So I’ve just signed a deal with Island Press to do a second edition of Human Transit, expected out near the end of 2023. A lot has obviously happened in the history of public transit since book came out in 2011, including real things like the pandemic, unreal things like the hyperloop, and some things that are real but overhyped, like microtransit. So in addition to updating examples and graphics, I plan at least five new chapters:
- Why does transit matter? A simple explanation of the unique role of transit in the city, and how it relates to all the other transportation options.
- The Wall Around Your Life. This chapter explains the idea of access and is mostly content from this article.
- Against Specialization. This chapter would emphasize that transit succeeds when an extreme diversity of people find it useful: diversity not just of race/gender/age but also of trip purpose. This explains why demands for specialized services, which transit agencies receive all the time, usually lead away from the best networks for everyone. The chapter would caution against elite projection, argue against the binarism of choice vs captive riders. This is also where the connection would be made to equity/justice frames.
- Flexible or Fixed? A review of the demand response service options, largely from this article.
- Should we Redesign our Network? How to recognize when a network needs redesign, and how to think about that.
Still, it’s a new edition rather than a new book because at least half of the book doesn’t need updating. After all, much of it is talking about geometry, and that doesn’t change.
But I’d love to know what you think! If you feel like perusing the book again, I’d welcome your thoughts, ideally organized by chapter, about what I should edit. You can email me by clicking that envelope up on the black bar, or just leave a comment. If your comment turns out to be really useful, you’ll get an acknowledgment.
Thanks!
There should be some explicit debunking of the myth that the pandemic made transit irrelevant. This could include discussions of why expecations of permanent work-from-home situations are exaggerated and how transit still serves vital needs even when white collar workers are remote. The pandemic may stimulate some needed changes in transit (e.g. better frequency throughout the day and shifting from commuter rail to regional rail), but it hasn’t diminished the need.
I would love to see some discussion of public participation and the challenges inherent in getting the public to speak on transit. You’ve discussed this topic quite often in the blog, and it might be helpful to expand on it within the second edition.
I suggest exploring further the idea of geometry being at the core of good public transit design – even in scheduling it is vital, but often overlooked. For example, a periodic schedule represented as a time and distance chart is like a frieze symmetry, with translations equal to the headway. Mirror symmetry is necessary to achieve pulse connections with local buses. In my area of interest, rotational symmetry is at the heart of berthing plans for high use ferry terminals, where one berth may need to be shared by more than one ferry line.
Symmetry is a constraint, but it is also an enabler of efficiency and usability for passengers. Understanding symmetry makes the planning process a whole lot easier.
I would attach some diagrams for further explanation if I could, but this web page doesn’t appear to allow it.
Robin. If you want to send me diagrams, click the envelope icon on the black bar at the top of the blog. That sends me an email.
Thanks!
Jarrett
I would suggest more discussion on ‘technology as tool, not goal’. There are currently a few paragraphs spread through the book, and mostly your message is: choose your goals, choose good routes, then choose technology to make that work. I agree 100% but it looks like I’m in the minority. With elected officials and definitely in the media what I see is technology driving the conversation — “my residents want something fast and reliable: we need (rail)(subway)(streetcars)(VegasLoop).”
Perhaps this fits under your proposed chapter on elite projection. Elites don’t (often) ride transit and are even less likely to ride it as general purpose riders – so shiny things (technology) and speed (rail and grade separation) weigh super heavily into their thinking. Frequency, span of service, operating costs, and network structure – not so much. Unfortunately this is not a small little issue – arguments over technology seem to be the norm.
Yeah, I agree completely with this. I could even see a chapter (or part of a chapter) discussing the advantages and disadvantages of various modes of transit. If you are going to cover microtransit, this would be the place to put it.
I think the broader notion of *infrastructure as a goal* is worth looking at.
More recently, I’ve seen a few examples of expensive BRT stations being built that only have a 15-30 minute frequency service, or train stations being built in poor locations. They’re usually built on similar premises to technology-driven transit projects – i.e. induce development, seen to support transit. They are arguably better than vanity technology as frequency can always be increased, station access improved, and new development supported. But that doesn’t mean the opportunity cost involved is any less painful. An example below.
https://twitter.com/jedwinmok/status/1566892402945703942
Service-driven infrastructure is something that decision leaders miss when they only attend the ribbon-cutting ceremony. You can frequently have really useful service without major infrastructure; you shouldn’t have major infrastructure without having/planning really useful service to run using it.
I love that you emphasize frequency, noting that riders perceive a minute of wait time as being longer than a minute of in-vehicle time. But I’d suggest stressing that frequency is so powerful that simply increasing frequency–even without concurrent increases in density–will increase ridership with elasticities of around 0.4-1.1. Or that density around most Calgary C-Train stations are no higher than Portland MAX stations, yet the C-Train achieved quadruple the MAX’s per-mile ridership, because the C-Train ran peak branch frequencies of 5 minutes, vs. the MAX’s 15 minutes.
It’s also important to note that frequency is relative–15 minute headways aren’t that bad for a 30-40 min commuter rail trip but are horrible for a 10-15 minute light rail trip.
In your fares chapter, it’d be great if you mentioned free transit. How free transit in Tallinn increased ridership but solely at the expense of walking and cycling while vehicle miles traveled slightly increased. How US cities usually can only have either free transit or more frequency, but not both, because transit operating funding is extremely scarce. How any cost savings from eliminating fare enforcement is greatly outweighed by the loss of fare revenues in all but the smallest systems. And how lowering fares won’t attract many new riders, since the full price of a monthly transit pass is far cheaper than owning and driving a car.
In your against specialization chapter, perhaps mention that commutes account for less than 20% of all trips (source: USDOT) and that these “commutes” excludes visitors traveling around town, recreational trips, errand runs, and traveling to school. Therefore, as powerful as job density is for ridership, don’t worry if your city has sprawling employment. Instead, be like San Diego, whose Trolley rivaled the MAX in 2019 per-mile and total ridership despite having far more job sprawl, less TOD, and zero airport rail link compared to Portland. The Trolley’s secret? Connecting the city’s two biggest universities, biggest malls, and the busiest border crossing outside Asia to Downtown, creating ridership that’s so bidirectional and consistent throughout the day, that in 2020, the Trolley doubled off-peak weekday frequencies from 15 to 7.5 minutes between Downtown and San Ysidro. Consequently, the Trolley has consistently enjoyed the strongest post-COVID ridership recovery of any US and Canadian LRT in APTA’s report, even before the Mid Coast Trolley opened or the free youth pass.
Great stuff, thanks!
Good stuff, Alex. I’ll hit all these points. The San Diego Trolley is a really interesting case.
*A focus on the impact of bus only lanes and signal timings
*Pulse Schedules, and if they are feasible for US transit agencies
*Autonomous subway transit (grade separated light or heavy rail, similar to Milan or Vancouver)
*Airport transit, in regards to updated federal funding rules
*The role of private intercity bus and rail (Flixbus, Brightline, & Vonlane)
*The decline of the corporate shuttle bus after covid
*Transit in a remote work world, everywhere and all day instead of the commute pattern
*Electric Bikes
*Electric golf carts for city use, as taxis and as personal transport
If you cover microtransit, you might consider how it evolves into fix route service the more you use it. The easiest way I know of explaining this is to think of “airporters”. These are vans to and from the airport, which often involve making several stops along the way. Eventually they become more and more like fixed route, fixed time service.
I like to think of the van leaving the airport. To begin with, you want to serve several people (otherwise it is nothing more than an expensive taxicab service). You want to pick up people from the same area, rather than going all over town. Ideally it covers one major corridor. That way, riders who are dropped off first do not delay those who are farther away. Deviations from that corridor take time, relative to the total time of the trip. It is nice — if not essential — for those with lots of luggage (in a sense that is what you are paying for) but it is costly. It is cheaper to just have the van serve the main corridor. This is essentially a fixed route line.
Then you get into timing. It is the same idea. One option would be for the van to leave when it is full. The problem is that riders who got there early would have to wait a very long time. You could try and time everything (based on demand) but flights get delayed. A rider who is scheduled for the 4:00 PM van to the north end will be disappointed if they have to wait hours for the next van. Thus it makes sense to just run at a regular basis. Now you have a fixed route, fixed time system. It is essentially a hub and spoke system with the airport in the center.
One other thing I would emphasize about microtransit is that it isn’t more cost effective, even if we develop self-driving vehicles. Self-driving buses would save a huge amount of money, while costing less to operate. A fleet of buses with the ability to hold a thousand people is much cheaper than a thousand cars.
Microtransit is at best a highly subsidized system for low density, or special use cases. It is easy to assume that by making it “smarter” it somehow leap-frogs other systems, but that simply isn’t the case. The problem isn’t lack of information — it is more fundamental than that.
I also think this is a great example of how transit scales. As ridership increases, the system becomes more cost effective and better for riders. Microtransit could be used in a very low density area (although even then, the Swiss have had great success with a pulse system — https://www.sciencedirect.com/science/article/pii/S0967070X16301469). Either way, you eventually move to fixed routes running more often. And then even more often. Eventually you move to larger vehicles (on rail) which saves money, which then goes into improving frequency in other parts of the system. You dig tunnels, or run the vehicles above ground, to be faster. Not only is this better for riders, but it saves operational time. You move from a hub and spoke system to a grid. This means more transfers, but with the increased frequency, the transfers don’t have to be timed, and are not much of an imposition. Riders who aren’t headed to the main hub save a considerable amount of time, while the agency saves money as well. Transit scales.
Great idea, thanks!
Thanks, Ross. We’re on the same page about microtransit, as you know, and I agree with your analysis. Thanks for the link to the article on Switzerland. I knew of that example in general but it’s great to see this detail.
In one of the introductory paragraphs, you might consider emphasizing that transit often involves trade-offs. Ridership versus coverage is the obvious one, but there are plenty more that permeate the book. The same goes for microtransit, or modal choice (as listed above). I think people discussing transit often ignore this, and think there is a simple solution to the problem, when it is more about the various trade-offs.
I don’t think you mention “spines” in your book. I think this is an important idea. It is based on branching (which you do cover). I see it as a network pattern — a variation on hub and spoke. It makes sense for cities that are shaped like a bottleneck, with strong demand in the center (Seattle is a good example). As you mention in the book, most systems are a hybrid, containing elements of these patterns (Seattle is slowly building more of a grid).
Overall, I think the additions you have in mind sound great.
A few thoughts/suggestions off the top of my head:
1) Perhaps as part of your section on “why transit matters”, you could bring in a bit more about transit’s potential role in helping to address climate change. Many places around the world are needing to completely transform their transport systems to achieve emissions reduction goals – transit is clearly part of that, and it could be good to highlight this a bit more.
2) A bit more about the benefits and challenges of “through-routing” services from one side of a city through the centre and out the other side. This can lead to long routes (which can be unreliable) but also potentially reduces bus numbers and required infrastructure in the middle of cities where it can be hardest and most costly to provide.
3) A bit more about the role of rapid transit services in an overall PT network. This could cover things like confirming when you might need to make the “jump” from standard PT to rapid transit, as well as subsequent processes around choosing the right mode and sequencing implementation of corridors and the wider network.
“”3) A bit more about the role of rapid transit services in an overall PT network. This could cover things like confirming when you might need to make the “jump” from standard PT to rapid transit, as well as subsequent processes around choosing the right mode and sequencing implementation of corridors and the wider network.””
Agree. Would love to see more on this in the book and/ or in blog posts.
The climate change argument for transit is tricky. At a minimum, you have to weigh the emissions avoided vs. the carbon emitted from the buses, themselves. This ties right into the classic ridership/coverage tradeoff, since in order for a bus to be avoiding more emissions than it creates, it needs to be well-ridden, hence a “ridership” route. Whereas, a bus that just winds around some suburb getting 5 MPG with 2-3 people on board might actually not be any better for the climate than if everyone on the bus had just taken Ubers. That doesn’t mean that coverage routes shouldn’t exist, it’s just that their purpose is to provide a service for people who really need it, many of whom “just take Uber” would be a financial hardship, not to fight climate change.
And, of course, even for routes that do avoid more carbon than they emit, from a pure climate change perspective, you’d have to consider the opportunity cost for what else the money used to operate the bus could be doing to reduce carbon emissions. For instance, does running a popular bus route more often bring in enough new riders to reduce more carbon than spending the same amount of money on solar panels or heat pumps? Maybe, maybe not.
Overall, I think transit systems are certainly very important to have, and while they can help some in reducing carbon emissions, I think their carbon impact, at least in North American cities, is always going to be relatively marginal. I think it’s better to focus on non-climate related reasons for running a transit system, such as freedom of movement being a right for everyone, not a privilege for those with enough money to afford a private car or taxi rides.
If you have anything to see on the point, a geometry-driven discussion of the likely impact of autonomous vehicles.
It seems to me that the supposed “killer argument” against transit is shifting from “people don’t like it” to “self-driving electric cars will make it obsolete in 10-20 years time, so we shouldn’t be investing now”.
How about a section on how American infrastructure costs are much higher per mile then the rest of the world and ways we can reduce it?
That issue is in the more capable hands of folks like Alon Levy, but I’ll try to find a place to refer to it.
More information on rapid transit specifically. Your discussion of branching divides frequency is critical, especially for how it can limit frequency on rapid transit, but could be expanded. Much serious discussion of transit comes up around planning for rapid transit in a region — light rail, subway/metro, bus rapid transit — that it is important to discuss that aspect.
Further, this might be beyond topic a little, but discussion of at least connection with intercity transport should be included. Maybe also discuss intercity rail and bus service, airlines, and high speed rail. I can respect not getting too much into intercity transport, but at least talk about how to connect urban transit with intercity transport, such as services to train stations and airports.
Further discussion about the differences and distinguishing of local and express/’limited-stop’ services, particularly bus.
Caveats to density – on how even suburbia has sufficient density to justify decent 10 min bus service given decent rates of use.
At least where I’m from (Melbourne, Australia) the “quarter acre block” (or 1000m^2) is not seen – where that one hectare would have only 10 houses if every house had a quarter acre (0.1 ha) plot, almost every suburb has at least 20 houses a hectare.
Much of the new outer suburbs is also built to the same density, or even slightly denser than various middle suburbs today (that were once the fringe in the 1950’s, 1970’s etc). Admittedly, I don’t know if houses in new suburbs are getting closer together in the US, but I am suspecting it might be happening with housing affordability getting worse (very much a problem where I’m at too)
This is important is that density is often misused as an excuse for poor public transit patronage when it’s a failure to provide busses more than every 40 minutes.
Importantly, it also means that driving up density (medium density has its own merits) will not automatically help increase transit ride share if services aren’t improved, which isn’t a given (building out suburbs before adding the train station).
This also gets knotted into messy definitions of density – does the city include jus the central bit, or the suburbs too? Are there large greenbelts or other natural spaces that skew the jurisdiction’s density lower – giving an erroneous low density for areas where people live (rather than in a desert)
Either way, acknowledging that cities do not need towers, or can even have single family housing AND frequent viable transit (though density itself can be promoted for other reasons, such as sprawl reduction)
Sounds like you’ve read Paul Mees! Yes, there will be more in the next edition about necessary and sufficient land use.
Hopefully not just land use, but also street form. I’ve noted for a while (not here yet though) that suburbs are mostly dense enough for good transit, but the long winding streets to cul-de-sacs means you can’t get a bus to within walking distance of most people. The streetcar suburbs in my city feature larger lots than the latest developments, going in right now – but the latest developments put most people out of walking distance of anyplace where you can get a bus.
You should also point out that planning developments around transit would be good. I have been watching a new development go in near my house. They dug up enough dirt to put in a 6 rail subway, put in one water pipe, then filled it back in. A little vision and only a little more money: that development could have had great subway service. Instead it will have no transit, and the form means it never will.
Well, true, although a working subway is vastly more expensive than just digging the hole.
Ideas to explore:
Does electric bus BRT have a higher benefit to cost ratio than LRT?
How should transit agencies pivot now that a large chunk of the primary customer base, choice riders at peak times, are gone?
In which direction does transit access move the needle on real estate? We used to discuss value capture for (some) TOD; has that flipped to a subsidy?
A couple of concepts that I’d like to see become pervasive in urban planning, including mobility …
– how inclusive design, from the start and throughout, is more successful and less expensive than accommodations after main design
– how to choose / design / build near-term solutions that can be groundwork for and not barriers to long-term solutions
I think the section on Laguna West in the original book was quite salient and prescient of a lot of developments that would continue to take place throughout the 2010s. It would be interesting to expand on that through more linearity/density examples. Examples of low-density, unwalkable sprawl are becoming more understood by urbanist-oriented decision-makers, but I think it’s still a learning exercise for many to understand the relationship between transport and housing outcomes. High-density walkable vibes are often enough for people to think they’re supporting urbanist outcomes when they ultimately hinder transit planning.
Others have mentioned it here too, but expanding the book’s rapid transit discussion would also be quite useful as this is increasingly dominating (and sometimes sidetracking) debates over improving transit. I’ve found that decision-makers often jump to grand conclusions about technology/priorities/etc, ahead of simply iterating on things like existing frequent service.
This is what I’m thinking about today:
1. Politicians are all excited about making transit free, but a product that doesn’t work for you when you have to pay is not suddenly going to work for you when its free. Which is better for a low-income person – extending a bus route so that a $2 bus fare replaces a $20 Uber ride or making a bus that doesn’t take you to your destination free instead of charging you $2?
2. Related: places like Indianapolis are all excited about their new BRT lines while conveniently overlooking the fact that so much of the built-up area of the city does not have any bus service at all. It’s sexier to have a new rapid transit line covering 1% of the transit service area than to have a basic network of bus routes covering 90% but leaving so much of the area without service is doing a disservice to people who need transit. Props to Phoenix and Las Vegas for at least trying to cover their ever-expanding sprawl.
3. To me, the increased interest in equity, as expressed by, amongst other things, the book by Paul Comfort mentioned in the post above suggests a recent move back towards the “coverage” end of the patronage-coverage axis. Are we finally going to listen to the people who actually need transit rather than envision our well-paid government employee car-owning selves taking transit a handful of times per year to “exotic” destinations like the ball game or the airport? If the answer is yes, then I say “great”. It’s not an either-or, either – well-designed “coverage” routes can easily grow to be high patronage.