Barriers to entry and the path of least resistance
Many people will volunteer that taking transit or eating locally is a good thing to do. But yet, people do not undertake sacrifices in droves to make “right” choices, unless there is no alternative. If we aim to change the choices people make, it helps to have a good way of thinking about the decision-making process. Here’s my way, which I find useful as a general conceptual framework:
People generally follow the path of least resistance given their situation and reject other paths due to higher barriers to entry. Moreover, the pool of resources for overcoming barriers is limited.
To be more precise, what matters is the perceived resistance and barriers.* We can be influenced by culture or limited information to perceive a choice as being more or less difficult. Though I think it takes a great deal of culture to get us to willingly refrain from littering, or to get us to recycle — and even then, we need those options to be easy.
Barriers to entry are anything that comes between the path of least resistance (or between no choice) and another choice. These could be a lack of information, difficulty of usage, discomfort of use, price, how long something takes, etc. — and there is both commonality and individual variation in the factors. The current path of least resistance may well have had its own barriers to entry.
Take the example of driving. At some point, the motorist had to learn to drive and had to buy and register a car. Following the framework suggests that overcoming those barriers actually was the path of least resistance, e.g. lacking a car severely restricted mobility. But once the car is in your driveway, the barriers to entry are few: the perceived marginal cost of gas/maintenance/depreciation (low), the price of parking (generally free), the price of the roads (generally free), and congestion to face (now we might be getting somewhere).
The view offers some specific guidance to changing choices:
Remove barriers to entry: provide frequent transit service for which one doesn’t need a schedule, let people use their cell phones or credit cards to get on transit, eliminate marginal costs through universal passes, get rid of helmet laws for cycling, make available bicycles that are easy to use in all conditions.
Put up barriers to entry: increase congestion, increase (or make more obvious) the marginal cost of driving — motorists won’t take into account the costs of highway infrastructure or parking if they’re never faced with them. Make roadways more narrow, thus decreasing the speed of least resistance (instead of hoping that speed limits will work).
Take advantage of the barriers that already occur, such as the large initial barriers to driving. Provide sufficiently good alternatives to car ownership — transit, cycling, walking, carshare — that it simply never makes sense to take the plunge and buy a car, and maybe not even to get a driver’s license. It’s much harder to get people not to use the car they own than it is to get people to continue not owning one.
Demolish false perceptions of barriers to entry, through education, good information and maps, and marketing.
The overwhelming implication of the perspective is this: to change people’s decisions, focus less on convincing and more on making the choice easy. Instead of attributing choices to something intrinsic about the person, or blaming the person for not doing enough, start changing the path of least resistance.
*To be even more precise and general would require adding in a notion of utility.
Another perspective on the Iron Horse Trail
This post is cross-posted to the TriTAG blog. Go there to leave any comments.
It can be insightful to take another perspective on something we’re used to. Yesterday I walked the length of Kitchener-Waterloo’s Iron Horse Trail and photographed it from its most common vantage points — the roads crossing it. There is little immediately evident in these photos, but I will explain below. Read More…
Don’t be fooled by middling transit frequency
Most bus routes here are infrequent, with buses no more often than every 30 minutes. What would count as frequent? I’m sure the majority of those who don’t use transit think every 15 minutes is frequent. But I’d bet most would change their minds after going to a bus stop and waiting 10-15 minutes (or more) for such a “frequent” bus. In the abstract 15 minutes isn’t much time, but at a bus stop it’s forever. Which is why truly frequent service is every 8 minutes or better, which doesn’t allow much time for getting annoyed with waiting.
If using schedules or trip planners, any frequency increase certainly improves the transit experience. But though we are tempted to think of 15 minute service as being frequent, it just is not good enough to be painlessly used without schedules. The difficulty with middling headways of 15-20 minutes is that while it seems like one should be able to use the service without a schedule, in practice doing so is not advisable. Even schedule-using riders may be more lax following the timetable than with avowedly infrequent service, and so more likely to just miss the bus — leaving them to wait the full 15 minutes until the next one.
Transit agencies, politicians, and transit advocates should call a spade a spade, and only refer to transit services as frequent if they can be used painlessly without a schedule by most people — and not just the transit-dependent, either.
Bold lines: Guaranteed high-frequency bus routes
I’ve written before about the complexities and uncertainties of bus systems. Here I outline one simple way to make transit networks more accessible to riders and would-be riders: guaranteed high-frequency routes. Creating such routes and marking them in bold on maps makes clear what portion of a system is accessible without a schedule, making possible spontaneous trips and more natural transit use.
Lines should be marked in bold on route maps if they run (for example):
- At least every 10 minutes Monday through Saturday from 6 am to 10 pm and Sundays and holidays from 8 am to 8 pm;
- At least every 30 minutes the rest of the time.
Express routes should be clearly delineated from other routes on the same corridor. Branching is okay, but only sections of a line meeting the frequency criteria may be marked in bold. Streetcar lines, if any, should meet bold line standards, perhaps being drawn in a separate color for clarity. Maps should list the service frequency for bold lines, and bus stops along bold routes should clearly indicate their status as such.
Currently there is not a single Grand River Transit route that qualifies. Route 7 mainline is pretty close to bold between King & Ottawa and King & University, but lacks night service. On University between Westmount and Weber, routes 8 and 12 could together be close as well.
Bus routes necessitating a schedule are only accessible to committed bus riders, and are unfriendly to casual users. A guaranteed high frequency on selected routes makes those lines easier and more pleasant to use for the choice rider as well as for the regular transit user. Minimum nightly service assures users that they will not be stranded, which encourages use during all hours.
Transit systems without bold lines should try implementing a small network of them. And systems that already have lines that qualify should be making a big deal of it. Simplified pocket-size maps ought to be freely available to show the transit-novice and the transit-averse which routes are easy to use.
It is difficult for low frequency service to spur enough ridership to “justify” high frequency service; it’s an uphill battle of incremental service increases in tandem with small ridership increases (and sporadic service cuts for good measure). Bold lines allow a transit network to pull itself up by the bootstraps through strategic allocation of resources into a network structure that is qualitatively different and more accessible to riders.
Contradictions in light rail opposition
I just sent in a letter to the editor at the Record:
It is insightful to contrast the two main complaints about the recommended light rail proposal.
Some say that we shouldn’t build expensive and inflexible light rail, that we should spend less money on an expanded bus system. They claim buses are just as good as trains, but are cheaper and more flexible. Well, to say nothing of the positive impact of a visibly permanent route, I will note that people in the real world vastly prefer trains to buses.
Just look at the Cambridge residents angry that they’re not yet getting the train! They’re upset because they’re getting left with only buses, and it speaks volumes about people’s true feelings about them. Those who have the choice will continue to avoid buses, but it is precisely these people who we need to be enticing out of their cars.
I, for one, would prefer seeing trains to Cambridge sooner rather than later. However, if we don’t build light rail at least in K-W, no one will leave their cars, and all of us in the region will bear the resulting costs of sprawl and roads.
Reality-based transit planning
Discussion, planning, and design of transit systems that aims to get people to use them must be based on what people actually do rather than on what they should do. The more in line with people’s actual use a transit system is, the more useful and economically efficient it is.
It is irrelevant how far people should want to walk to get to transit. The reality is that they are not willing to walk far, though the more complete reality is that they are willing to walk further to get to a train than to a bus.
It is irrelevant that people could plan around infrequent service. Very few are willing to do so.
It is irrelevant that people should take the bus to get to where they need to go. Either they will or they won’t, and empirical evidence all over the world says that they will do everything in their power to avoid buses.
This is why I have been talking about perception and simplicity. Because at the core of a transit system are people who have to navigate it using their faculties of perception and memory (and more), and who also do not derive equal pain or pleasure from the various modes of transportation. Paternalistic transit planning based on what people should do is self-defeating.
The uncertainty of buses
A lot of the issues with regular bus systems can be explained as uncertainty:
You don’t know which route you should take, where to find that bus route, when the next bus is supposed to arrive, how early or late it will actually arrive, and where to get off. The only thing that is generally certain is that you’d prefer not to deal with the uncertainty.
Why the Rapid Transit route should not be split
As promised in the last post, I will discuss the issue of the splitting of the route in downtown Kitchener and Waterloo that was recommended by the Region of Waterloo report. I understand the motivations, but I think it is a very short-sighted move and that the region would be better served by the alternative of routing both directions along Caroline in Waterloo, and along Charles in Kitchener.
Part of the appeal of a light rail line in the first place is that it is simple. Fixed stations, trains going both directions, no schedule needed. But the recommended split route segments have stations on them, which complicates the route. Having stations on different streets for the two directions would make giving directions based on station location more difficult, and it would also create strange disparities in ease of access to transit based on the direction of one’s destination. In other transit systems where the directions go along separate one-way streets for much longer stretches than proposed here, the split is easy to understand and deal with, but here it is a very ad hoc measure. This is not a huge issue with the proposal, but it adds needless complexity and costs without the benefits to compensate.
Waterloo holds festivals that require the closing of King Street to traffic. The proposed way to deal with having one direction going along King Street is to double-up on the one Caroline track during any festivals. This will be difficult to do in practice, will require a serious amount of planning, and will include the real possibility of failure — which might not prove dangerous, but would certainly bring the entire system to a standstill. Additionally, splitting the route here is incredibly short-sighted because of the impossibility of single-tracking when (not if) trains will be running at a very high frequency.
King Street in Uptown Waterloo is already a fairly reasonably developed area and attraction, especially with the newly-opened public square. It is one street over from Caroline and no trouble to get to. On the other hand, Caroline is surrounded by parking lots that are prime for transformation. This is going to be happening very soon in the case of the Balsillie School, which will be next to CIGI. But it will take a fair number of years after trains start running before Caroline Street gets fully built up, and until then King Street will continue getting the lion’s share of visitors. People will go to King Street and the public square, and probably enjoy the lack of a transit line separating the two sides of the street. And of course, leaving trains off King would allow the future conversion of that portion of it into a pedestrian mall. Side benefits of not having any part of LRT along King in Uptown Waterloo: more parking left on King for those who swing that way, and no need to have LRT go against traffic on Erb.
One more point about Waterloo: apparently the Catalina town houses (bounded by William, Caroline, Allen, and Park) are an important part of why the proposal has only one track on Caroline. Such influence should not be accorded to a non-historic development of perhaps 20 units. For comparison, the Bauer Lofts, also on Caroline, are nearing completion and will have 89 units. This is the right scale for the area, especially noting the adjacent Sun Life building and the proposed 144 Park condos. [Update: This 143-unit high rise has been approved.]
In Kitchener, the proposal has one direction along Charles and the other along Duke. This is even more problematic from the standpoint of complexity because those streets are not even within sight of each other. The two stations would be far away, and this would present issues with directions and orienting, and with the arbitrariness of business and residential proximity to only one direction of travel. It is far simpler to have both tracks along Charles Street, which is easy walking distance to anywhere else in downtown. Of Duke, King, and Charles, Charles is the one with the most opportunity for new development. And as in Waterloo, King Street and Duke Street will hold their own in attracting visitors.
I find the idea of businesses being seriously hampered by being out of sight of the transit route to be strange. Light rail is not a streetcar — people cannot get off at will. The main thing LRT does for people is facilitation of getting from station A to station B, a trip in which people are interested due to the surroundings of station B, and not just due to what they happened to glimpse along the way. Most importantly for businesses in the downtowns, LRT will bring many more people into the downtown areas from elsewhere, resulting in more than what the current business selection can handle. Worrying about direct visibility from the LRT is silly; if it is there, worthwhile, and walking distance from a station, they will find out and they will come.
As just one example, consider the Mission District in San Francisco. All the main transit lines — subway and frequent trolleybuses — are along Mission Street, from which most visitors make their way a block or two to Valencia, which has an entirely different street feel than Mission. Considering the distance between the successive 16th St Mission and 24th St Mission BART (subway) stops which service the district, the distance to Valencia is insignificant.
In transit, less is more
Following up my previous post on the role of perception in transit, I want to discuss the importance of simplicity of routes. In that post I mentioned the relative ease of using a bus that runs frequently. What that omits is the need to know where the bus actually goes. The purpose of most regular bus routes is to provide extensive service for an area, which generally means a winding route with many stops. What immediately follows from this is that it is difficult to be very familiar with more than just the routes you use routinely. This in turn makes it likely that you will simply not use the bus system for getting to somewhere unusual.
I must note that this is not always a predicament in bus systems. In cities with predictable street grids there can be bus routes that go straight along a particular street, or there can be bus routes that follow some important avenue. But inevitably there are common trips that do not follow the obvious streets, and so bus routes evolve to serve those trips.
The complexity of bus routes is a huge part of what makes bus transit a poor way to get people out of cars. A barrier to entry, if you will. If someone is reliant on buses entirely, they have no choice but to take the time to learn the route system, though they might still learn only the part that is relevant to them. Don’t expect people to willingly subject themselves to something difficult and esoteric if they can avoid it.
Though I won’t address it here, some complexity can be mitigated by properly designed information and maps.
The answer is probably some version of Bus Rapid Transit, a term that generally refers to individual lines with separate rights-of-way and express service. I believe bus systems should be holistically planned with BRT ideas in mind, in a way similar to subway systems. This entails simplified routes, limited stops as a matter of course, and maps that make it easy to understand both the overall route system as well as stop locations. Such a system could be parallel to the more comprehensive system. But it would be vastly more usable for most people.
This formulation exposes, however, a serious limitation of even a simplified bus system: the lack of permanent routes. BRT has little infrastructure and its vehicles can travel off route, so it still has this limitation, albeit to a lesser extent. The right way to think about the importance of unchanging routes is probably in terms of temporal simplicity and complexity. A simple bus system that changes over time becomes complex temporally. The effect is that people will still be wary of using the system to inform long-term decisions such as where to live and where to set up a business, as the bus routes could easily change and leave them out in the cold. Here the right sort of Bus Rapid Transit does offer a partial solution, but light rail and subways are and feel more permanent.
On the tipping point of buses
No, we’re not going to go around tipping over buses. Though it would make a great urban legend to counter that rural legend of cow tipping.
There are plenty of bus routes in Kitchener-Waterloo — and in most cities that are not New York, I would imagine — that run every half-hour or less frequently. On the other hand, the “mainline” 7 buses run every 7-8 minutes along their common spine. I claim that the difference between a 30 minute and a 8 minute headway is not just in quantity, but also in quality. This might be obvious.
The difference is in how one may use such routes. For 30-minute buses, on average you’ll wait 15 minutes; but sometimes you have bad luck and the bus is running late, so a worst case is something like 40 minutes. These are 40 minutes that you could have used to walk to your destination. Which is pretty bad, so you get yourself a bus schedule and plan your time around the bus stop time. If you don’t get there sufficiently ahead of time, you still run the risk of the bus arriving early and screwing you over. The obvious way out is to avoid all this hassle with some other means of getting from point A to point B, such as driving.
Why is an 8-minute headway different? The answer should be clear: if you need to take the bus, you go to the bus stop and wait. On average, that’s a 4 minute wait. Worst case, 8 minutes, which is manageable. You don’t need the bus schedule! If you know where the bus goes, then you can just go to a stop and expect the bus to ferry you to your destination. It’s easy, requires no access to schedules or the Internet, and especially requires no advance planning. (With a short headway, the schedule might actually give you a false confidence due to inherent variability in the route.)
The bus headway tipping point is probably somewhere in between 5 and 30 minutes, and is some kind of average of individual patience curves. Around that tipping point, more buses will lead to disproportionately more riders and a more user-friendly bus route. The greater point is that there is a serious qualitative difference between low- and high-frequency service, and this deserves more prominence in discussion of transit.