Speed matters

Of time, life and space. Or why transit speed matters.

The recent opening of the Finch West LRT has reopened the (probably niche) debate about what matters when we pick a mode and decide what to build and where, to improve the lives of transit users and try to lure more people away from personal motorized vehicles and onto public transportation.

As advocates, experts, users, and people in the Canadian transit sphere realized that the travel times scheduled for the Finch West LRT were substantially longer than what was promised in early planning documents underpinning the initial mode choice of the business case, a sort of reckoning occurred in this little corner of the virtual world. Many, rightfully, pointed out that this is not what was promised and that travel times remain substantially the same as the buses they replace. Others run to defend the project, stressing that speed is not the only metric that matters, probably not even the most important one.

While the extremely slow pace of Finch LRT is due to an accumulation of poor planning and operating choices, like extensive slow zones, a punitive CBTC system and inferior signal priority, and might hopefully increase after the trial period, there is indeed a tendency among some planners, urbanists, advocates and transit allies to dismiss speed as a key driving metric in the difficult task of “picking a mode”, specially when it comes to street-running trams. They stress that the obsession with speed is misleading, driven by a car-brain engineering perspective of transportation, and whatnot. I disagree. Strongly.

Of time, life and space. Why speed matters.

Let me be clear. I think it’s arrogant and condescending toward transit users to assert that speed is not really critical, that “slow is better,” and that absolute travel times matter little. Because they do, and it’s paramount. There is no technology, no “perceived travel time” abracadabra that can infinitely stretch one of the constants of our species: our biorhythm is organized in daily cycles determined by the constant alternation of day-night, sleep-wake cycles. Most of our days start in the morning and end at night at the place we call home (when we have one). The amount of time we can dedicate to reaching the places we need or want to go, whether for work, studying, doing errands, caring for someone, or enjoying life and the company of family and friends, is finite, and the more the time spent travelling grows, the more it’s taken away from other things.

Our time is finite. It’s one of the hard facts of life. As are our energies and will to spend time moving around, even when cocooned in a quality ride. Yes, there is plenty of literature and empirical evidence that confirms that the perception of time is not uniform and that some factors can induce us to tolerate, or even embrace, longer travel times. After all, I am one of those lunatics who regularly take the 11-hour Amtrak trip to New York, so I can’t deny that multiple factors determine our travel time budget and, hence, the speed we are willing to accept. But from the Marchetti’s 30-minute constant to more systematic recent studies, we know that the time we are willing to dedicate to daily travel remains remarkably similar across societies. If the transportation we have access to is slower globally, we will travel shorter distances. Vice versa, if we have access to faster transportation, we will gain access to a larger territory and what it can offer. When our trips get longer, we will eventually grow tired, skip going places, and reduce our “hunting territory.”

Some say that this can be good. That instead of looking for even faster means of transportation, we should foster a city of proximity, that is, the 15-minute city, or, as we called it in the less fashionable times when I was in architecture school, the mixed-use city (la ville mixte, for the French). That is absolutely something that we need to work for. Having essential services of everyday life within walking or biking distance is a vital ingredient of a car-lite, more humane way of life. But the idea that, in our modern times, we can have all we need or want within a 2 km radius is just as laughable as the center-less car-dominated city-region Utopia (or Distopia) of Broadacre City. There are oh-so-many services, functions, places of gathering, and professional and personal networks that will always transcend hyperproximity. That particular restaurant. That conference where you want to socialize in person with the broader national or even global professional space. That bookshop with a weird atmosphere and a friendly, flirty clerk. That one café that hasn’t fallen for the third-wave fad. That friend who has moved to the other side of the city, looking for cheap real estate. That single clinic in the entire region that has an MRI machine. And, well, that job that really fits your profile and you actually enjoy doing. People more intelligent than me have translated this evidence into various theories of why cities exist and why they keep attracting more people: because they offer unrivalled job markets in terms of quality and quantity (A. Bertaud), or because they offer consumption amenities like no other place (R. Florida).

Speed may just be another anodyne metric to describe the quality of movement. Nevertheless, it captures the very essence of the complex interdependence of time and space that defines modern, urbanized, advanced societies and the daily lives of the individuals inhabiting them. Average speed, or, more technically, “commercial speed,” together with frequency (another time-based metric), is a good proxy for how many places we can reach in our daily lives. It is not some heartless engineering obsession that we should dialectically (and artificially) oppose to a more humane vision of a low-pace life for urban flaneurs. It is a metric of opportunity and fulfillment. Being able to go more places, to pick between more jobs without investing innumerable more hours of your life in crawling around in snail-paced public transit, is a fundamental right and a way to ensure that more people can enjoy the fruits and opportunities of our modern prosperity.

Finally, another recurrent argument is that reliability matters much more than speed. So a tram slower or as slow as the bus it replaces is still a good value for money because it’s reliable. I disagree. Reliability is necessary, but not sufficient. Don’t take me wrong, reliability is paramount. It’s a metric that represents how well a given service will fulfil the two basic promises written in a timetable (which is the “contract” between the provider and the users). It tells me when I, the rider, can expect to depart from a particular stop and when I will arrive at my destination so that I can plan accordingly. But so it’s travel times, because that’s one of the driving factors that will push me to decide whether I will take that transit trip, I will opt for another mode, change my destination or give up and skip that activity altogether.

A matter of trade-offs, layering and urbanism.

At this point, the reader may think I’m obsessed with absolute speed, so I advocate for just building metros and TGVs everywhere. Well, obviously not. It is intuitively well known that, with transit, the faster we go, the more we lose coverage, because as soon as we get into a segregated right-of-way, stop distance is the primary determinant of the average speed of transit, all else equal. So, the further you go, the faster you need to go, and larger city-regions, which will have complex Origin-Destination (O-D) patterns of varying distances, need a layered transit network to offer competitive travel times to as many O-D pairs as possible.

Another way to define modes is to characterize them by their “typical” average speed and how that relates to a few key aspects of their urban integration. Credits: Marco Chitti

In great transit cities, especially large metro areas whose commute basin spans for dozens of kilometres, transit is layered across modes to provide for different speeds depending on the O-D. There isn’t a single way to layer modes, as you can see from a few diagrams of idealized models of mode-chains or mode-layering across the globe that I prepared for a presentation (in French, sorry). One is not intrinsically better than the other, but just optimized to serve certain types of O-D better and is intrinsically linked to the “urbanism” of a place and its history (i.e., how origins and destinations are spatially distributed as a result of economy, society and policy over time).

An idealized model of transit in a Canadian city, with an underdeveloped high-capacity, higher-order grade-separated “fast” transit serving the inner city fed by local, skip-stop and express buses irrigating the vast expanses of the suburbs.

An idealized model of the Central European model, with a dense, layered system of legacy trams, metros or pre-metros, serving the inner compact city and “suburban” and “regional” trains reaching further out, with only local buses as a short-range last-mile feeder.

An idealized representation of the Australian model, where fast and frequent suburban rail represents the backbone of regional transit across vastly sprawling regions, with legacy or modern trams and buses as proximity transit in the core and suburban buses as feeders.

The Japanese megacity model, with seamless interoperability between trains and metros, offering an incredibly dense rail network across the region, primarily fed by walking and biking and, marginally, by a weak local bus network.

If you have followed me for a while, you’ve seen that I’m enthusiastically cheering (sort of, for Italian standards) every news of Bologna’s upcoming tramway, while I can’t spare any occasion to stress my skepticism of “LRT” or tramway projects in many North American cities, including Toronto and Montréal. The reason is that, in the vast expanses of our sprawling American and Canadian metropolises, what we really lack, in most cases, is still the fast, higher-order transit that makes travelling across the region feasible in a reasonable amount of time. Trams, when built well, are probably better suited for where they were initially, the streetcar suburbs whose development they made possible, but are a poor value-for-money (especially at $200+ million per km) for suburban corridors, at least most of them. Bologna is a medium-sized, very evenly dense and compact Emilian city where a 12-15 km trip brings you from one edge to the other of the central built-up area, with an extensive and vibrant low-traffic core and limited higher education and tertiary job sprawl. And the suburban and regional railway network is already quite well developed, so the trade-offs between slower travel times, better coverage, and lower costs offered by choosing trams instead of a light metro as the higher-order inner-urban transit make sense there. The average speed urban trams can offer (15-18 km/h) is good enough to achieve an optimal level of accessibility and tolerable travel times for most across its relatively small commuting basin.

Wrapping up, speed matters, but how much speed depends on the where and how of local urbanism.

A corollary: cheating on speed is a planning crime.

I’ll go ahead and conclude with a reprimand to my fellow planners. It’s gonna be a bit harsh, but I want to pass the message in a neat, straightforward way. Dramatically misrepresenting the actual average speed and running times your transit infrastructure can achieve is not a minor planning “misdemeanour.” It’s a major “crime”. Travel times directly inform ridership models. In the typical 4-stage agent-based model, the third step, the modal choice of the modelled “agents”, is highly dependent on two functions: cost and time. Underestimating travel times (or frequencies) leads to excessively optimistic forecasts, especially for the modal shift of high-earners and high-car-ownership agents. Moreover, a reduction in travel times, both direct for transit users and indirect for drivers through reduced congestion, is the key benefit that can turn a cost/benefit analysis for transportation projects positive. Misrepresenting running times in the very early stages of planning, such as during early scoping (business cases), mode and alignment choice, can lead to a severe misunderstanding of the benefits and costs of the different options, biasing decision-makers’ choices…or reinforcing their bias, which is often the case.

Small discrepancies are inevitable between the very first hypotheses and the actual running time. Reality always finds some way to surprise us, in ways we didn’t think about, even in the most scrupulous and honest planning exercise. And we all develop an optimism or a pessimism bias, as Giandomenico Majone liked to stress. But that’s why it’s even more fundamental that we understand what it takes to achieve certain speeds, especially in the context of street-running transit. We should develop better tools to quickly identify which downstream design choices can cause us to egregiously miss our target planning speed, like intersection design, stop spacing, signalling, and alignment geometry. Understanding this is paramount, as we buy our way from planning to design, one compromise after another, some of them very consequential and poorly understood across the decision-making chain. Helping political decision-makers to understand the consequences of certain planning decisions, even very technical ones (like installing CBTC on a tram, a rather odd choice or not adopting international best practices for signal aspects), is the duty of an ethical professional.