Photo by Oran

When I wrote up all of the Transit Master Plan’s High Capacity Transit corridors, many of which the Council has killed for now, I tried to make the point that there is no one super-metric that is arbiter of which corridor is the most important one. On the other hand, I spilled many words explaining one metric, ANC/NR, that may have given people the impression that it was the truly important one.

ANC/NR is basically a measure of how much you’re spending to get another person to ride transit on one of their trips. While that’s a really important consideration, a simple thought experiment illustrates its limitations.

Imagine there are two lines. Both have only two stops, at each endpoint. One is 1 mile long and will attract 10,000 riders. Another is 2 miles long and will attract 16,000 riders.* As they are technically almost identical, line 2 costs the same per mile, or twice as much overall.

More after the jump.

So which is the best to build? There’s obviously a lot of “it depends” since I haven’t specified the problem very well. A lot of people here would, holding all else equal, say “line 1”. And that’s a valid viewpoint; ANC/NR, for sure, would tell you that line 1 is 20% more cost-effective.

But would-be customers of line 2 are almost certainly creating more congestion and spewing out more greenhouse gases as they keep driving. So you could instead measure passenger miles. But at some point you’re inhibiting the growth of dense, non-auto-oriented neighborhoods and applying centrifugal force to the population. The really long trips are commuter trips, and before long you’re building a peak-commute oriented network.

And if you’re a line 1 partisan, we can tweak the numbers to your breaking point. What if line 2 has 19,999 riders? Line 1 still has a miniscule edge in our cost-effectiveness, but do you still think it’s better? If not, you’re conceding that there are other metrics that enter the calculus.

Conversely, someone who prefers line 2 has to address increased ridership on line 2. If there are 30,000 boardings on line 1, that’s still fewer passenger-miles than line 2, but is it really worse?

This is not intended to endorse any particular line. It’s a way to clarify your thinking about what’s important when choosing between various alternatives.

* If you think 1 mile is too ridiculously short for a transit line, feel free to double the numbers.

** Athough the 4th/5th connector has some similarities to line 1 with respect to the other HCT corridors

27 Replies to “Riders vs. Miles”

  1. Since the future worker will inhabit a cloud more than an office, what is the 21st century role of transit?

    Like LINK, it seems like it will be to ferry people to the remaining centralized institutions such as sports events, as all commerce, retail, education and other resources for living life have long since been decentralized and are serviced by personal transit with door-to-door stops.

    1. There is of course no reason that digital interaction will be limited to office work, so the need to centrally locate business will change – though the random interactions that cities have historically provided still have value – which is why we see concentrations within urban areas of medical, creative, law, etc…

      That applies to other institutions such as sports (which most folks watch remotely already).

      With walkable neighborhoods, the personal transit that will be necessary/desirable will typically have two shoes, that you put on that morning.

      1. The need to centrally locate / cluster businesses is not going away at all. It’s been slightly ameliorated. But the fact is, name a manufacturing industry, it will concentrate. Service and entertainment and intellectual operations will continue to cluster, even if less so than before. Even agriculture clusters.

        Heck, even if you forget jobs, social activity clusters — people don’t want to do *all* their social activity on line and would rather live within transport distand of their friends.

      2. The “string of villages” model, with clusters of housing and entertainment and business linked by interurban rail, is a viable one. And Kent can be part of that. The sidewalkless car-dependent sprawlburbs which spread across the landscape in all directions, however, cannot.

    2. The primary model for organizing economic activity is the cluster. Take Silicon Valley itself: a primary proponent of the cloud, yet its creation and continued existence is due to things that are optimized by being restricted to a locale: relational contacts, physical supply chains, academic foundations, investment by local government, etc.

      The advantage of the cloud is in its ability to easily and quickly upscale/downscale, centralize expertise in specific technical competencies (so companies can offload that particular workstream and hire for other competencies), and to have all this accessible by the complete range of devices, from smartphone to tablet to PC.

      I don’t really see what that has to do with urban planning. The cloud isn’t the same as RAS.

      1. People will always congregate and they’ll always want to be close to one another. I don’t see since the Internet came out that the mass migration to the city has slowed down at all.

      2. Silicon Valley has long since been the center of it all.

        In fact, there is no more center.

        That has everything to do with planning…none of it necessarily “urban”.

        No one is necessisarily RASsing…because everyone needs to talk to everyone in the 21st century workcloud.

        The problem with “transit” is that it is too limiting.

        The worker constrained by travel can only do so much.

        The worker unconstrained by clouding can do anything.

      3. [John] I don’t think anyone has suggested that we turn off the Internet. It’s not an either/or: communicate online or in person. It’s both. Though you’re happy to work in your basement away from human contact, I’d argue that you are and always will be the exception.

  2. Love Oran’s picture. If only there were a way to get Link in the picture, underground (of course it’s really a bit east, but conceptually it’s right there).

    I prefer whatever line connects the most dense area to the most dense area (or potentially dense, in both cases). Removing car trips shouldn’t be the point of transit. Building a well-connected, walkable city should. (that said, I get your point)

    1. Right there with you.

      Another fun exercise would be to measure based on number of new commercial transactions (=jobs, taxes) that could occur due to new trips. Can you measure pent-up demand that is limited by lack of mobility and poor connections?

  3. This is why I dislike the use of passenger-miles when it comes to measuring transit. It assumes that moving people long distances is an end to itself, rather than something we have to do given current suburban development patterns. Even the use of “trips” can be tricky. People often point to transit’s low share of total trips, ignoring that drivers tend to make more individual trips and that transit users are more likely to live in dense neighborhoods where they can forgo a trip altogether and just walk to the store. Portland has started to promote the idea of the “Trip Not Taken” as a goal of public policy, so that rather than focusing on growing ridership as an end unto itself, the goal is to get people and destinations closer together.

    1. The most destructive double-talk metric that we see on a regular basis is the “subsidy-per-passenger-mile,” fed to a clueless media that reports it as if it were synonymous with “subsidy-per-passenger-trip,” in order to make suburban-commute transit look more competitive with real urban transit.

      You see this all the time with BART proposals, as subsidy/passenger-mile numbers are the only way to make it look competitive with other modes. (The subsidy/passenger-trip numbers are eye-popping.)

      The MBTA also uses throws out misleading subsidy/passenger-mile statistics when trying to build political consensus for its necessary but very costly commuter rail operations.

      BTW, I believe that “walking” usually one of the modes considered as part of the measurement of mode-share for trips. So having walkable trips factored in should not in and of itself reduce transit’s apparent measured share.

  4. All measures have some kind of implicit or explicit relationship to one or a certain set of conclusions and solutions. That is why no single measure is often “correct” when making a decision based on multiple objectives. That is why most studies have a large set of measures that correlate to a smaller set of objectives where are in turn are derived from a few goals.

    My favorite example is looking at the rate of vehicular fixed object collisions. This essentially measures the rate at which vehicle depart the roadway and hit a tree/building/pole/etc. If this is the only measure you use to decided where to make safety “improves” the only logical action from a roadway design perspective is to remove fixed objects from the clear zone (the area adjacent and parallel to the roadway), which effectively results in the highwayization of arterials. While this measure would likely improve safety in one regard, for cars that depart the roadway, the impact on overall safety, or the best strategy to improve safety depends on the situation is probably much more multifaceted than this one measure indicates.

  5. Realistically, among the most important metrics are the ones that get you more funding. For example, Central Link agreed to do an in-depth ‘Before and After Study’, as part of receiving $500 million from the FTA.
    They want to know what was predicted in five major categories versus what was actually accomplished after two full years of operation, which ended in this July.
    In a nut shell, here’s the five categories:
    a) Scope of the Project
    b) Service levels
    c) Capital Costs
    d) Operation and Maintenance Costs
    e) Ridership Patterns
    …incremental ridership (projected and actual)
    …origin/Destination patterns
    …farebox recovery ratios
    It would make sense for ST to continue reporting the metrics required for the Nat’l Database, as that’s what they will be judged against.

  6. Did the council actually say no to the HCT corridors, or do you just mean they weren’t included in the $60 fee for this year? We always knew it would take several years to find funding for all of them. But if the council actually said no to the plan, that would be a bigger problem.

    1. They can change it as they like later, but it appears that the other corridors are unlikely to even get planning money. It’s hard to see how things can move forward in that context.

    2. Well, the study isn’t even finished, and next year’s council might take up another corridor. We couldn’t expect the $60 fee to take up all five corridors when so many people have been clamoring for years about sidewalks and potholes and the car people want something from McGinn.

      1. Mass sidewalk programs could have very interesting results. People who don’t support bikes, buses, trains, may support sidewalks…. yet I have suspicions that the longer a district has had sidewalks the more popular non-private-auto transportation will become….

      2. Nathanael, a mass sidewalk program for Seattle would entail a lot more general tax revenue, or LIDs. Drainage requirements make new sidewalks very expensive; the Ped Plan’s Tier 1 projects come out to about $840 million, and we’re currently spending about $15 million/year total on all pedestrian projects, not just new sidewalks.

      1. That’s nothing compared to the piteous whining we’ll hear from your crowd after your electoral wishes get crushed in November!

Comments are closed.