In the two months since The Transit Brief launched, and from the very first post, the issue of massive escalations in the cost of building transit— and specifically subways — has been central.

The issue has recently gained significant salience (we take some credit), particularly in Ontario as the cost of the Ontario Line subway — a great project in many ways — ticked over triple its initial, already high projected cost to more than $30 billion CAD.

This led to a somewhat promising media storm. Finally, the message is getting through: a Globe and Mail piece talked about the absurdity of going over budget by so much; the Toronto Star tried to elucidate the cause of the escalation by talking to some local researchers; The Transit Brief’s own Reece Martin wrote about the avoidance of the problem — ironically as it might have finally broken through; and then just yesterday the Globe and Mail wrote on the topic again, urging policymakers to fix the problem — we fully endorse this.

But, the question we have received time and again, and which must be properly addressed if we are to create a fix is … what exactly is causing this problem? In this post, which will be in-depth, we attempt to answer — ultimately highlighting at least 52 factors that contribute to high costs.

The Problem

The problem if we are to define it is fairly simple. Transit projects — from subways to bus rapid transit — in Canada, and in the English-speaking world writ large, cost far more than projects in many other developed and developing countries, even when adjusting for inflation and differences in the relative value of currencies and inputs. This is, crucially, despite the fact that most transport projects are built to functionally equivalent fire safety standards and similar accessibility standards — even in developing nations.

Some politicians, including the Premier of Ontario, have feigned ignorance by pretending that the upset is regarding inflation, when in reality the cost of building transit has escalated dramatically after inflation adjustment.

The importance of the problem goes deeper. Other large projects often face similar issues, from energy and road infrastructure to defence, and often for similar reasons. Despite being rarely discussed, many of these problems of public projects have private-sector parallels, with major delays to big software releases or bugs growing increasingly common as one example.

Time is often interchangeable with money, and increasingly these projects are not only expensive, but also slow. As we will see, there is a complex interplay between project timelines and cost.

Zooming in on transit, high costs have — such as in Toronto — led to a lowering of ambitions; the “Transit City” light rail plan was in no small part a response to increasing costs for subways.

Unfortunately, cost increases have not been limited to subways. That means that by the time the central projects of “Transit City” were being built, those cut-down suburban streetcars were clocking in at the same prices as the subways that had motivated them.

The cost of high costs is real. In the United States, only New York is currently building a new subway through its city centre (if northern Manhattan can be called the city centre), new subways are rarely constructed, and even new light rail lines are few and far between — Austin massively scaled back one such plan recently. The shocking contrast with China’s explosive growth, but even with expansion in places like France and Korea — countries without enormous growing urban populations and wealth — highlights the problem.

In much of the rest of the English-speaking world transit is more highly valued, and so enormous projects the scope of which are not seen in America have happened across many cities. But, this has been a “powering through” and huge sums of money have been invested.

We recently looked at how the opportunity cost here might play out in Montreal. By our estimates, if the new Blue Line extension program there cost something closer to what the most efficient developed countries pay, Montreal could have still built it, but it could have also built over a dozen other transformative urban infrastructure projects — hospitals, highway removals, street redesigns, and new university campuses — in the same budget envelope.

These higher costs also mean worse transit. The public shows significant interest in subway washrooms, which are common in Asia, but these are costly. If subways cost far less to build, such features would not land as “yet another ask” of stretched projects.

In most cases, this also means less is built than might otherwise be. Toronto is getting a much-needed and very exciting new city centre subway with the Ontario Line and some extensions, but Santiago, Chile, a similarly sized, but much less affluent city, is building four new major subway lines and its own extensions.

And we have more to say. An upcoming post collaborating with JRUrbaneNetwork (a top commentator on public infrastructure in Asia and China specifically) will look at a new measure that more accurately considers the value of building transit that may make the price of Toronto’s Ontario Line look even higher.

The post will consider the capacity of transit actually delivered for the dollars spent, since much like with construction time, high costs have bled across into other domains leading to reduced scope. Furthermore, it will discuss the increase in project costs with respect to stable, and sometimes falling, project complexity.

So why is it expensive?

On the matter of why transit costs so much, there are three important takeaways. For one, there is rarely any one reason; for two, applicable reasons vary from city to city; and for three, research is ongoing.

It may not be satisfying to hear, but there is not one thing we could fix that would suddenly allow New York, Toronto, and London to pump out new subways like the low-cost champion cities of Madrid, Shenzhen, or Istanbul. This also means that those claiming the problem is just consultants, or P3s (public-private partnerships or design-build projects), should be viewed with extreme scepticism.

A downstream problem is that reporters, including in the Star reporting above, often will cite a few specific issues mentioned by experts interviewed — who are either ignored or fail to point out that there are a plethora of problems — and then make it sound as if those issues are the issues — as in the only or primary ones. For example, not building for extended periods, or poor transparency.

This can mask the true nature of the problem, because with sufficient expertise and time one can enumerate endless exceptions. For example, Vancouver has built continuously and has still seen consistent cost escalation, while Montreal has seen cost escalation despite not engaging in large numbers of P3s — in fact, the REM, a rare bright spot in the Canadian transit costs landscape, closely resembles a P3.

The most accurate way to characterize the problem, then, is that it is a grab bag of problems, many shared across many cities, some less common, that contribute to high prices.

We mentioned research above, and the best way to understand what is driving costs is to look at the numerous factors that research has identified as contributing, as well as additional ones that a wide range of observers, professionals, and advocates have identified.

Probably the best singular resource for information on transit costs is the aptly-named Transit Costs Project at New York University, which has both a large database of projects and their prices, and a series of case studies that look at what went wrong and right in different cities. Other useful resources — particularly in the Canadian context — include the RCCAO report on subway costs in Toronto, a comparative study released last year from the University of Toronto’s School of Cities, and a recent report released by the Canadian Standards Association. In the next section, we synthesize the findings of all of these resources into a common atlas of cost-driving factors and relations. It should be noted that not all factors enumerated in the literature are included, as we focused on relevance to Canada in particular. We have also added significant context where we think it adds color to a particular cost driver.

The Factors

Perhaps the most important factor for high costs in our view is not really a direct factor at all. Instead, it is the recognition, clearly visually outlined in the Transit Costs Project final report, that the problem leading to 5 or 10 times more expensive infrastructure in Toronto or New York is not generally that any one thing is being done ten times worse, but that factors are multiplicative — they compound.

For example, imagine workers in Toronto are paid twice what workers in Seoul are, and then imagine the subway has stations twice as large as is needed, and then imagine that the procurement process leads to overheads that double the price paid for these things. Nothing is in isolation costing five times as much, but since the workers being paid twice as much build something twice as big, and the process of organizing this costs twice as much as well, you get these factors multiplying out to 8 times. The visualization below beautifully illustrates how this looks for New York City.

It is beyond the scope of this post to try to determine the exact numerical impact of these factors for any given project(s), though the logic discussed above can be helpful for considering how factors interact. It is also worth considering how time and risk also interrelate. Some cost increases are not multiplicative!

Factors also have not appeared all at once. Many have diffused as different jurisdictions learn from one another, or experience similar cultural changes.

Another important quip (made off the cuff and yet deeply insightful!) by Matt Yglesias (of Slow Boring and Vox fame) at one of the Transit Costs Project events, that your correspondent (who has worked on at least a few major transit projects) can validate — is that if there was a single factor that could be said to be causing everything, it is that costs just don’t seem to be a priority.

This is reflected by the indifference to the issue that is so widely seen, as well as its historically low salience. But it’s also perhaps more directly related to the fact that while people will make loud noises and protest about the impacts of a nearby train line, or demand additional community benefits on top of new infrastructure and its spinoffs, (almost) nobody will yell at a policymaker for greenlighting even the most expensive subway ever built.

On the inside of projects — at least in Canada — this manifests as just little mind being paid to the cost implications of decisions, and of processes of the project itself. This is reflected by the fact that soft costs — that is, not heavy construction, but planning and engineering — now sometimes cost as much as the entire project might in a place like Italy or Chile.

Finally, it’s important to recognize the role of change. Canada used to build transit projects for similar prices to many of the other countries discussed favourably here, even in fairly recent memory. Clearly active changes that we have made in the way we design, build, plan, and govern have taken us from competitive to among the most expensive in the world.

But, these changes were often made to avoid past problems. It’s key to recognize that at decision points after a big project that particularly when something was unpopular or went wrong we often responded not by getting better, but by doing something different and this different thing — though we may not have recognized it at the time, and may still not — was often worse. It is also possible that when being influenced to change — that is, to learn — that those learnings may be negative. For example, imagine the trend to fare-free transit sweeping America spurred on by headlines from other cities, even as early adopters like Tallinn, Estonia abandon the policy.

What follows is a categorized breakdown of the major factors driving transit costs. While we have tried to categorize appropriately, many factors bleed across categories. While we’ve tried to frame factors in terms of how they increase costs, some are purely positive:

Design

Design refers to the way stations and lines are laid out and architected.

1. Transit stations are often designed to be larger than they need to be to serve their function. This can mean being much longer than the trains served, or featuring significant non-passenger spaces, or void space. This means more materials, more excavation, more labour, and more time. For example, older subway stations were often much smaller, and yet generally do not have serious capacity issues. Office and utility spaces were less plush, might have been found above ground, and were often shared.
2. A related issue is that stations often have many entrances, even when fewer might do.
3. There is often not much variability in the size of stations, and the provision of entrances and other amenities, even if only busier stations justify these extra costs.
4. Station finishings are often standardized, but architectural designs where more savings might be found are often poorly standardized. Vancouver has done well with this, as many new SkyTrain stations have a standard layout, with finishings being the main differentiator. Standard designs could increase economies of scale for all kinds of mechanical and ventilation equipment — for example, being able to order the same escalators for every station instead of slightly different ones.
5. Station art and finishings rarely have a huge impact on costs, but grand open spaces and bespoke designs for each station with minimal standardization (As seen on Toronto’s York University subway extension) do. Standardizing the expensive thing — architectural design and layout, while letting the finishings be distinct would be more cost effective.
6. The underutilization of the Swiss-German phrase “Organisation vor Elektronik vor Beton,” – “Organization before electronics before concrete” which implores the use of better management (i.e., more disciplined operations) before the use of technology (i.e., advanced signalling that enables tighter operational margins) before the use of concrete (i.e., new infrastructure).
7. Related to the last point, significant waste occurs when expensive physical structures are constructed as necessitated by bad organizational practices or outdated technology. Many transit systems could have built, or could build, smaller stations if a far less expensive investment in automated signalling enabling higher frequencies was made.
8. Poor understanding of low-cost designs. Cities internationally use different station design techniques, such as Seoul with alternative stair arrangements, and cities domestically have delivered low-cost stations — for example, Lincoln, an elevated SkyTrain station in Metro Vancouver, was constructed for just $28 million in 2016, less than 1/10th the cost of stations on the also-suburban Vaughan subway extension despite featuring platforms greater than 50% as long.
9. Planning alignments in ways that increase costs. For example, many projects like the “Transit City” light rail lines were overly prescriptive not only on corridor design (they had to run down the middle of streets), but also on the actual alignment used — for example, following a busy street when a parallel corridor with fewer obstructions was available nearby, but which would not satisfy a fixation on a particular roadway. Reuse of space in existing highway and rail corridors often is underexplored.
10. Expensive station designs poorly optimized for cost with little benefit. Examples include not locating elevated stations at the side of streets to reduce massing and height, or not removing fare gates as in German-speaking cities or Copenhagen to simplify layouts.
11. Related is considering how some tradeoffs might shift costs or improve cost-effectiveness. For example, removing fare gates might mean an operational cost in ticket inspectors, but perhaps that cost is lower over a long period of time (say 100 years) given current high costs, or could be combined with existing roles.
12. Cost-benefit analysis is often done for entire projects with somewhat questionable results (we build lots of projects where the calculated costs appear to outweigh the benefits); it unfortunately seems to be less frequently employed for individual design decisions. For example, if a station must be made larger to accommodate extra escalators or elevators, it might be rational to ask how valuable the benefits are.
13. Poor technology choice driving high costs. The Eglinton Crosstown, for example, featured dramatically larger stations than the Canada Line, driving higher costs despite featuring the same capacity, because the trains have an inefficient layout. Many European cities use trains that are small by North American or Asian standards — Madrid, for example, uses among the smallest subway trains in the world.

It’s important to note that design factors really do mean considering tradeoffs in a way that is less true of many policy and construction factors. Tackling design factors generally means giving something up, usually along with a recognition that the thing being given up is not so important to the transport utility of a project. By comparison, more productive construction or efficient procurement can make any design less expensive.

An important note is that, as costs rise, the need to cut sometimes valuable things, such as additional entrances, increases, which can reduce a project’s value and sometimes make things even worse. Because of this, design factors might in many cases be the last thing to tackle, and the first thing to add back if costs come down.

This highlights both a real benefit of having low costs — being able to invest in more “nice-to-haves” — but also acts as a reminder that many things, like fare gates, are not actually necessary, and may have costs we don’t appreciate, like adding friction to the experience of using transit.

We were hesitant to discuss this, because no project needs to tackle all of these factors to achieve low costs. However, low costs are the result of choices, and who is to say which are the right ones? Some might argue fare gates are needed, in the same way some might argue that disruptive construction is unacceptable. Most low-cost jurisdictions do some of these things, and getting to low costs often will mean that some of these considerations need to be “in the mix.”

Processes

Processes refer to how projects are constructed, by whom, and by what methods.

1. Slow, drawn-out projects with laborious processes. Time is a central element. Labour is paid based on time, equipment depreciates based on the time it is used, planners and engineers charge by the hour, and disruptions and construction tie up space that has direct and indirect costs over time. Because of this, a slower project will often be more expensive. The transit projects in recent Canadian memory that were among the most cost-effective — the Canada Line in Vancouver and the REM network in Montreal — were both completed unusually quickly.
2. Overbuilding the right-of-way. The “right-of-way” of a project determines how it functions, but when considering a “grade-separated” metro project that features no level crossings, tunnelled, elevated, and at-grade alignments are all options. Choosing tunnels where elevated might work, or elevated where at-grade might work, increases costs. Tunnelling is particularly risk-laden. Vancouver has historically done a good job of this, with fairly minimal tunnelling and all manner of existing rights-of-way used for elevated guideways. Even Toronto — when it built for less — put the subway at grade through Rosedale, one of the most affluent parts of the city.
3. Overbuilding of tunnels. Relatedly, deeper and larger tunnels increase costs by increasing the amount of excavation that must be done to reach them — often dramatically increasing station size and encouraging frequently more expensive construction methods like mining, as well as increasing risk.
4. Expensive construction methods. Construction methods matter. Cut-and-cover construction excavates less material for tunnels, is simpler, and is faster, which reduces costs.
5. Construction methods that are not context-aware. Tunnel construction method selection can be a little too prescriptive in armchair-engineer circles. Tunnel boring can be done cost-effectively and efficiently, but will require greater depth than cut-and-cover, meaning if it is to be competitive it needs to make that up elsewhere — for example, through shorter alignments. Mined tunnels have been common historically in Montreal, Stockholm, and Sydney and were cost-effective because of favourable rock. In fact, it was Montreal’s very expensive Blue Line extension that abandoned the city’s traditional mining in favour of tunnel boring — highlighting the potential for negative learning.
6. Using complex construction methods where not required. There is a crossover point where mined stations, or even stations within a tunnel bore, might make sense — but not because their costs ever come down, but because the alternatives grow very expensive. Sometimes this may be required, but more and more often these methods are being touted not to get out from between a rock and a hard place, but as a capitulation to NIMBYs or inflexible organizations.
7. Poorly mapped utilities increase risk and work required to build new infrastructure, especially underground.
8. Poor construction productivity and coordination. Given the large volumes required, low construction productivity is a major cost for transit projects.
9. Onerous union rules — for example, around hours worked or procedures — can dramatically increase costs and reduce labour availability. Many low-cost jurisdictions in Europe, for example, are heavily unionized, but with better relationships and more reasonable policies. This, of course, probably means less benefit to each union member, but more union members and work for them — a more equitable distribution of benefits.
10. Excessive contingency budgets create room for cost increases, since allocated funds will rarely not be spent. Cost-effective jurisdictions keep the contingency tight at around ~10%, which creates an incentive for efficiency and puts a high cost on absolute cost increases.
11. Low-trust relationships and inflexible processes. Flexible, trusting relationships allow contractors and project teams to change minor project details and designs on the fly when conditions change — for example, finding an unmarked utility or a building drawing that is incorrect, instead of slowing down the whole project at great cost to accommodate a bureaucratic change process.
12. A fear of disruption. Often public backlash means scaling back construction practices that might be loud, or that may be highly disruptive even if they limit total disruption duration. The Canada Line’s cut-and-cover construction allowed it to finish in just a few years, while the Eglinton Crosstown took over a decade. The Crosstown’s bored tunnels and combination of mined and cut-and-cover stations was sold as less disruptive, but still took up significant surface area and meant that a slightly lighter disruption footprint lasted much longer, likely meaning more disruption overall.

State Capacity and Policy

This refers to the government’s ability to deliver projects, how it does so, and how politics and interactions between state and non-state actors work.

1. Political meddling drives high costs, often even just by suggesting “more study” when constituents complain. In many functional countries, politicians serve to simply provide a yes or no for major decisions, not to micromanage. This is not a universal rule, but countries with high costs do often see more political interference with planning and engineering processes, which slows them down and leads to worse outcomes. One can imagine how NIMBYs can force a more expensive alignment, or residents might demand politicians interfere to help them secure additional benefits.
2. Politicians rarely use their power to improve project efficiency. Politicians rarely help navigate interagency disputes or apply pressure to various parties to ensure a project moves forward efficiently. This can be seen in Toronto, where the City of Toronto is not in charge of new subway construction, but still appears to hold sway over projects — slowing them down and increasing their costs.
3. The non-professionalized transit planning process. In many Anglosphere jurisdictions, a select segment of the public who show up to often inconveniently timed public hearings will vote on different alignments. In many countries, this process is done the same way we might plan a new sewer.
4. Poor and inefficient processes, notably for utility management. The use of expensive change orders in many North American projects is a great example of this, and highlights an adversarial dynamic between the public and private sectors. A great example raised is that cities like Paris and Rome are not lacking in old and poorly mapped underground pipes and wires — and sometimes even ancient ruins — but these jurisdictions tend to have better processes for dealing with surprises and things like utility relocations. The REM in Montreal had a legislated ability to force all kinds of responses to permits and the like, and it seems no small surprise that it has come in as one of the most affordable Anglosphere transit projects in recent memory.
5. Extraction. A big problem with big projects is that parties interacting with them see them as an opportunity to extract concessions and valuable new amenities — such as above-market-value property compensation or unnecessary sound barriers. This is also often a problem with utilities and other public agencies that can get their infrastructure replaced with other people's money — often thanks to vague or poorly conceived processes. The same effect sometimes exists with residents or activists who frame transit — an amenity — as pollution and extract additional benefits. For example, the Transit Costs Project details a surprisingly expensive bike path promised along Boston’s Green Line extension. We don’t think such projects are bad per se, but bundling in this way is inefficient and bad for transit. In some cases, like York Region's Viva network, the entire point of the “transit” project appears to be funding road widening and attracting development, so the whole project is really just extraction with some transit icing.
6. Use of public-private partnerships. It’s important to note that P3s are not a guarantee of high costs; several of Canada’s more cost-effective recent projects have actually been P3s. But P3s are rarely if ever used in the most cost-effective jurisdictions, and create an elevated floor for costs because of the need for long-term returns for the proponents. They are not necessary for cost-effectiveness and bar us from the most cost-effective conceivable projects.
7. A lack of transparency. In many countries, the cost of materials for projects as well as all manner of planning and contractual documents are publicly available not to mention construction progress information. This has often appeared as an anti-corruption measure, but also helps civil servants better understand pricing and better price future work.
8. Lack of open reflection. While sometimes Canadian transit systems have engaged in public inquiries when things have gone wrong, they are not common enough. Careful reflection and analysis of mistakes that politicians might want buried is important for understanding how to do better, but also for not making rash reactionary decisions. That a public report discussing what went wrong and what was learned from the Eglinton Crosstown is missing is completely egregious, and further wastes tax dollars by not at least capturing lessons from the project.
9. Over-the-top fire safety. Since officials and authorities responsible for dialing up fire codes and safety rarely need to consider tradeoffs or the costs of new regulation, we often create expensive, over-the-top design changes for things like ventilation, exits, and separation walls — even when projects built without these things generally operate safely. Requirements are often not reasonable, as well as being grounded on unlikely and worst-case assumptions that force expensive designs. This is in some ways a form of extraction.
10. Bundling of operations and maintenance. Being able to bundle operations and maintenance of a project allows the state to have less responsibility, outsource work, and is in some ways an insurance policy for infrastructure with unexpected problems, but this is expensive. States are not individuals and generally should not pay premiums for hands-off, high-simplicity products. The bundling of these costs with project capital costs also reduces cost salience in both the public and government, and reduces transparency.
11. Understudied policy cost-benefit. Sometimes different policy approaches may have valuable cost-benefit profiles. For example, the Canada Line’s disruptive cut-and-cover construction costed more than people realized because it lead to lawsuits. However, these lawsuits totalled a small sum, that is likely only a few percentage points of the cost of deeper tunnels. Being willing to pay out benefits — extraction, or face lawsuits is probably something which should be considered more often.
12. Offloading of risk. Trying to shift risk to the private sector is expensive, and ultimately generally futile, because while a private consortium can walk away from a project, a government generally cannot. The Ottawa O-Train’s many problems, as well as those on Edmonton’s Valley Line tramway, financially fell on the consortia running the projects. In Ottawa’s case, performance issues seemingly never got fixed, and now all Canadian P3s are apparently more expensive because of the higher odds of being held to account. In Edmonton, the city decided its own transit operator will take over from the P3 consortium on the Valley Line.
13. A lack of public sector capacity and capability. A public sector that is not capable and lacks the human resources to carry out its role in projects can create all kinds of problems. Even when the private sector is heavily involved in a project, you need a public sector knowledgeable enough to know what makes sense, and if needed, call their bluff. A public sector capable of taking on projects itself also creates a competitive threat. Istanbul Metro is one such organization, and many cities had capable organizations before they were hollowed out, or at the very least cut down. In Istanbul, some of the rolling stock was actually designed by the agency and built domestically by Turkish firms. The Bay Area’s BART also received much fanfare when it took some of the work for its new trains in-house, which led to significant savings. Consultants can be useful as manpower on demand (though appropriate public staffing remains critical), and they are used to some extent for various purposes around the world. But they need to be carefully managed lest optioning, huge meetings, and other forms of wheel-spinning drive huge bills. A capable public sector will generally be a better manager. It’s important to note that bringing things in-house is so valuable because individuals rarely make significantly more working at a consultancy; instead, high costs go to company overhead. An interesting solution to this problem can be public-sector consultants, often working at a large transit authority. This allows smaller cities to hire reputable and knowledgeable talent on a temporary or project-by-project basis, and has strengthened the national standing of agencies like RATP in France and Milan Metro in Italy. Having such consultants makes keeping talent in the public sector more straightforward.
14. Purchasing too much private property for construction and structures. Since the public — or specific members of it — often really dislike closed parks, cut trees, or other disruption, there is a tendency for new projects to build things like stations and do construction on acquired private property at great cost instead of publicly owned land and rights-of-way.
15. Imitation of expensive jurisdictions. This sets baseline expectations high, and makes learning the wrong things and missing the right things likely. In an attempt to justify high proposed costs for platform screen doors in Toronto, TTC staff mentioned that they looked to New York’s MTA, famous for expensive subways, elevators, and other projects — and also an agency that has not implemented platform screen doors.
16. Following trends. Emulating construction and procurement methods without a critical eye to whether they are applicable, or have even been successful. Barcelona’s Line 9 and 10, which are famous domestically for construction problems and delays, were more famous in the English-speaking world for their use of a novel “in-bore” station design to cope with Barcelona’s dense historic centre and other underground infrastructure. Vestiges of this idea were then utilized in other contexts, like low-density San Jose, at enormous cost.
17. Stop-and-start development. Stopping the construction of projects for long periods increases the risk of losing professionals and institutional expertise, as well as industry capacity to build and public understanding of the costs and benefits of large projects.
18. Lowest cost as a primary deciding factor for contracts during bidding, and the undervaluing of technical expertise.
19. Uncompetitive construction markets and dynamics. Istanbul again provides an interesting example. Projects are carefully structured to maximize competitive bidding, and the construction market overall is large and competitive, with many capable firms. In Canada, the market is highly concentrated.

The Transit Brief Cost-Driving Factors

We also have some factors we’ve observed over the years and across geographies that we think ought to get more attention, and which are not featured in great depth in the cited literature.

1. Technical system standardization. While not widely explored elsewhere, in China nearly all metro trains are built to just a handful of capacity levels and designs. Since all infrastructure is built around the trains, China’s hundreds of subway lines end up having vehicles, systems, and stations that are remarkably similar from city to city. This makes for economies of scale, but also encourages learning and the transfer of knowledge from place to place. In Canada, the REM, Canada Line, and Ontario Line — all automated metro lines — could have all used the same trains, just with different numbers of cars, allowing for greater economies of scale with future orders.
2. Cost consideration coming too late in a project. Oftentimes cost blowouts or downward pressures come only as budgets are squeezed, or a better sense of actual costs pushes estimates up. Unfortunately, as a project progresses, the options for reducing costs narrow significantly. Removing platform doors or pleasant finishings saves almost nothing, but redesigning a mostly built station is not practical.
3. A decentring of costs. As mentioned above, there is a sense that costs do not matter. They have low salience with the public, public servants, and even many politicians. With so little attention paid to costs, is it any wonder that with all of the other demands on a project, cost is the one that flexes?
4. Unquestioned requirements. Questioning requirements has come into vogue thanks to Elon Musk, but it has long been a sensible policy for public projects. Often things like new bike paths, extra entrances, underground office space, and more move through a plan because they are not questioned. Better transparency and itemization would help by making the unusualness of these things, and the extent to which they are unnecessary, more obvious.
5. Media narratives. Media reports make a big deal of cost overruns irrespective of how reasonable the initial budget was, and how good of a value the project is after the overrun. This and delays create a vicious cycle where slack is added to budgets and timelines, and then eaten all to avoid bad headlines.
6. Doing very little and then rapidly scaling up construction could potentially run into issues of construction capacity. Toronto, for example, has gone from building nothing, to one subway extension, to a tram and a subway-surface hybrid line, to a new subway line, three extensions, and multiple trams and major regional rail projects in under a decade — all while in a massive housing building boom. It’s hard to imagine this does not drive prices up further.
7. Cancelling projects is too uncommon. While politically expensive, the unwillingness to pause or cancel projects means contractors and agencies know that governments have to move forward, often without consideration of the costs.
8. Safeguarding and planning. More long-term planning, and reserving space on the surface as well as underground for long-term planned projects, could make those projects less expensive with time as pieces fall together, as well as de-risk them. This would require sticking with plans to some extent, but the benefits might be so large that not building everything you protected for may be acceptable.
9. The nonessential nature of transit. In many jurisdictions where transit is expensive, it is not so heavily used. This weakens politicians’ understanding of the projects and their benefits, but also weakens institutional pressure to deliver good projects and get better — after all, it’s just a nice-to-have.
10. Willingness to pay. If anything could signal that the current status quo is acceptable, continuing to fund projects regardless of price does just that.