four years. Compare this to the electoral promise that the recently re-elected Mayor of West Yorkshire, Tracy Brabin, made. Brabin pledged that at the end of her four year term, construction would begin, not finish, on a new tramline in Leeds, the largest city in Europe without a mass transit system.

This isn’t a dig at Brabin for a lack of ambition; in fact getting spades in the ground after four years is pretty quick in the UK, where tramlines often have to wait nearly a decade to gain the necessary approvals and funding from central government. Rather it’s a demonstration of how putting decision-making, approvals, and funding at the right level for local transport projects can make it much easier to build.

The Madrid regional assembly could raise taxes to cover the cost of construction, approve projects, and pass laws that greatly reduce the environmental assessment burden that the metro expansion faced. Mayors in the UK have none of these powers, which consequently makes building new projects a constant back and forth between central and local government, adding delays and cost, while diluting the opportunity and responsibility to deliver projects quickly and cheaply.

Alon Levy, of the Marron Institute at NYU, wrote a critique of my Works in Progress piece on how Madrid was able to triple the size of their metro in 12 years at per mile costs 10 times lower, arguing that focusing on decentralization is a mistake,

“The mentality, emanating from the UK but also mirrored elsewhere in Europe and in much of the American discourse, is that decentralization is obviously good, so it must be paired with other good things like low Spanish costs. In truth, the UK shares high costs with more decentralized countries, and Spain shares low ones with more centralized ones. The emphasis on decentralization is a distraction…”

I’m a big fan of Alon’s work and the wider Transit Costs Project. The in-depth case studies are well worth reading and contain important new data about how to build infrastructure cheaper in Anglosphere countries which if implemented would make our cities better connected and more prosperous. But I think the point about decentralization here is misconstrued and misses the wider political point about getting politicians to care about building quicker and cheaper. If countries like the UK want to lower the costs of construction of local transit projects through making construction quicker, building state capacity to deliver projects, and making trade-offs for lower cost alternatives, we need to get politicians who care about building transit cheaply.

At the most basic level, what do politicians care about?

Winning elections.

By its very nature of having local impacts, a local transport project will generally not be a decisive issue in a national election, which means that national politicians are not incentivised to care much about delivery and will instead focus on the national issues that could be decisive (e.g. defence, health care, or the economy). Of course there are instances where infrastructure in marginal seats can make a difference, hence why HS2 is in tunnels, rather than above ground, through the Chilterns (a national park near London in which many wealthy people live). But when the national government is funding a local project for these political reasons (e.g. the Chiltern tunnel), there is not a group of voters that want the project to go ahead but also be cheap.

Since there is not a group of voters who care about a project being built cheaply and quickly when local or regional infrastructure is managed nationally, there is not a politician who is directly threatened by higher costs or a failure to deliver. For example, the Lower Thames Crossing may have the most expensive planning application ever at £300 million, but when you divide that across all the households in the UK, it’s less than a tenner: hardly a make or break electoral issue. What’s more, most taxpayers in the UK won’t benefit from a specific project, so, given the difficulties of restraining costs, if they wanted to save that ten pounds, they would be more likely to want to scrap the project, or prevent projects like it happening at all, which is what usually happens in the UK.

In contrast, in sub-national elections, building local transport could be a decisive issue. We’ve already seen in Madrid how the PP’s pledge to deliver more metro helped them electorally. Before the metro expansion pledge, they had never won a regional election in Madrid, since the construction began they have yet to lose an election. In conversations I’ve had with senior PP politicians, they are well aware how the success of building the metro is tied to their electoral prospects.

The detailed results also demonstrate the electoral appeal for building local transport. In the run-up to the 1999 elections, the PP pledged to construct a circular metro line between five historically working-class cities to Madrid’s south. The party increased their vote share by just 0.1 percentage points across the Madrid community, but they increased their share of the vote by 3 percentage points across those five cities that would be connected by the new line.

To see the electoral appeal to politicians of opening metro lines quickly and cheaply, just look at when Madrid’s projects opened. In 1999, 2003, and 2007, there was a spate of openings of new lines and stations in the run-up to the regional elections in each of those years, allowing photo-ops and demonstrations of the success of the incumbent government ahead of the vote. These got so plentiful that a rival political party complained that the presence of these openings was a misuse of the PP’s office for political gain.

Alberto Ruiz-Gallardón, Mayor of Madrid, and Esperanza Aguirre, President of the Community of Madrid at the opening of the extension to Madrid Airport’s Terminal 4, taken three weeks before the 2007 regional elections

It’s not just Madrid that shows the benefits of having local/regional elections fought on plans to expand public transit with the powers that a position holds. France has experienced a tram renaissance with 21 French cities building a tramway this century. French mayors can run on a promise to build a tram in their six year term, and then have the power to do it. Approvals are through Declarations of Public Utility, done mostly locally with approval from the Department’s Prefect. The most important source of funding is the Versement Transport, which is a payroll tax approved locally. In all, French trams can be planned, funded and approved locally within two years, and then overseen and constructed in two more years, compared to 13+ years in Britain. This is part of what has led to costs being about half as much per mile as the UK.¹

Now Levy’s response to this is that Anglosphere countries have different levels of decentralization, but all have high costs.

“The OECD fiscal decentralization database has the UK as one of the more centralized governments, with, as of 2022, subnational spending accounting for 9.21% of GDP and 19.7% of overall spending, compared with Spain’s 20.7% and 43.6% respectively – but in Australia the numbers are 17.22% and 46.2%, and in Canada they are 27.8% and 66.5%”

There are three problems with using this statistic to determine how decentralised a country’s local transit construction is:

The first issue is that simply accounting for subnational spending isn’t a very good measure for decentralization of decision-making surrounding infrastructure. Madrid’s metro and West Yorkshire’s tram would both count towards subnational spending, but only one of these places has the ability for elected politicians to approve, fund, and expedite rules around construction. Most infrastructure in the UK is largely funded by the national government, but delivered by local governments, so it would be considered subnational spending under this definition, while the central government clearly sets the parameters for the expenditure. The same divergence between spending and decision-making around regulations affects American infrastructure, where partial federal funding of projects leads to a number of added time and money wasting regulatory burden, which is discussed below.

The second challenge with using these figures is that subnational spending often means spending across a large region. A Canadian or Australian state is much larger than a French city or Spanish autonomous region, diluting the political salience of a local transport project. New South Wales has 8 million people spread across 300,000 square miles, Avignon has just 92,000 people in only 25 square miles. This makes it much less likely that politicians would be incentivised to care about delivering a project quickly and cheaply.

Finally, Singapore and Hong Kong, both cited in Levy’s article, don’t have competitive elections, making them of little use to debunk the political nature of having powers at the right level. They are also tiny, dense nations with very high salaries and lots of tall buildings with deep foundations, which increases the costs of metro construction.

While America had the second highest percentage of spending at the subnational level in the OECD database Levy links to, that does not mean infrastructure is completely decentralized. In America, almost no large transit project goes ahead without federal grants, which trigger a number of cost-raising policies.

• The infamous National Environmental Policy Act requiring immense environmental impact statements is the result of federal funding mandates. These on average take 4 1/2 years, but can take up to 17 years or more.

• Federal agencies also have to create extensive reports of the impact of infrastructure spending on historic properties and archaeological sites, which often included extensive digs (by the government's own estimate 90% of all archaeological excavations in the U.S. are done under the auspices of Section 106 of the National Historic Preservation Act).

• Federally-funded projects have to abide by extensive Buy America rules on steel, iron, manufactured products, and, recently added to the list by the 2021 infrastructure act, construction materials. These either drive up project costs or make some impossible due to the absence of particular domestically-made products, which then must be acquired through an extensive waiver process.

• Since the 1970s federally-funded transportation projects in the U.S. are required to use about 10% of their contracting dollars on certain types of businesses owned by minorities, which has been shown to drive up costs (highlighted in Ezra Klein and Derek Thompson’s new book Abundance)

• The federal Davis-Bacon and related acts require many federally-funded projects to pay union-level wages even if cheaper contracts could be had from non-union outfits. Under 13(c) of the 1964 Urban Mass Transportation Act, local transit agencies receiving federal funds have to certify that they will protect existing employees and collective bargaining agreements.

• All federal funds also come with extensive grant application and reporting costs that would obviously be unnecessary if these were carried out on the state or local level.

These requirements all add costs and delay, but because local projects need to abide by them to get funding, they dilute the ability, and thus responsibility, of local politicians to control costs.

The policy implications of this argument that local politicians should be empowered to deliver local transport are quite straightforward. To get politicians to have an electoral incentive to care about transport projects, they need to have the powers to approve, fund, and oversee transit projects that roughly match up in scale with their electoral patch.

In Britain, that could mean three key changes:

• Devolve the Transport and Works Act Order process, which approve new tramlines and some underground projects, to Metro Mayors.

• Give Metro Mayors new abilities to fund projects (similar to how local taxes such as Mayoral Community Infrastructure Levy and Crossrail Business Rates Supplement funded a large fraction of Crossrail).

• Mayors should be allowed to streamline environmental assessments for clean transport projects.

Giving mayors these powers also gives them the responsibility to deliver. Future politicians will be able to win office on pledges to deliver transport projects that they will be in complete control of. If they fail to live up to their promises to build, voters will be able to hold them accountable, further incentivising delivery. Once projects complete, voters will then be able to reward successful politicians with re-election.

It’s important to reiterate that putting power to fund, approve, and expedite construction at the right levels won’t guarantee lower costs. After all, there's still the matter of completing in-house designs with improved state capacity, accounting for trade-offs in design and construction, and actually streamlining and overseeing construction. But what it will do is incentivise local political entrepreneurs to care about building transport because it could be decisive towards them winning and holding office.

Thanks to Samuel Hughes, Ben Southwood, Judge Glock, and Sam Dumitriu for their comments on drafts.

1. France doesn’t have the lowest cost per mile of other countries, which can partially be attributed to the extra public realm improvements that tramways in France tend to make. Perhaps this is because elected mayors are also concerned with enhancing the quality of their city centres alongside the provision of new transit.