In New York, a well-publicized homicide by pushing the victim onto the subway tracks created a conversation about platform edge doors, or PEDs; A Train of Thought even mentions this New York context, with photos from Singapore.
In Paris, the ongoing automation of the system involves installing PEDs. This is for a combination of safety and precision. For safety, unattended trains do not have drivers who would notice if a passenger fell onto the track. For precision, the same technology that lets trains run with a high level of automation, which includes driverless operations but not just, can also let the train arrive with meter-scale precision so that PEDs are viable. This means that we have a ready comparison for how much PEDs should cost.
The cost of M4 PEDs is 106 M€ for 29 stations, or 3.7M€ per station. The platforms are 90 meters long; New York’s are mostly twice as long, but some (on the 1-6) are only 70% longer. So, pro-rated to Parisian length, this should be around $10 million per station with two platform faces. Based on Vanshnook’s track map, there are 204 pairs of platform faces on the IRT, 187 on the IND (including the entire Culver Line), and 165 on the BMT (including Second Avenue Subway). So this should be about $5.5 billion, systemwide.
Here is what the MTA thinks it should cost. It projects $55 million per station – but the study is notable in looking for excuses not to do it. Instead of talking about PEDs, it talks about how they are infeasible, categorizing stations by what the excuse is. At the largest group, it is accessibility; PEDs improve accessibility, but such a big station project voids the grandfather clause in the Americans with Disabilities Act that permits New York to keep its system inaccessible (Berlin, of similar age, is approaching 100% accessible), and therefore the MTA does not do major station upgrades until it can extort ADA funding for them.
Then there is the excuse of pre-cast platforms. These are supposed to be structurally incapable of hosting PEDs; in reality, PEDs are present on a variety of platforms, including legacy ones that are similar to those of New York, for example in Paris. (Singapore was the first full-size heavy rail system to have PEDs – in fact it has full-height platform screen doors, or PSDs, at the underground stations – but there are later retrofits in Singapore, Paris, Shanghai, and other cities.)
The trains in New York do not have consistent door placement. The study surprisingly does not mention that as a major impediment, only a minor one – but at any rate, there are vertical doors for such situations.
So there is a solution to subway falls and suicides; it improves accessibility because of accidental falls, and full-height PSDs also reduce air cooling costs at stations. Unfortunately, for a combination of extreme construction costs and an agency that doesn’t really want to build things with its $50 billion capital plans, it will not happen while the agency and its political leaders remain as they are.
The state is to a large extent a coordinating body. Even the more extractive aspects of it, like historically the military, succeeded or failed not by who was the most brutal (they all were brutal) but by who was most efficient at organizing large groups of people.
Coordination in public transit is especially important, because it’s a system with many moving parts: infrastructure, equipment, timetable, development. These do not accrete spontaneously, not in any society that has also invented cars; transit-oriented development in the 21st century looks different from historic development before mass motorization. Organizational capacity makes the difference between a state that grows around mass transit, like Japan or South Korea or Switzerland or Sweden or increasingly France, and one that grows around cars even when the goal is nominally transit first, as is common in the United States but also most of Southeast Asia.
So in general, better coordination means overall better public transit. But it specifically means better investment – more targeted at the right places. And this is especially visible in mainline rail, which is less self-contained than urban metro lines. The right way to plan is to get different bodies to cooperate, such as different railroads and government agencies. And then there is the wrong, American way.
Coordination versus wishlists
In theory, the United States has mechanisms to get different agencies to talk to one another. The Northeast Corridor planning process understands that the corridor has many users and owners: Amtrak, MBTA, Connecticut DOT, MTA, New Jersey Transit, SEPTA, MARC. To ensure they collaborate, there are layers set on top of them, like the NEC Commission.
And yet, the NEC Commission’s plans are not worth the paper they are written on, and the people involved should not work in this field or in government again. The problem is that their idea of coordination is to ask each of the above agencies what its wishlist is, collate the responses, and staple them together.
The wishlist staple job is the opposite of coordination. Coordination means sitting down with intercity and regional rail operators, figuring out their service needs, and writing down a timetable with associated infrastructure plan that maximizes service at minimum cost. Even the accidental moves toward coordination that do exist, like the MBTA plan to complete electrification of the Providence Line and run modern EMUs rather than diesels under catenary, do not figure into the plan: Amtrak still wants a third track on the Providence Line, which such electrification obviates even if Amtrak cuts its Boston-Providence trip time in half. The third track was said to cost $400 million years ago; I do not know if it is still its budget or whether costs are higher now. One such unnecessary project at a time is what it takes to turn what should be a $15 billion project into three-figure billions.
This wishlist mentality is present whenever bad planners (e.g. all Americans) try to do something that involves more than one agency. It’s assumed that different parts of the government must constantly be at one another’s throats. Unless one agency dominates, the only solutions in this mentality are either to do a staple job, or subordinate all agencies to one new hierarchy, typically run by people who have never run transit service and do not respect those who have.
How to plan mainline rail better
Three of the legs of coordinated planning – infrastructure, rolling stock, timetable – are coordinated in an excellent way in Switzerland. (Switzerland is unfortunately too NIMBY for modern TOD.) This does not mean slavishly copying every single Swiss decision, but it does mean that it behooves planners to learn how Swiss rail planners got Europe’s best rail network on a limited (though not quite austerity) budget.
The way it should work is that everything begins from the timetable. Trains must run on the same fixed interval – typically hourly, but denser services should be planned around shorter intervals like 30 minutes or smaller divisors of the hour. This provides the base level of coordination: connections between trains at major stations are to be done at times that are compatible with this interval.
If the trip time between major stations (“Knoten”) is just a bit too long for timed connections at both ends, it means that the trains should be sped up. This is the run trains as fast as necessary maxim, beloved by many high-speed rail opponents who bring up that maxim far more often than they bring up how much rail tunneling Switzerland has built.
Everything must come based on this plan. The choice of rolling stock must be compatible. Switzerland chose bilevel EMUs, because its use case is urban stations with a surplus of platform tracks but limited platform length; the bilevel trades off higher on-train capacity per unit of train length for lower egress capacity, and in a country where the main train station has 26 tracks, the bilevel is the correct choice. Maybe in another environment it is and maybe it isn’t; in New York it is not.
The slate of infrastructure projects must likewise be based on total integration of operations and capital planning. This means being able to trace delays to their source, using data to figure out what the most problematic areas are, and fixing them. Swiss trains are not inherently punctual; delays in the 5 minute range are routine. What sets them apart is that the infrastructure has been designed, at minimum cost, to ensure that delays don’t propagate, whereas in Germany, cascading delays are more common, and the less said about the United States, the better.
Swiss integration, to be clear, operates in an environment that is highly federal, has a smattering of private railroads interoperating with SBB, is stingy about public spending, and has in most cases Western Europe’s most privatized economy. And yet there is no separation of infrastructure and operations, in contrast with the trend in Britain and the EU.
Coordination and saying no
A planning agency that has to work with operators to ensure they all collaborate has to mediate conflict in many cases. This is the origin of the wishlist mentality: by planning overly expensive systems with maximum separation between operators, conflict is avoided, at the minor cost of an order of magnitude increase in the budget.
A better way to mediate is to either propose compromises, or outright saying no. Investment that is not part of the coordinated plan is extra and infrastructure plans should not burden the taxpayers with it. If different bodies conflict, sometimes one is right and the other is wrong, and the infrastructure planners should say so; sometimes who is right and who is wrong is consistent, sometimes it isn’t. Moreover, if bodies refuse to coordinate, it’s important to be able to say no to overall plans.
All of this interfaces with previous posts on this subject. In particular, the infrastructure investment program, whether it’s a regional Verkehrsverbund or an intercity system like the NEC Commission, should consist of subject matter experts. Senior politicians should understand that those experts are paid to maximize the efficiency of an enormous infrastructure program and therefore defend their expertise against attacks.
This is the start of a multi-post series, of undecided length, focusing on institutional, managerial, and sociopolitical issues that govern the quality of infrastructure. I expect much of the content to also appear in our upcoming construction costs report, with more examples, but this is a collation of the issues that I think are most pertinent at the current state of our work.
Moreover, in this and many posts in the series, the issues covered affect both price and quality. These are not in conflict: the same institutions that produce low construction costs also produce rigorous quality of infrastructure. The tradeoffs between cost and quality are elsewhere, in some (not all) aspects of engineering and planning.
The common theme of much (but not all) of this runs through procurement. It’s not as exciting as engineering or architecture or timetables – how many railfans write contracts and contracting regulations for fun? – but it’s fundamental to a large fraction of the difference in construction costs between different countries. Some of the subheadings in this post deserve full treatments by themselves later, and thus this writeup is best viewed as an introductory overview of how things tie together.
What is procurement?
Procurement covers all issues of how the state contracts with providers of goods and services. In the case of public transportation, these goods and services may include consulting services, planning, design, engineering, construction services, equipment, materials, and operating concessions. The providers are almost always private-sector, but they can also be public companies in some cases – for example, Milan Metro provides consulting services for other Italian cities and Delhi Metro does for other Indian ones, and state-owned companies like RATP, SNCF/Keolis, and DB/Arriva sometimes bid on private contracts abroad.
The contracting process can be efficient, or it can introduce inefficiencies into the system. In the worst case I know of, that of New York, procurement problems alone can double the cost of a contract, independently of any other issue of engineering, utilities, labor, or management quality. In contrast, low-cost examples lack any such inefficiencies in construction.
The issue of oversight
On the list of services that are procured, some are more commonly contracted out than others. Construction is as far as I can tell always bid out to private-sector competition, including equipment (nobody makes their own trains), materials, and physical construction. Design and engineering may be contracted out to consultants, depending on the system. Planning never is anywhere I know of, except some unusually high-cost American examples in which public-sector planning has been hollowed out.
The best practice from Southern Europe as well as Scandinavia is that planning and oversight always stay with the public sector. Even with highly privatized contracts, there’s active in-house oversight over the entire process.
The issue of competition
It is necessary to ensure there’s healthy competition between contractors. This requires casting as wide a net as possible. This is easier to do in environments where there is already extensive private- and public-sector construction going on: Turkey builds about 1 million dwellings a year and many bridges, highways, and rail lines, and therefore has a thick domestic market. In Turkish law, it is required that every public megaproject procurement get at least three distinct bids, or else it must be rebid. This rule is generally easy to satisfy in the domestic market.
But if the domestic market is not enough, it is necessary to go elsewhere. This is fine – foreign bidders are common where they are allowed to participate, always with local oversight. Turkish contractors in Northern Europe are increasingly common, following all of the local labor laws, often partnering with a domestic firm.
Old boy networks, in which contractors are required to have a preexisting relationship with the client, are destructive. They lead to groupthink and stagnation. A Turkish contractor held an Android in front of me and, describing work in Sweden, said, “If a Swede said it’s an iPhone, the Swedes would accept that it’s an iPhone, but if I did, they’d check, and see it’s an Android.” In Sweden at least the domestic system is functional, but in a high-cost environment, it is critical to look elsewhere and let foreigners outcompete domestic business.
It is even more important to make sure the experience of bidding on public contracts is positive. A competent contractor has a choice of clients, and a nightmare client will soon lose its business. Such a loss is triple. First, the contractor would have done a good job at an affordable price. Second, the negative experience, such as micromanagement or stalling, is likely to increase costs and reduce the quality of work. And third, the loss of any contractor reduces competition. In the United States, we have repeatedly heard this complaint from contractors and their representatives, that they always have to deal with the “agency factor,” where the agency can be the MTA or any other transit agency, making things difficult and leading to higher bids.
Good client-contractor relations
The issue of avoiding being a nightmare client deserves special scrutiny. It is critical for agencies to make sure to be pleasant clients. This includes any of the following principles:
- Do not change important regulations midway through the project. In Stockholm, the otherwise-good Nya Tunnelbana project has had cost overruns due to new environmental regulations that required disposal of waste rock to the standards of toxic waste, introducing new costs of transportation.
- Avoid difficult change order process (see below for more details on itemization). It should be everyone against the project, not the agency against the contractors or one contractor against another.
- Avoid any weird process or requirements. The RFPs should look like what successful international contractors are used to; this has been a recent problem of American rolling stock procurement, which has excessively long RFPs defining what a train is, rather than the most standard documents used in Europe. This rule is especially important for peripheral markets, such as the entire United States – the contractor knows what they’re doing better than you, so you should adapt to them.
- Require some experience and track record to evaluate a bid, but do not require local experience. A contractor with extensive foreign experience may still be valuable: Israel’s rail electrification went to such a contractor, SEMI, and the results are positive in the sense that the bid was well below expectations and the only problems stem from a nuisance lawsuit launched by a competitor that bid higher and felt entitled to the contract.
- For a complex contract, the best practice here is to have an in-house team score every proposal for technical merit and make that the primary determinant of the final score, not cost. Across most of the low- and medium-cost examples we have looked at, the technical score is 50-70% of the total and cost 30-50%.
- Do not micromanage. New York’s lowest bid rules lead to a thick book of regulations that force the bids to be as similar to one another as possible in quantity and type of goods, to the point of telling contractors what materials they are allowed to use. This is bad practice. Oversight should always be done with flexibility and competent in-house engineers working in conversation with the stakeholders and never with a long checklist of rules.
Contracts should permit as much flexibility as practical, to allow contractors to take advantage of circumstances for everyone’s benefits and get around problems. This is especially important for underground construction and for construction in a constrained city, where geotechnical surprises are inevitable.
Most of the English-speaking world, and some parts of the rest (Copenhagen, to some extent Grand Paris Express) interpret flexibility to mean design-build (DB) contracts, in which the same firm is given a large contract to both design a project and then build it. However, DB is not used in the lowest-cost examples I know of, and rarely in medium-cost ones. If design is contracted out, then there are almost always two contracts, in what is called design-bid-build (DBB). Sources in Sweden say they use single build contracts, but they often use consultants for supplementary engineering and thus they are in practice DBB; Italy is DBB; Turkish sources claim to do design-build, but in reality there are two contracts, one for 60% design and another for going to 100% and then doing construction.
The Turkish system is a good example of how to ensure flexibility. Because the construction contractor is responsible for the finalized (but not most) design, it is possible to make little changes as needed based on market or in-the-ground conditions. In Italy and Spain, the DBB system is traditional, but the building contractor is allowed to propose changes and the in-house oversight team will generally approve them; this is also how the more functional American DBB contracts work, typically for small projects such as individual train stations, which are within the oversight capacity of the existing in-house teams.
DBB can be done inflexibly – that is, wrong – and often when this happens, everyone gets a bitter taste and comes out with the impression that DB is superior. If the building contractor has to go through onerous process to vary from the design, or is excessively incentivized to follow the design to the letter, then problems will occur.
One example of inflexibility comes from Norway. Norwegian construction costs are generally low, but the Fornebu Line’s cost is around $200 million/km, which is not as low as some other Nordic lines. Norway uses DBB, but its liability system incentivizes rigidly adhering to the design: any defect in the construction is deemed to be the designer’s fault if the builder followed the design exactly but the builder’s fault if the builder made any variation. This means small changes do not occur, and then the design consultants engage in defensive design, rather than letting the building contractor see later what risks are likely based on meter-scale geology.
Itemization and change orders
Change orders, and defensive design therefor, are a huge source of cost overruns and acrimony. Moreover, because of the risks involved, any cost overruns are transmitted back into the overall budget – that is, every attempt to clamp down on overruns will just increase absolute costs, as bidders demand more money in risk compensation. California is infamous for the way change orders drive up costs. New York only avoids that by imposing large and growing risk on the bidders (including, recently, a counterproductive blacklisting system called disbarment, a misplaced effort by Andrew Cuomo to rein in cost overruns); the bidders respond by bidding much higher.
Instead of the above morass, contracts should be itemized rather than lump-sum. The costs of materials are determined by the global, national, and local markets, and the contractor has little control over them; in fact, one of the examples an American source gave me of functional change orders in a DBB system is that the bench at a train station can be made of wood, metal, or another material, depending on what costs the least when physical construction happens.
Labor costs depend on large-scale factors as well, including market conditions and union agreements. The use of union labor ensures that the wages and benefits of the workers are known in advance and therefore unit costs can be written into the contract easily. Spain essentially turns contracts into cost-plus: costs depend on items as bid and as required by inevitable changes, and there is a fixed profit rate based on a large amount of competition between different construction firms.
The upshot is that itemized costs prevent the need to individually negotiate changes. If difficult ground conditions or unexpected utilities slow down the work, the wages of the workers during the longer construction period are already known. It is especially important to avoid litigation and the threat thereof – questions of engineering should be resolved by engineers, not lawyers.
Here, our results, based on qualitative interviews with industry experts, mirror some quantitative work in economics, including Ryan and Bolotnyy-Vasserman. Itemization reduces risk because it pre-decides some of the disputes that may arise, and therefore the required profit rate to break even net of risk is lower, reducing overall cost.
The impact of bad procurement practices
One of our sources told us that procurement problems add up to a factor of 2 increase in New York construction costs. Five specific problems of roughly equal magnitude were identified:
- A regulation for minority- and woman-owned businesses (MWB), which none of the pre-qualified contractors in the old boy network is.
- The MTA factor.
- Change order risk.
- Disbarment risk.
- Profit in a low-competition environment.
MWB and disbarment are New York-specific, but the other three appear US-wide. In California, the change order risk is if anything worse, judging by routine cost overruns coming from change orders. California, moreover, is very rigid whenever a contractor suggests design improvements, as Dragados did for its share of California High-Speed Rail, even while giving contracts to contractors that engage in nuisance change orders like Tutor Perini.
Aligning American procurement practice with best practice is therefore likely to halve construction costs across the board, and substantially reduce equipment costs due to better competition and easier contractor-client relations.
I want to go back to the problem of early commitment as I explained it two months ago. It comes out of research done by Chantal Cantarelli and Bert van Wee about Dutch cost overruns, but the theory is more generally applicable and once I heard about it I started seeing it in play elsewhere. The short version is that politically committing to a megaproject too early leads to lock in, which leads to compromised designs and higher costs. The solution, then, is to defer commitment and keep alternatives open as much as possible.
The theory of lock in
The papers to read about it are Cantarelli-Flyvbjeerg-Molin-van Wee (2010), and Cantarelli-Oglethorpe-van Wee (2021). Both make the point that when the decision to build is undertaken, it imposes psychological constraints on the planners. They are not long or difficult papers to read and I recommend people read them in full and perhaps think of examples from their own non-Dutch experience – this problem is broader than just the Netherlands.
For example, take this, from the 2010 paper:
Decision-makers show evidence of entrapment whenever they escalate their commitment to ineffective policies, products, services or strategies in order to justify previous allocations of resources to those objectives (Brockner et al, 1986). Escalating commitment and justification are therefore important indicators of lock-in. The need for justification is derived from the theories of self-justification and the theory of dissonance which describe how individuals search for confirmation of their rational behaviour (Staw, 1981; Wilson and Zhang, 1997). This need arises due to social pressures and “face-saving” mechanisms. The involvement of interest groups and organizational pushes and pulls can also introduce pressures into the decision-making process, threatening the position of the decision-makers, who may feel pressure to continue with a (failing) project in order to avoid publicly admitting what they may see as a personal failure (McElhinney, 2005). “People try to rationalize their actions or psychologically defend themselves against an apparent error in judgment” (Whyte, 1986) (“face-saving”). When the support for the decision is sustained despite contradicting information and social pressures, the argumentation for a decision is based on the need for justification.
The focus on face-saving behavior leading to escalation is not unique to the literature on transportation. In international relations, it is called audience cost and refers to the domestic backlash a political leader suffers in case they back down from a confrontation they were involved in earlier; this way, small escalations turn into bigger ones and eventually to war, or perhaps to a forever occupation.
There are a number of consequences of lock in:
- Projects will follow designs set long ago, especially ones that were hotly contentious. For example, California High-Speed Rail has stuck with the decision to build its alignments via Palmdale and Pacheco Pass, since the possibilities of changing Palmdale to the Grapevine/Tejon alignment and Pacheco to Altamont Pass both loomed large (there was a NIMBY lawsuit trying to force a change to Altamont). However, at the same time, there are plans to potentially run the partially-built system without electrification, since that issue was never in contention and is not part o the audience cost.
- There are unlikely to be formal cancellations. California is again a good example: high-speed rail lives as a hulk, not formally canceled even when the governor said of the idea to complete it, back during the Trump administration, “let’s be real,” defending the initial construction segment between Bakersfield and Fresno as valuable in itself. Formal cancellation is embarrassing; a forever construction project is less visible a failure.
- Prioritization is warped to tie into real or imagined connections with the already-decided project. California is not as clear an example of this as of the other two points, but in New York, once the real (if not yet formal) decision to go forward with Second Avenue Subway was made in the 1990s, the Regional Plan Association tied in every proposed expansion plan to that one line.
Cantarelli-van Wee treat early commitment as a problem of bad planners, who become psychologically wedded to potentially incorrect solutions. However, it is instructive to shift the locus of moral blame to surplus extraction by political actors, such a local politicians, power brokers, and NIMBYs.
In the story of HSL Zuid, much of the extra cost should be blamed on excessive tunneling. In the flat terrain of Holland and near-coastal Brabant, no tunneling should have been needed. And yet, the line is 20% underground, partly to serve Schiphol, partly to avoid taking any farmland in the Groene Hart. The Groene Hart tunneling has to be understood in context of rural NIMBYism (since at-grade solutions to habitat loss exist in France).
In this formulation, the problem with lock in is not just at the level of planners (though they share most of the blame in California). It’s at the level of small actors demanding changes for selfish reasons, knowing that the macro decision has already been made and the stat cannot easily walk away from the project if costs rise. These selfish actors can be NIMBY, but they can equally be local power brokers wanting a local amenity like a detour to serve them or a station without commercial justification. In Germany, an extra layer of NIMBYism (albeit not on connected with lock in – we have late commitment here) is demands to include freight on high-speed lines, in order to take it off legacy lines, which design forces gratuitous tunneling on high-speed lines in order to moderate the grade.
California is a good example of a non-NIMBY version of this. The state politically committed to building high-sped rail in the 2008 election, for which it showed clear maps of the trains detouring via Palmdale and going to San Francisco via Pacheco Pass. By the time further environmental design showed that the Los Angeles-Palmdale route would require tens of km more tunneling through Soledad Canyon than anticipated to avoid impact to an ecologically sensitive area, the state had already pitched Palmdale as a key high-speed commuter suburb, and Los Angeles County made housing plans accordingly. The county subsequently kept agitating for retaining Palmdale even as other alignment changes in the area were made, turning Palmdale into its pet project.
The planning literature undertheorizes and understudies problems arising from localism. In conversations with people in the European core as well as the United States, there’s an unspoken assumption that the community is good and the state is bad. If the community demands something, it must represent correction of a real negative externality, rather than antisocial behavior on behalf of self-appointed community leaders who the state can and should ignore. It doesn’t help that the part of Europe with the least community input is the Mediterranean countries, which Northern European planners look down on, believing any success there must be the result of statistical fudging.
The solution: late commitment
To reduce costs and improve projects, it’s best to delay political commitment as late as possible. This means designing uncertain projects and only making the decision to build at advanced stages of design – maybe not 100% but close enough that major revisions are not likely. The American situation in which there is no regular design budget so agencies rely on federal funding for the design of the projects they use the same federal funding for leads to bad outcomes over and over. California, which went to referendum without completing the environmental design first, takes the cake.
Late commitment is thankfully common in low- and medium-cost countries. Germany does not commit to high-speed rail lines early, and, judging by Berlin’s uncertainty over which U-Bahn extensions to even build, it doesn’t commit to subways early either. Sweden is investigating the feasibility of high-speed rail but rail planners who I talk to there make it clear that it’s not guaranteed to happen and much depends on politics and changes in economic behavior; overall, Nordic infrastructure projects are developed by the civil service beyond the concept stage and only presented for political negotiation and approval well into the process. Southern European planners com up with their own extension programs and politically commit close to the beginning of construction.
I was on a panel at Eno’s symposium on costs, talking with other teams investigating comparative construction costs. We worked off a list of questions Eno’s Robert Puentes had sent us before, knowing that the list was too long for five people (me, Eric, Laura Tolkoff, Ethan Elkind, Romic Aevaz) to cover in an hour. So for more completeness, here are my responses – and pay attention specifically to issues of scope and what we should be doing in the future. In particular, as we’re getting funded to do other things, we will likely have room in the budget to add a few more cases, and hire people who can put them together.
What were your key takeaways on the extent of our cost premium, and key cost drivers?
I blogged this just before the panel. The only major headers I’ll add are poor interagency coordination in the United States, especially for projects that are or touch commuter rail, and a political system full of real and imagined veto points. The imagined veto points are not unique to the US – the UK, Germany, and the Netherlands all have visible problems with excessive tunneling on high-speed rail projects coming from NIMBY demands, NIMBY demands that at least in the first two cases are paper tigers that the state can ignore if it doesn’t mind a few news cycles with negative headlines.
Questions on scope
There were three separate questions on this, since our approaches differ – Eno has more cases covered in less depth (and we made sure to pick disjoint comparison cases from theirs), Berkeley focuses on California projects. So we went through questions about what our respective scopes and limitations are:
- Could you walk us through the general scope and bounds of your work?
- What were some of the limitations you ran into when collecting information on costs/timeline, and what recommendations would you have to improve data reporting for projects?
- What are some of the lingering questions or areas for future study that your teams have flagged?
The answer to all three is that our scope – the six cases – looks at specific issues rather than general ones. The forest comprises trees and cannot be studied as an ecosystem until one understands the biology of the tree species therein. But then, understanding the biology of the tree species requires understanding the ecosystem they have evolved in; the reason we do cases simultaneously is that hearing about issues arising in one place informs our work on other places.
That said, I think it matters that none of our six cases is typical. Medium-cost environments like France, Germany, and Japan are unfortunately not in scope; I’ve read a lot of work on cot issues plaguing Grand Paris Express, but unfortunately not in any global or even just European comparative sense. All of our cases are Western (for infrastructure purposes Turkey is a Western country); this matters because, while European and East/Southeast Asian costs are broadly the same, both covering the entire global range short of American costs, there are notable differences in how they build, so it’s plausible that there re things one side does right that the other doesn’t in both directions. All of our cases are first-world or, in Istanbul’s case, 1.5th-world.
This means that we would like to add cases. Attractive targets include anything in Spain, to beef up our set of low-cost examples, and then cases that represent examples we didn’t study, that is places that are medium-cost, non-Western, or not in or in the penumbra of the developed world. My suspicion is that medium-cost examples will interpolate practices – Germany and France both vaguely appear to mix good Scandinavian or Southern European behavior with bad British and American behavior, each in its own way. But I do not know and that’s why we’d like to add cases. In middle-income countries like Russia, Mexico, Brazil, and China, and in low-income ones like India or the Philippines, I do not really know what to expect and my only explanation so far is completely different from any first-world pattern.
We should have a budget for this, but I don’t yet know how many cases we can juggle in addition to where we’re going to shift the main of our attention starting in early 2022, that is high-speed rail and a synthesis for the Northeast Corridor. Most likely other people will write the cases (for pay of course) and we will supervise in between looking at the history and technical data of the Northeast Corridor.
I’m in the middle of an online symposium at Eno about construction costs; I talked on Tuesday and I think there will be a recording made available later. The conference is good by a lot of standards, including the “do they tell me things I don’t know about costs” standard. But for now, I want to address one point made repeatedly in interviews re some low-cost cities: the argument from consensus. It’s wrong, and leads to very wrong conclusions.
What is the argument from consensus?
In Eno’s lowest-cost comparison cities, like Madrid, there’s political consensus in favor of building more subways. Repeated panels gave this example of how in the 1990s and 2000s, PP and PSOE both supported subway construction and promised to build more in their election campaigns for the Community of Madrid. PSOE in fact attacked PP saying its proposals were unrealistically ambitious, but then the Madrid Metro expansion opened as planned. In such a political environment, no wonder planners had leeway to build the system without much interference.
Why is this wrong?
Britain has bipartisan political consensus in favor of both Crossrail and High Speed 2, both of which are explicitly supported in both parties’ manifestos. Its construction costs are still Europe’s highest.
More to the point, the United States is not uniformly an environment where public transportation is a partisan flashpoint, because in most of the cities that build subways, the Republican Party does not exist. The closest thing to a Republican Party in New York is the Manhattan Institute, which criticizes unions incessantly but does not call for ending public support for the subway and refrains from making the anti-transit arguments made by national Republican outfits like Reason and Cato. The last Republican mayor, Rudy Giuliani [sic], tried to expand the subway to LaGuardia. I’ve met New Yorkers who view Giuliani as a savior and who are ant-immigration climate denialists and thy too think the city and state should make the subway better – if anything they treat its poor state as evidence that the Democrats can’t govern.
This consensus does not lead to low costs. Why would it? There is no respect for planners or engineers. The consensus in New York means every governor installs their political cronies at the head of the agencies involved. None of the mechanisms that make Madrid Metro work is present.
Just having more political support for subway construction is not going to by itself make things better. American states where investment is safe from cancellation do not do better than ones where one election could spell doom for the investment program. Sea changes are required, not just more public support.
In three cities that I know of, there are plans to deal with an incipient regional rail capacity crunch by building a new tunnel: Tel Aviv, Hamburg, London. The route in question in all cities already has regional rail service making frequent urban stops as well as longer-distance intercity trains. Setting Tel Aviv aside – new tracks are not necessary there at all – both Hamburg and London have a choice of what to build in the tunnel. In both cases, the answer must be intercity rail and not regional rail. This affects Crossrail 2 in London, currently shelved but still in active planning, as well as plans for Hamburg Hauptbahnhof-Altona capacity improvements.
The dominant factor in the cost of an expensive urban railway in a constrained environment is the stations. Low-cost countries build very cheap stations, but that’s true in outlying areas, urban as they may be, and not in city center areas under and around older subways. What’s more, Britain and Germany are not low-cost countries. German costs are somewhat higher than the global median, British costs among the world’s highest. Thus, keeping down station costs is paramount – and express tunnels have fewer stations than local tunnels.
Normally, the express vs. local issue is not relevant to a new urban rail line. Yes, more stations are more expensive, but on a line designed to open up service to a new area, more stations also provide more access, so the extra cost is often worth it. This is true even for urban subways that act as relief lines, like Second Avenue Subway, a relief line for the Lexington Line: more stations provide better local access and therefore increase the line’s relief value.
However, when the problem comes from regional or longer-distance capacity, all of this goes out the door. Crossrail 2 includes a long tunnel from Central London all the way to Wimbledon not because of purely local needs but because of very high rail usage along the South West Main Line. In Hamburg the problem is similarly about the main line between Hauptbahnhof and Altona – local traffic is saturated on the S-Bahn, and all other trains have to squeeze on the remaining two tracks of the Verbindungsbahn. Thus, capacity expansion should involve a tunnel with the fewest number of stations, on the most express services.
And yet both cities are doing it wrong. Hamburg is planning an S-Bahn tunnel, with the existing S-Bahn route then given over to regional trains, to be segregated away from the intercity trains. But there are already two Hauptbahnhof-Altona S-Bahn routes – that’s not where the service need is. Instead, new tunnels should go between the stations without stopping, to reduce costs, hosting intercity trains while regional trains take over the existing intercity tracks.
London is likewise planning on an undulating connection between Clapham Junction and Wimbledon, with links to other parallel north-south lines in the area. This is not good planning – those new stations are inordinately expensive and not needed for network connectivity. If there is no way to six-track the South West Main Line above-ground by replacing the sloped berm with retaining walls, it’s the fastest trains that should go underground, to save money on stations. Crossrail 2 is a £31.2 billion project; I don’t think Paris has spent this money on all Métro and RER lines in the region to date combined, and Grand Paris Express, at a broadly similar cost, includes 160 km of tunnel. It’s necessary to economize and build the tunnels that are necessary, and not the ones London would like to have.
I did a poll on Patreon about cost issues to write about. This is a close second, with 11 votes; other people’s money won with 12, whereas neighborhood empowerment got 8 and will not be on my docket.
There are infrastructure investment programs defined around the specific projects funded: Crossrail, Second Avenue Subway, Grand Paris Express, Nya Tunnelbanan, the Toronto RER, Marmaray. Then there are programs defined around costs, where one constantly hears the project defined by its budget rather than what it produces; these are all in the United States, and include the entire slates of Los Angeles and Seattle rail extensions, and to some extent also California High-Speed Rail and Gateway. The latter appears like bad practice for cost minimization.
What’s the problem?
In isolation, I’d have expected cost-centric plans to be more prudent with the budget – there’s less room for overruns. This is related to Swiss practice, which is project-centric but also requires projects to go to referendum on the precise amount, which has disciplined cost overruns in most cases and also kept absolute costs low. However, the fact of the matter is that the only places that use cost-centric plans have high costs, having recently risen from levels that were not so bad 20 years ago. So why?
My suspicion is leakage. This is getting to be less general and more specific to the situation of California and its sales tax measures, but the way it works is, there is an amount that proponents think they can go to ballot on, and then they work the slate of projects backward. In theory, this is supposed to discipline the planners into better behavior: the amount of money is truly fixed, and if costs go up, it delays the entire program. In practice, there is no prior discipline about what infrastructure should be included, and thus the slate is decided politically on a place-based plan.
Further leakage occurs when buying off additional interest groups. Soon enough, one useful if very expensive subway line, like the Purple Line Extension in Los Angeles or the Ballard-West Seattle LRT, is bundled into a huge program alongside bus operating subsidies, road money, and low-usage lines to lower-density areas.
I can’t prove that this is the result of budget-centric planning. The comparison examples I have – all high-cost, politicized North American projects – exhibit leakage as well, but less of it. The Green Line Extension in Boston had extensive local leakage in the first iteration of the project, like the Somerville Community Path, but it wasn’t paired with less useful infrastructure elsewhere. Second Avenue Subway Phase 1 was paired with East Side Access and the Broadway subway in Vancouver is paired with a SkyTrain extension deeper into Surrey toward Langley; in both cases, the less useful projects are nonetheless more useful on a likely cost per rider basis than any of the American West Coast leakage and compete with the more useful projects. ESA is probably going to end up $60,000/rider, not much worse than GLX and probably about the same as the Purple Line Extension depending on how much transit-oriented development Los Angeles permits.
Place-based politics is a scourge and should be eradicated whenever possible. What it does wherever it is not suppressed is create political identification among local and regional power brokers not with the piece of infrastructure but its cost. The reason is that evaluating transportation needs is too technocratic for the attention span of a local politician, whereas the budget is a straightforward measure of one’s importance.
Once local actors are empowered, they make further demands for irrelevant extras (“betterments”), or construction techniques that spend too much money to avoid real or imagined negative local impact. People with a local identity don’t care about public transit much – public transit takes riders to other localities, especially city center, whereas the locally-empowered minority of people who work locally has little use for it and drives everywhere.
Local empowerment is not unique to budget-based infrastructure. It was a major drag on GLX and at least a moderate one on SAS Phase 1, neither of which is budget-based. The Central Subway and BART to San Jose projects are both place-based vanity, for Chinatown and San Jose respectively, but even these projects are smaller in scope than the Los Angeles or Seattle ST3 leakage. There’s just more surface area for it when advocates lead with a budget, because then every local hack sees an opportunity to make a claim.
The place-based politics in the Northeast is much broader and more regional: SAS, a city project championed by then-Assembly speaker Sheldon Silver (D-Lower East Side), was balanced with ESA, a suburban project serving the base of Governor George Pataki (R-Peekskill, but the state Republicans were based on Long Island). Metro Vancouver’s place-based extraction follows the same schema: if Vancouver gets a useful SkyTrain extension, Surrey must get an extension too regardless of usefulness. Massachusetts is likely to be in a similar situation with the TransitMatters Regional Rail program: RR serves the entire east of the state, and must be balanced with a Western Massachusetts project, for which we propose the still-useful East-West Rail program connecting Boston and Springfield.
In contrast, the situation in California and metropolitan Seattle is much worse – useful lines in Los Angeles are paired with many layers of leakage, as different groups make claims on the pot of money. This way, Los Angeles doesn’t build as much useful transit as New York and Boston even though its construction costs are comparable to Boston’s and much lower than those of New York, and even though it makes large amounts of money available for transportation by referendum.
Are jobs a cost or benefit?
Like place-based extraction, the use of infrastructure as a jobs program is terrible everywhere in a modern developed country where construction is not a labor-intensive zero-skill job, and should be eradicated. And like place-based extraction, I think – and am less certain than on the other points – there is more surface area for this when the program is about a budget and not a piece of infrastructure.
The mechanism is the same as before: once money becomes available, local labor groups descend on it to make claims. Promises of job creation are thus always local, including beggar-thy-neighboring-state demands for local rolling stock construction. These occur for both budget-based plans (like the Los Angeles light rail fleet) and project-based ones (like the new Red and Orange Line cars for Boston, built in Springfield due to place-based extraction). However, it’s easier to make a claim when the political discussion is about how to spend $X and not how to optimally produce a desired piece of infrastructure.
The way forward
The American West Coast’s problem of budget-based planning is, thankfully, easy to solve, because it’s been solved in other parts of the same country. The Bay Area has less of it than Southern California and the Pacific Northwest (but it’s not free of it – many of the specifics of California High-Speed Rail’s failure come from Bay Area power brokers hoping to use it as a slush fund). The Northeast doesn’t have it at all. Los Angeles is likely to be forced in that direction anyway, because it’s running out of sales tax capacity – the already-approved measures are spoken for through the 2050s.
The impact is likely not a matter of straight construction costs in dollars per kilometer. Rather, it’s about leakage. Los Angeles and Seattle do not have unusually higher per-km costs by American standards; in the 2000s Los Angeles looked like the good part of America and in the 2010s Seattle did, but since both have converged to much higher figures. The problem is that a smaller share of the Los Angeles Measures R and M spending goes to useful expansion than the capital budget in places that have project-based planning. This is what needs to be fixed through transitioning to project-based planning, costs aside.
I have found something a European city does worse than the United States in public transit. Paris has just announced its new bilevel design for the RER B, currently the only line running single-deck trains due to restricted clearances. The new double-deckers, dubbed MI 20, are expected to cost 2.56b€ for 146 trainsets, each 104 meters long, for a total of 168,600€ per meter of train length.
I’ve criticized Paris’s use of double-deckers in the past. The cost premium for a double-decker, usually around 25-50%, at best matches the gain in seated capacity, and leads to other capacity problems with access and egress, which are of especial importance on urban rail like the RER. Not for nothing, bilevel trains are not used in Tokyo except for the occasional first-class car (“green car”), which is less crowded by design than the legendarily crowded subway and regional rail cars.
However, this is a lot worse than the usual premium. The only comparably expensive bilevel I can find is the Stadler KISS order for Caltrain, which at $230,000/m for the base order (and only $160,000/m for an equal-size option) comes at a large premium over usual KISSes (both around 130,000€/m) due to client interference and micromanagement coming from low competence by American railroaders.
But the KISS is a high-performance train, at the expensive end in Europe, too. Moreover, it is fully bilevel, whereas the MI 20 has a mix of single- and double-deck cars, with high-platform boarding. Comparable split-level trains go well below 130,000€/m. Canalblog has a compendium of recent Coradias: the single-level example for Milan is 6.25m€ per 84-meter train, or 74,400€/m, and the mixed single- and double-deck examples are 96,700€/m in Luxembourg and 117,600€/m in Germany. The mixed-deck Siemens Desiro HC has a range of costs: its RRX order is 1.7b€ for 82 150-meter trainsets, which is 138,200€/m, but a smaller order for the Berlin RegionalBahn is 300m€ for 21 six-car and 2 four-car trainsets, or 89,600€/m, which is a high but not unheard of cost for a single-decker, let alone a double-decker. The Desiro HC is being delivered to Israel as well, at a cost of 900m€ for 60 trainsets totaling 330 cars, or 109,100€/m.
There’s nothing special about Paris that justifies such a cost – the highest in the world so far, even beating the Americans. Rather, the problem is most likely that Paris thinks it’s special and won’t buy a standard platform. Canalblog points out that the Coradia Duplex formed the basis of the X’Trapolis, currently delivered for the RER D and E – and the X’Trapolis’s first tranche spent 29% of its budget on design and engineering, driving the cost up to a stratospheric 21.83m€ per train of length 112 or 130 meters. Even averaged over the entire order of 255 trainsets, at which point economies of scale kick in and the bespoke design is less harmful, the cost is 121,400€/m, which is 25% more than the more standard Luxembourg design.
Update: Clem Tillier asked me about Madrid’s recent Cercanías order. This is a mix of Stadler trains and Coradias, both mixed single- and double-deck; the Coradias, using the same platform as the Paris X’Trapolis trains and built in Spain rather than in France, cost 1.447b€/152 trainsets, or 95,200€/m, and the Stadlers, mixing KISS and FLIRT technology, cost 998m€ for 24 100-meter trains and 35 200-meter trains, or 106,200€/m.
In a megacity like Paris, it’s tempting to think one is special and must have special equipment. But the resulting high costs are particularly damaging in such a city. The RER B runs every 3 minutes at rush hour, which means that high rolling stock costs are proportionally a bigger problem than on a less frequent system. The cost premium of the order over standard single-deck trains is a factor of around 2; half the cost is 1.3b€, which would be enough to build some necessary tunnel extensions, like quad-tracking the combined two-track tunnel for the RER B and D between Châtelet-Les Halles and Gare du Nord, or if RER investment is not desired then around 6 km of tunnel for Grand Paris Express after the latest cost overruns.
France needs to let go of its pride and recognize that Paris is merely the largest city in the Union, with the same standards and regulations as the other 440 million of us who do not live in Ile-de-France. Vanilla Coradias and Desiros that work elsewhere should also work for the RER, with minor tweaks to take into account high platforms and the loading gauge, both of which the vendors are experienced in dealing with due to common intra-European variation. The people who sign extravagant contracts may feel special about the train design, but the passengers who end up not getting the investment the cost premium would have gone to are going to keep feeling packed on rush hour RERs. The region ought to do better and hire managers who are better than this.
I did a poll on Patreon about cost issues to write about. This is the winning option, with 12 votes; project- vs. budget-driven plans came second with 11 and I will blog about it soon, whereas neighborhood empowerment got 8.
OPM, or other people’s money, is a big impediment to cost reform. In this context, OPM refers to any external infusion of money, typically from a higher-level government from that controlling an agency. Any municipal or otherwise local agency, not able or willing to raise local taxes to fund itself, will look for external grants, for example in a federal budget. The situation then is that the federal grantor gives money but isn’t involved in the design of where the money goes to, leading to high costs.
OPM at ground level
Local and regional advocates love OPM. Whenever they want something, OPM lets them have it without thinking in terms of tradeoffs. Want a new piece of infrastructure, including everything the local community groups want, with labor-intensive methods that also pay the wages the unions hop for? OPM is for you.
This was a big problem for the Green Line Extension’s first iteration. Somerville made ridiculous demands for signature stations and even a bike path (“Somerville Community Path”) thrown in – and all of these weren’t jut extra scope but also especially expensive, since the funding came from elsewhere. The Community Path, a 3 km bike path, was budgeted at $100 million. The common refrain on this is “we don’t care, it’s federally funded.” Once there’s an outside infusion of money, there is no incentive to spend it prudently.
OPM modifying projects
In capital construction, OPM can furthermore lead to worse projects, designed to maximize OPM rather than benefits. Thus, not only are costs high, but also the results are deficient. In my experience talking to New Englanders, this takes the form of trying to vaguely connect to a politician’s set of petty priorities. If a politician wants something, the groups will try pitching a plan that is related to that something as a sales pitch. The system thus encourages advocates and local agencies to invest in buying politicians rather than in providing good service.
This kind of behavior can persist past the petty politician’s shelf life. To argue their cases, advocates sometimes claim that their pet project is a necessary component of the petty politician’s own priority. Then the petty politician leaves and is replaced by another, but by now, the two projects have been wedded in the public discourse, and woe betide any advocate or civil servant who suggests separating them. With a succession of petty politicians, each expressing interest in something else, an entire ecosystem of extras can develop, compromising design at every step while also raising costs.
The issue of efficiency
In the 1960s, the Toronto Transit Commission backed keeping a law requiring it to fund its operations out of fares. The reason was fear of surplus extraction: if it could receive subsidies, workers could use this as an excuse to demand higher wages and employment levels, and thus the subsidy would not go to more service. As it is, by 1971 this was untenable and the TTC started getting subsidies anyway, as rising market wages required it to keep up.
In New York, the outcome of the cycle of more subsidies and less efficiency is clearer. Kyle Kirschling’s thesis points out on PDF-p. 106 that New York City Transit’s predecessors, the IRT and BMT, had higher productivity measured in revenue car-km per employee in the 1930s than the subway has today. The system’s productivity fell from the late 1930s to 1980, and has risen since 1980 but (as of 2010) not yet to the 1930s peak. The city is one of a handful where subway trains have conductors; maintenance productivity is very low as well.
Instead of demanding efficiency, American transit advocates tend to demand even more OPM. Federal funding only goes to capital construction, not operations – but the people who run advocacy organizations today keep calling for federal funding to operations, indifferent to the impact OPM would have on any effort to increase efficiency and make organizations leaner. A well-meaning but harmful bill to break this dam has been proposed in the Senate; it should be withdrawn as soon as possible.
The difference between nudging and planning
I am soon going to go over this in more details, but, in brief, the disconnect between funding and oversight is not a universal feature of state funding of local priorities. In all unitary states we’ve investigated, there is state funding, and in Sweden it’s normal to mix state, county, and municipal funding. In that way, the US is not unique, despite its federal system (which at any case has far more federal involvement in transportation than Canada has).
Where the US is unique is that the Washington political establishment doesn’t really view itself as doing concrete planning. It instead opts for government by nudge. A federal agency makes some metrics, knowing that local and state bodies will game them, creating a competition for who can game the other side better. Active planning is shunned – the idea that the FTA should have engineers who can help design subways for New York is unthinkable. Federal plans for high-speed rail are created by hiring an external consultant to cobble together local demands rather than the publicly-driven top-down planning necessary for rail.
The same political advocates who want more money and care little for technical details also care little for oversight. They say “regulations are needed” or “we’ll come up with standards,” but never point to anything concrete: “money for bus shelter,” “money for subway accessibility,” “money for subway automation,” etc. Instead, in this mentality the role of federal funding is to be an open tab, in which every leakage and every abnormal cost is justified because it employed inherently-moral $80,000/year tradesmen or build something that organized groups of third-generation homeowners in an expensive city want. The politics is the project.