A report from the TransitCenter has discovered something that's obvious to transit riders but not always to our urbanist elites: Transit succeeds when it is fast (in terms of total trip time and reliable).
While we know this from the actual human behavior we call ridership, it's also nice to see it confirmed in people's conscious thoughts, in the form of surveys. Actual behavior is a better signal than surveys when the two contradict, but when behavior and surveys agree, the survey adds something useful: a sense that people are not only making certain choices, but are conscious of those choices and able to discuss them. That, in turn, is good for intelligent policymaking.
They commissioned a study of 12,000 people across the country, representing large cities from each of the major regions, asking them a variety of questions on their attitudes towards and use of transit. While the document is full of interesting insights, this chart is really key:
Regardless of the age of the respondent, the two most important factors for choosing a travel mode: total travel time (walking/cycling/driving + waiting + riding), and reliability.
The table below reports on a variation on the same question:
Look at the top 6 responses to the question of "I would ride transit more if…":
- it took less time (travel time)
- stations/stops were closer to my home/work (travel time)
- it was clearly the less expensive transportation option
- the travel times were more reliable (reliability)
- there were different transit modes available
- it ran more frequently (travel time and reliability)
Four of the top 6 responses to this question concern travel time and reliability. How quickly can I get to my destination? How sure am I that the trip is going to work consistently? And how much does it cost? These are the same questions we ask when choosing to drive, or picking a commuting route.
The low ranking of frequency is not too surprising, because it simply shows that people are reacting to total trip time and reliability, and are not all that fixated on specific elements of that total. The problem with all surveys of this type is that they are divorced from the actual math of how optimal travel times are achieved for the most poeple. When this is factored in, frequency is paramount in delivering the desired outcome. On the other hand, shorter walking distances (item 2), if defined as a goal, is deterimental to total travel time because it implies lower frequencies and thus longer waits.
Note what isn't important in the respondents' assessment of their own mode choice factors: "nicer" vehicles, "more comfortable" seats. When it comes to how they make actual choices in their actual situation, these are far less critical than whether the service is useful. Obviously, people have minimum quality standards for these things, but these data suggest that they are finding those minimum standards to be met by the transit they see around them in the US.
Given the relative unimportance of even the most basic civility features like cleaniness and comfort, what importance can be ascribed to the even more optional features such as romantic vehicles or public art? (Not questioning the larger value of these things, but only their relevance to people's mode choices.) This data needs to be faced by anyone who argues that we should have quantitatively less transit so that shelters can be cuter, or so that our experience will be more like Disneyland, or simply because — in the interests of "place mobility" – people should want to travel more slowly.
What matters in transit ridership? These are the big ones: Travel time. Reliability. Cost. If you want to get people out of cars and into transit, start there.
I’m beginning to see why the urbanist “flaneur” types tend to underappreciate speed and frequency at the same time as they overemphasize the luxuriousness of station/stop/vehicle/tech amenities (the Disney model):
If you don’t think frequency and speed are that important, then in a weird way it makes sense to clamor for high-art transit shelters and high-tech vehicles (sexy pods on rails) because people will be spending (wasting) more time waiting *in* them and traveling *on* them.
If you prioritize frequency and speed, then shelter design and vehicle luxuriousness become less important because you won’t be spending as much time in and on them: If I can show up at the bus stop with the expectation that I’ll only have to wait a couple minutes before the bus comes, then I don’t need a high-art shelter to assuage me, because I won’t be waiting there long enough to really notice it.
NYC subway is hardly sexy, but most people don’t care because they’re not lounging around forever at the stations or on the trains. And prewar streetcars often had no formal shelters at stops because their frequencies were so high you barely even had to wait at the curb – not so for today’s spartan-service lines.
Ok, a small peep defense of urbanist flaneurs. In terms of convenience, if not “speed”, higher quality service trumps cute shelters, we realize that. Please note how Jeff Speck, for example, tweets about service disruptions & inconveniences on WMATA on the one hand and, on the other, regularly boasts his morning wake-up time to airport arrival time using WMATA. We have been getting a whole lot better as a lot (thanks to Jarrett).
You may pile on us, but please think about the ways urban designers can help you. I know that good network service is also an urban design challenge. Right? Let’s be a team.
For me, an important challenge is closing the gaps that exist at the connections of a high frequency grid, making the place around the connection safer and more convenient for pedestrians. There’s great potential to think a lot about improving that experience holistically with great placemaking strategies that also provide convenience. In fact, a creative approach to addressing solutions for this, improving transit and connections together as a strategy for capturing mode share and creating economic opportunity in places transit could help activate, could be a perfect challenge to address for the Knight Foundation’s City Challenge. http://knightcities.org/ I’d be eager to join any team willing to collaborate on a proposal.
I just happened to be looking at the report below after reading your article. Jump to 10.3 for a similar survey of transit attitudes for residents living on corridors that are relatively well served by transit. It is interesting that crowding (somewhat related to frequency) and cost on significant factors.
According to the survey most users over 30 want the impossible: high speed and stops located everywhere they might want to go. To achieve the former one must sacrifice the latter. High speed demands few stops meaning that it’s quite likely the train/bus simply doesn’t stop where you want it to.
To achieve high speed one must also sacrifice access time. When stops are farther apart passengers have to travel farther to reach one. When stops are in a dedicated lane passengers have to wait at crosswalks. When the system is grade separated passengers have to climb stairs or wait for elevators.
Designing a transit system that will appeal to the maximum number of people is complicated. Adding one more stop may bring in another 100 passengers, but may increase travel time for 100 others by enough that they choose to stop using the service. There is no single right answer, even when everyone you serve is young and fit.
Trying to serve the diverse needs of the entire population is much more complicated. You have to account for people who will walk 1200m and those who can’t walk more than 100m. You have to serve those who have no other means of getting around while still managing to appeal to people who own cars. You need lots of seats for seniors, the disabled, the pregnant and those who simply won’t take transit unless they can sit down. At the same time you must strip out seats so others can bring their scooters, wheelchairs and strollers, and so the vehicle can fit more standees at rush hour.
Posted by Bregalad: “According to the survey most users over 30 want the impossible: high speed and stops located everywhere they might want to go”
You can achieve something pretty close to this with a multi-tiered approach – high speed limited-stop backbones combined with local feeder services.
I was debating about the same thing with a Facebook friend. He says that people would prefer a line that have buses with A/C, 2+2 high seats and Wifi but low speeds and a not reliable headway over a punctual and quick line without any extra amenities. I think exactly the opposite…
BTW, are there any similar studies in other countries?(preferentially in third world countries)?
@Bregalad – As a person approaching 70, I sure do appreciate closely-spaced stops when my arthritis acts up. In this respect, I discovered that Los Angeles had a good approach to the problem on their heavier routes. These routes had both a local and a “rapid” service operating in parallel. With a pass in hand, I could (and did) board at a local stop and transfer to the rapid on the same route at the first common stop. Then I took the Rapid the rest of the way.
Of course, less heavily used routes only had the local service; but they also did not stop at every stop as there was often no one needing to board or disembark at those stops.
I am surprised by the lack of any mention in the report of the availability and cost of auto parking as factors in mode choice. (Except for single reference to parking at stations.) Parking is a huge factor in mode choice decisions and it is directly related to the comparable speed and cost of transit vs. driving.
John, the survey does mention parking, and it receives very low priority. That’s because most transit riders do not care much about parking because they walk to the station or bus stop.
Parking is not a huge factor in mode choice decisions except maybe for a relatively small number of rich commuter rail riders.
Sadly, too many transit agencies ignore the needs of the majority of their riders and value the opinions of that minority of rich commuter rail riders and waste many, many millions of dollars on destroying transit-adjacent land for parking lots.
That’s how the MBTA ends up spending nearly $100 million on two parking garage palaces, while the actual train service runs into the ground, neglected and rapidly decaying. And then they wonder why ridership is declining.
It’s pathetic that a major transit agency does not understand the basics of ridership generation: frequency and reliability. To focus on parking is completely irrational — but sadly, it is also a typical American approach. Pure incompetence at work.
Hello Matthew, Parking does count in bus only towns, here is survey from Waikato region Council in New Zealand, for city of Hamilton.
For existing bus uses travel time does not look to be important, non bus uses do see time as important
When asked: Way they take the bus – (1) No parking worries – (2) Save money – (3) Have no car
When asked: What stops you using the Bus. – (1) More convenient to use my car – (2) Timetable doesn’t suit – (3) Preferred to walk or bike (Hamilton’s urban area is about 100 sq km which is similar central Paris).
Interesting congestion is seen as least important.