What does ‘progress’ mean for public transit?
I say it means solving the problem endemic to transit operations, which I described in detail last time. I call that problem the Endless Emergency: the fact that public transit, across North America, always seems to be in crisis. Transit service generally ranges between poor and adequate, always falling short in some way, in some combination of poor speed, frequency, reliability, cleanliness, and/or orderliness.
When confronted by these shortcomings, operators and their advocates respond that the issue is funding. If only the government—city, state, province, federal—would pay more, the system would be able to function as it should.
From last time: Twitter users exemplifying the Endless Emergency. It’s still the case that the claim about BART is untrue; before the pandemic, BART’s farebox recovery ratio was between .6 and .7
This struggle, for more funding in support of inadequate service, seems perpetual; an endless emergency. But the reason the emergency is endless is because of the choices we’ve made. If we had made different choices in the past, or did things differently today, we’d get different outcomes.
Today, we’ll explore one of the things we could choose to do differently. Today’s hypothesis is that the root of transit's dysfunction is the fact it does not cover its costs with fares, and fixing this problem would fix everything else.
Solving this problem will first require us to accept that transit operation is a business, not a charity nor a cost centre, and should be run like one.
The Logic of Cost Recovery
Consider a typical North American transit system, as per this, my favourite chart:
Also from last time. Source: Canadian Urban Transportation Association, Why Public Transit Needs Extended Operating Support, 2021
As it shows, no North American agency covers its operating costs through fares; Toronto can cover about two-thirds, Los Angeles less than a fifth, everyone else in between. The rest of the costs are covered by government subsidy. This dependency creates a cascade of problems.
Firstly, depending on external sources for operating funds fundamentally skews how operators make decisions.
Coca-Cola, to the best of my knowledge, makes its money selling soft drinks. At any given time, its leadership has a variety of options it can pursue to make more profit, but they all rest on the fact that profit flows from one place: the net of drink sales after production costs and overhead. The firm’s business strategy depends on playing with the variables in this equation. It could run a marketing campaign to increase sales; it could optimize its bottling plants to run at lower cost; it could adjust the price point of a Coke (multiplied by expected sales) to increase the total revenue gained. There are always a variety of approaches the firm might take, all resting on a common understanding that the goal is to increase revenue or decrease costs, such that profit goes up.
A transit operator is not so fortunate. As per the chart, it only gets about half of its revenue from riders, and almost all the rest from government. This makes all its decisions more complicated.
Take route planning, for example. A system that covered its costs would run routes that had enough rider demand to justify them. Routes that didn’t have enough riders would be cut; ones where riders were crowding the vehicles would get more service, to absorb that latent demand.
Because operators depend on government subsidy, they are unable to proceed in this fashion. Route planning stems from a variety of motives: serving riders, but also maximizing geographic coverage, and satisfying political direction. This confusion means that service can be run even when density can’t support it.
For example: years ago, I was involved (he said with deliberate vagueness) in determining an appropriate service pattern for a transit route. The efficient thing to do was to run only express service, which is what I recommended. This recommendation failed because of some constituencies that were used to receiving local service, whom the relevant government preferred not to antagonize. Therefore local service had to continue, despite the fact that it would lose most, but not all, of its riders to the parallel express service, and as such lose significant sums of money on every single trip. Everyone involved recognized this, but canceling local service was nonetheless off the table. The government paid the piper, so the government got to call the tune.
Or consider capital upgrades. We might think that an agency will invest in new infrastructure or assets, like electric buses or better maintenance bays and so forth, but only on expectation that the improvements offered cost savings (short or long term). Absent such an expectation, it would reject the opportunity.
We might think this, but we’d be mistaken. I could (but won’t) name any number of transit operators that, I happen to know, were forced by government mandate to experiment with hydrogen-powered buses. The operators did so for precisely as long as they were required to, at which point they abandoned the technology as expensive, unreliable, and ill-suited to local conditions. They would not have invested money and time in such misadventures on their own… but it wasn’t a decision they could choose not to make.
They couldn’t choose not to make it because their reliance on subsidy made them captive to the whims of their funding partner. Nor is that capture the worst of it; as discussed last time, reliance on external subsidy introduces warped incentives into how both labour and management conduct themselves.
In most industries, wage growth is constrained by overall profitability: if Coke isn’t selling enough product, there isn’t enough money to give raises, and everyone involved knows this. But because transit operators aren’t constrained by fare revenue, the unions negotiate knowing that agencies can always seek more subsidy… and management has a similarly weak incentive to push back. Such behaviour, over long timescales, is a major cause of the Endless Emergency.
The perversity of the arrangement affects management in other ways. Without market discipline, how do you evaluate performance? There are many possibilities: ridership numbers, or cost control, or good labour relations, or absence of media criticism, to name only a few. Each metric creates different incentives, meaning trade-offs are difficult to parse. Worse, poor managers are always able to shift the goalposts: whatever metric is going well, that one is highlighted, and the rest set aside. The result is a system where good managers are handcuffed, and bad ones always have excuses ready to hand.
There is a better way.
The Business Case for Cost Recovery
Contrast transit operators with other public-service providers. Electric utilities, for instance, operate as regulated monopolies but generally cover their costs through user fees. In Ontario, for example, the Ontario Energy Board regulates what electricity distributors charge for power. These fees aim to cover the reasonable costs of operation; fund capital maintenance and upgrades; and permit a (specified and regulated) return on equity. At the same time, the system rests on an understanding that electric utilities are expected to recover their own costs from their revenues and not receive subsidy.
Consequently, utility executives have a clear incentive: to make a return on equity by serving customers efficiently while maintaining their infrastructure. If they can’t do this, the solution is not a bailout, but to find new executives who can.1
Transit executives should be expected to do the same.
Specifically, transit agencies should operate as regulated monopolies with a clear mandate: serve customers while covering costs from fares, on understanding that the agency will not receive subsidy.
Regulators would ensure fares are set high enough to maintain good service and provide modest returns on equity, but no higher than necessary to achieve these goals.
Given this constraint, the utility's path to growth would be simple: serve customers well within those regulated rates.
But what would it mean to ‘serve customers’? Coke executives sell Coke; electricity executives sell power; what does transit sell? Right now, operators are lumbered with a variety of mandates. They must simultaneously:
Maximize ridership
Provide service across their entire geography
Maintain low fares, while subsidizing certain riders with even lower, or no, fares (e.g., veterans, students, the elderly)
Promote environmental sustainability, among other aims
But they cannot drive hard at four goals simultaneously. Instead, they should try to drive hard at the most important one, the one that supersedes the others. That is ridership.
The Ridership Imperative
An operator focused on cost recovery would need to set fares high enough to cover operating costs and maintain its assets. And like other regulated utilities, transit agencies would have their fares capped at this level. For a given ridership level, the fare would be high enough to provide the revenue necessary to cover costs and earn a modest return on equity, but no higher. Arranging matters this way would create a crucial dynamic: with fares fixed and subsidies eliminated, the only way for an agency to increase its revenue would be to attract more riders.
This alignment of incentives would transform how transit agencies operate. Instead of juggling multiple competing priorities, they would have a single clear imperative: maximize ridership within the constraint of cost recovery. Every decision, from route planning to service frequency to customer experience, would be evaluated based on its impact on ridership. Better service would attract more riders, generating more revenue that could be reinvested in still better service, creating a virtuous cycle.
An operator that focuses first, last, and only on ridership as its key metric would behave rather differently than what we’re used to.
Fares would be different. The cost to use the service would be the same as, or in fact greater than, the cost to run it. Transit users would find their budgets stretched, and hard; to judge by current farebox-recovery ratios, fares would have to rise by at least half of what they are now, and probably double.
Route planning would be different. Instead of spreading service thinly across wide geographic areas, operators would concentrate frequency on corridors where demand is highest. This might mean fewer total route-kilometres, but each route would run frequently enough to be genuinely useful; useful enough that patrons could plan their lives on the expectation that the service would be there when they needed it.
Labour relations would be different. Staff would know there were limits to the wage-and-benefit packages the employer could offer, and that extended strikes would shrink the envelope for future years by driving away business. Union politics wouldn’t become sweetness and light, but it would be no worse than prevails in private industry.
Management would be different. Under regulated fares, an agency's financial health depends entirely on attracting and retaining riders while controlling costs. This focuses management attention on two primary metrics:
Service Quality, measured by ridership-per-service-hour. This metric shows how many people choose to use each hour of service provided. Improving this metric means making service more attractive to riders, through whatever means: better frequency, reliability, cleanliness, or customer experience
Operating Efficiency, measured by cost-per-rider. This metric ensures service-quality improvements are delivered sustainably. Reducing this metric means finding ways to serve each rider more efficiently, whether through better route design, improved maintenance practices, or better technology
Secondary metrics, like on-time performance, customer satisfaction, and cleanliness scores might help managers understand how to improve these primary ones, while distance-based metrics like cost-per-passenger-kilometer might help optimize route structures and fare policies. But these are the trees, while the forest—the core mission—remains clear: attract more riders while controlling costs.
This clarity would help managers make better decisions. Should we increase frequency on a route? Add a new express service? Invest in electric buses? Each choice can be evaluated based on its impact on ridership and costs. Unlike today's confused mandates, this would create a direct link between management decisions and agency success.
As an aside: the idea that transit operations should be self-sustaining isn't a fantasy. To the contrary, it's how transit worked in North America for decades. From the 1800s through the mid-20th century, some transit systems across North America operated profitably. New York's subway lines, for example, were built by private companies, as were Toronto’s earliest railways, which were built by a private operator.2 So the question isn't whether transit can support itself with fares alone. We know it can because it has before.
Answering the Critics
At this point I should acknowledge that this plan would be incredibly controversial.
I imagine the first complaint would be that this is ‘privatization’, and putting profits over people. But if we don’t object to managing our power grid this way, we shouldn’t object to running our transit system this way either. In both cases, the goal would be the same: operational self-sufficiency in the service of public need.
By contrast, the loudest complaint would be from users, who would suddenly have to pay much more, at least initially. The answer here is that they are only expected to pay what the service costs to operate, and it should be strange to expect that they shouldn’t. Is it better if the bulk of a rider’s trip is paid by some taxpayer who never uses the service and never receives direct benefit from it? Aside from all of the perverse incentives I’ve already described, this situation seems straightforwardly unfair.3
To go further, I believe that many current users would be pleased to pay more, if in return they got a service free from the curse of the Endless Emergency: service that was more frequent, more reliable, cleaner, and more secure.
The hardest complaint to answer is that this approach amounts to class warfare.
In a world where ridership is king, the first thing operators would abandon would be surface routes with insufficient ridership to sustain themselves. Who is using these routes? It tends to be those with no alternative: students, the elderly, recent immigrants, and people in poverty. Taking service away from these people would make them even worse off, depriving them of ready access to the places they need to visit to live—grocery stores, doctor’s offices—as well as the places some of them need to visit to improve their situation: school or places of employment.
Such a move would be politically, and even morally, indefensible. And here is where subsidy comes back into the picture. There has always been a place for public support for the least well off, and it applies here as well. It merely needs to be provided in a way that doesn’t warp the system as a whole.
There are two ways to do this. Critically, neither is a subsidy to the transit operator.
Firstly, we could subsidize the rider. If we, as a society, want to assist the poor (or the elderly, or veterans, or any other class of resident we think needs help), then give them the money. Transit operators everywhere are increasingly moving to farecards linked with digital accounts to manage payment rather than cash. A government can simply provide a regular payout to these digital accounts to offset, in part or in whole, a given patron’s use of the system. In this way the patron gets the help they need but the system’s incentives remain healthy.
Secondly, we could subsidize the route. As David Levison once proposed, “…money-losing routes [should be] operated under contract by the transit utility in exchange for specific revenue from the jurisdiction that route serves. Transit organizations would at least break even on the operation of the route. The ‘deficit’ would shift from the transit utility to the public sector, which would have a clearer picture of the costs of its wants.”
Again, subsidy enables an area to get service it would not otherwise have, but that subsidy is in plain view, enabling politicians, civil servants, and the public to determine what they want to support, and to what extent. Further, transit operators remain focused on the goal of maximizing ridership, a metric they can track and improve.
Behaviour Follows Incentives
More could be said. In an ideal world, for instance, transit would be free of subsidy, but so too would urban roads. By choosing not to implement congestion charges, we implicitly subsidize trips by car. The costs of road use are high: not only to government, which must maintain these assets, but also to every road user, whose trip imposes congestion that delays every other user. Imposing a cost-recovery model on roads via a congestion charge would align user incentives better, put transit and auto trips on the same field, and lead to improved outcomes in urban transportation.
But that’s a topic for another day. Our present concern is addressing the warped incentives that plague transit, which would go a long way to solving the Endless Emergency. A focus on ridership would improve service, and improved service would in turn grow ridership, in a virtuous circle. Such a change would be hard, and there are many entrenched interests that stand in the way. But implementing it would make for a better world; it would be progress.
This post is part of the series on Progress and Public Transit. The first outlined the problem we have to solve: the Endless Emergency. This post was about how we could address it by changing our transit operators, but we have other tools in our kit. We can also change our cities, and our technology. Those subjects will be the focus of the next entries in this series.
Respect to Steve Newman and Kevin Kohler for feedback on earlier drafts.
For more on transit as a regulated utility, see David Levinson (linked above), How to Make Mass Transit Financially Sustainable Once and for All; Matt Yglesias, American transit agencies should prioritize ridership over other goals; and Jarrett Walker’s book Human Transit, revised in a 2024 second edition.
Changing Lanes on the Road (virtually)
A reminder that the day after this post is published, on Wednesday November 20 at 11:00am EST, I’m offering a webinar on The Ever-Expanding Uses of Autonomous Vehicles. This will be the inaugural webinar in a series hosted by the Canadian Automated Vehicle Initiative (CAVI), an association for all stakeholders in the automated-vehicle ecosystem in Canada.
The webinar will describe how automated driving will appear, across sectors, in the coming years. It will pay special attention to the implications for our businesses, infrastructure, and economy, and the policy changes automated vehicles will require. Though the examples will be primarily Canadian, the webinar will be of interest to people outside Canada as well.
The webinar is free of charge. You may register to attend here.
Thanks for reading Changing Lanes! Please let us know how we’re doing by answering the poll below. And if you’d like to respond to this post, please leave a comment.
None of which is to suggest that regulated utilities are a paradise where no incentives are misaligned. Most electrical utilities in the USA are allowed, as part of their return on equity, a specified percentage of capital investment, created a perverse incentive to spend as much on capital expenditure as they can; meaning regulators have to scrutinize these decisions to forbid bad investment; leading to overinvestment, second-guessing, and many other bad outcomes.
Having noted all this, I don’t think those problems are insoluble. What’s more, I think that they’re a better class of problem than transit operators indulge today.
What changed? The history has been told in detail, but briefly: the rise of the automobile firstly created fierce competition, especially given massive public investment in highways and suburban infrastructure, reducing the customer base; and secondly congested city streets, making buses and streetcars slower and less attractive. Meanwhile, government fiat prevented fares from rising proportionate to costs, and further compelled service in areas ill-suited for density. Taken as a whole, these changes made cost recovery impossible, so operators had to rely on government support. This began a vicious cycle: subsidies led to confused incentives, which led to more-expensive service, and the need for more subsidy.
I mentioned that Toronto’s early railways were built by private operators. These refused to run unprofitable service, so in 1912 the city founded a public agency to do it, which was not today’s TTC, but the Toronto Civic Railways company (TCR). As a public agency, some politicians and citizens believed the TCR should operate fare-free. Even then, this was recognized to be a bad idea. Unfortunately the TRC compromised with another bad idea, offering fares at two cents a ride, below cost recovery. As noted on Transit Toronto’s history of the episode, TCR head R.C. Harris tried in vain to change this policy, arguing in 1918 to the city’s administration:
...it cannot be urged that this practice is sound, wise or business like [sic]. Every public utility should be made self-supporting. It is unfair that the citizens at large should be compelled to make good annually, through the tax rate, the deficit created by reason of preferential treatment accorded a section of the community. Furthermore, it provides one of the most potent and damaging arguments against public ownership…
I can’t fault Harris’ reasoning.
Transit Toronto further notes that “the TCR would be dissolved in 1921 and [operated] in deficit for the duration of its existence”.
You make a compelling case, and I agree that subsidizing the rider and expecting the transit agency to break even is a far better model of transit provision.
That said, I don’t think this would be viable in isolation.
First, as long as the bus is stuck on the same roads as other traffic, it cannot be faster than driving.
Second, as long as driving is intensely subsidized and full of sunk cost, driving will be/feel much cheaper for the middle class.
This leaves is where we are today, transit is not valuable to people with cars EXCEPT to get to places that are expensive to drive (which usually means either expensive to park or infeasible to park). So the only real purpose of transit is to provide some degree of mobility for people who can’t afford or can’t operate cars. That’s just not a big or rich enough market for a successful business to operate in.
If we dropped the gas tax and instead raised car tax (more fair in an era of electric cars) and adopted congestion pricing or a VMT based system, *then* I think your proposal could really work. But heavily subsidizing drivers is very popular, and efforts to make them pay what it costs to maintain the roads have largely failed politically, so I’m not very optimistic we can change this any time soon.
EDIT: I expanded these thoughts a bit to a note and re-stacked this -- I wish substack hadn't got rid of the button to both comment and re-stack at the same time! -- hope that gets more people to read your idea :)
https://substack.com/@burlesona/note/c-77900684
The argument for bringing the revenue/cost ratio up to 1.0 (plus a certain level of profit?) is fair, but transit (public and private) suffers from uncontrollable circumstances that make it extraordinarily difficult to "compete" for ridership with the alternatives. There are two ways to level the playing field (while holding other variables such as land use, density, socioeconomic conditions, geography, etc. constant): pricing ALL urban transportation similarly (in effect, road pricing or congestion pricing); and/or spending capital ($illions, with an M or a B) on dedicated transit infrastructure to make it time-competitive in particular corridors (subways, LRT lines, busways).
There is no operating cost recovery scenario in the world where transit service generates enough money to be able to build its own separate infrastructure, so governments will continue to have a role in building transit infrastructure (especially since the perceived political risk of blowback on road pricing is so high as to paralyze elected officials on that front). So creating a "transit infrastructure investment agency" that is separate from "transit service operating company" would be a necessary part of your strategy.
Hong Kong was often cited as a "profitable" transit agency, but that was a function of top-of-the-world population density, severely constrained road alternatives, and the transit company's function as a developer to both build apartments (demand) on their property/stations and to cross-subsidize operations from real estate profits. That particular scenario does not translate to very many other places in the world.
Nevertheless, the discussion might benefit from a comment on how "attracting maximum ridership" works in other parts of the world. In parts of Africa or Asia, for example, what is the mode split where there is essentially no public transit and a completely profit-driven free market reigns for urban transportation? And what is the operating revenue/cost ratio and subsidy situation in places like Europe and Japan where there is exemplary public transit and relatively high ridership?