the same “empty buses” fallacy, over and over

This morning, Andrew Sullivan, whom I usually find intellectually engaging, featured a confused article about transit productivity from Eric Morris on the Freakonomics blog.  It's the old line about how because buses are often empty, they're not a very efficient transit mode.  I first rebutted it three years ago and the rebuttal hasn't changed at all.

I quickly wrote the letter below. But the big announcement is after the letter!

Eric Morris on the Freakonomics blog has fallen into the familiar trap …

To put my remarks in context: I’ve been a transit network design consultant for 20 years, and am also the author of the blog HumanTransit.org and the book Human Transit (Island Press, 2011) which rebuts many of the false assumptions in this article.

Morris's argument rests on the false assumption is that transit agencies are all trying to maximize ridership as their overriding objective.

In 20 years as a transit network design consultant working across North America, Australia, and New Zealand, I’ve never encountered a transit agency that pursues a ridership goal as its overriding purpose. Transit agencies are always required to provide large amounts of service despite predictably low ridership, for reasons including basic access for seniors and the disabled and the perception that service should be delivered “equitably.” While equitable is a slippery word that means different things to different people, its effect is to justify service spread all over an urban region, even into areas where ridership is inevitably low (usually due to a combination of low density and street networks that discourage walking).

In my own work, I refer to these predictably low-ridership servics as coverage services because they are tied to a coverage goal that conflicts with a goal of maximum ridership. Typically the coverage goal is stated in the form “__% of residents and jobs shall be within ___ feet (or meters) of transit.” This goal requires service to be spread out over areas where prospects for ridership are poor. I then encourage transit agency boards (or Ministers) to think consciously about what how their service resources should be divided between ridership goals and coverage goals.

If this method ever becomes common, it will be possible assess bus services that are trying to achieve high ridership. Only that universe of services is relevant to discussions about whether bus services provide ridership effectively.

A more extensive geometry-based discussion of exactly this issue, and how it needs to be managed in policy thinking is in Chapter 10 of my book Human Transit.

Regards …

The big announcement: I'm not going to do this much anymore.  Here is my response, but hey, regular readers, any of you could have written this, right?   After all, the rebuttal has been on this site for three years!  Could everyone please bookmark that, or bookmark this, and just send a link whenever you see this same argument?  Would save us all much time.  Thanks!

20 Responses to the same “empty buses” fallacy, over and over

  1. Jeremy November 15, 2012 at 9:37 pm #

    I also debunked this story from a different angle in Episode 8 of my podcast. No matter what the purpose(s) for your services, you will always have buses be empty near the terminus and layover spot, as well as variations in loads (most often caused by school crowds and other irregular travel patterns). And of course, even routes that are crush-loaded much of the day may be “empty” at night, but that’s not a good reason to eliminate overnight service (which is, by definition, a coverage service).
    It’s easy for the anti-transit crowd to cite empty buses to score cheap political points. It may become a political problem but it’s not a real problem.

  2. Dexter Wong November 15, 2012 at 11:03 pm #

    Quite right, Jeremy! I have ridden the bus that runs past my house from end to end and have seen just that. Over the years it went from 40-foot buses to 60-foot articulated buses but the pattern is still the same, empty near the terminals and full in the middle during the day.

  3. asdf November 15, 2012 at 11:39 pm #

    So, when your agency is considering cutting a coverage route in your area, how often do you have to use it for you to feel morally justified in complaining about the decision?
    Does that bus have to be your everyday trip to work, or if you use the bus to visit friends, is that good enough? If you ride it once a week or once a month, is that good enough? What if the bus is merely your emergency ride home in case you have a problem with your bike, but to date, that problem has never actually happened, so you have never ridden that bus?

  4. J Medley November 16, 2012 at 1:43 am #

    Odd. No one ever makes an issue of the paving of rarely used roads.

  5. James A. November 16, 2012 at 1:54 am #

    I saw that post as well, Jarret, but as is now my standard practice when confronted with things originating at Freakonomics (a site that trades in dazzling the sophomoric with pseudoscience), I declined to respond.
    More troubling to me was the fact that in the “Ctd.” part of the thread the only response included from a “certified transit expert” endorses Morris’ arguments and uses them as a rallying cry to support…(wait for it)…PRT.
    And as if advocates of proven transit modes haven’t spent enough years our lives patiently explaining that, while the semi-PRT people mover in Morgantown is all well and good, a universal solution for every transport need it does not make, we now have to sit back and watch the PRT fanatics unite with the automated car boosters. All the while ignoring the fact that a four lane road has four lanes — no more, no less — regardless of whether a human, a computer, or Gozer the Destructor is behind the wheel.
    Thanks for the link; next time I’ll send it in.

  6. Thomas November 16, 2012 at 2:23 am #

    Odd, but we never get complaints about empty roads.
    We have rural roads they carry few trips at all.
    We have roads all over town that are empty every Sunday morning at 6 AM.
    Even at morning rush hour, when traffic into town can be backed up extensively, lanes leading away from downtown are virtually carless.

  7. EN57 November 16, 2012 at 4:42 am #

    Odd, that many of Jarrett’s pronouncements on transit are hardly ever rebutted here.

  8. BBnet3000 November 16, 2012 at 5:42 am #

    From an environmental perspective, the issue of low-ridership buses is also How To Get More People On The Bus, which is also an issue of land use, pricing (of driving more than of the bus), and other things.

  9. John Haugeland November 16, 2012 at 6:04 am #

    To be honest, I don’t understand your rebuttal.
    This seems like someone saying “the size of the sun is X” and someone else yelling “BUT THE PURPOSE OF THE SUN IS NOT ITS SIZE.”
    It turns out that it’s valid to measure the efficiency of the bus system even when the designers’ goals aren’t efficiency.
    One particularly compelling case of when to do this is when the stated goals of the designer have seriously diverged from the practical needs of the larger community.
    You appear to be arguing for moral rather than practical goals. I am not saying you’re wrong, but I think you should probably recognize that moral goals as a sole design criterion typically lead to systems so inefficient that they fail to deliver the service quality an efficient immoral system would to the people meant to be defended.
    Pittsburgh’s bus system used to be designed this way. It was terrible. The second they stopped, everyone got much happier. (Then six months later the governor pulled the plug on funding, so nobody remembers the magical revolution that happened when the moralists and the economists were forced to take a back seat to some engineers for once.)

  10. Matt MSP November 16, 2012 at 7:46 am #

    Seems like people have mentioned it, but it’s true in practice too… People are more likely to take a bus during busy times when they know they can count on the same bus at slow times!
    It’s nice knowing I can take a bus home at 5:15 every day, but I could also take the bus home at 9:30 if I stay out late or work late, or I could take the bus on a Sunday if I decided to go to a Vikings or Twins game downtown.

  11. Wieseltierfan November 16, 2012 at 8:00 am #

    Sullivan has a weakness for things that sound contrarian. Like when he pushed The Bell Curve. He tends too much toward the confrontational, and away from the analytical. He spent several years attacking anyone who opposed the Iraq war, and then switched to attacking neocons wherever he could find them. Maybe he’s interesting on things where moral certitude and anger are good things (arguably torture) but when analysis is needed, he tends to be weak. Where moral certitude gets in the way of analysis (like mideastern politics) he is especially bad.

  12. Tom West November 16, 2012 at 8:00 am #

    I see a lot of cars in *rush hour* running at just 20% capacity. Such a waste of taxpayer-funded road space.

  13. djconnel November 16, 2012 at 8:06 am #

    I agree with John: none of this rebuts the efficiency argument. “Safety buses” themselves may not be inefficient relative to fuller buses, as they enable people to take the higher demand buses, but nevertheless the efficiency of the net system is what it is: so many $ and so many gallons and so many pounds of CO2 per passenger mile. And that efficiency is not good.
    You could argue there’s indirect benefits such as promoting denser development or whatever, but the reality is transit resources are limited, and using blanket moral-based arguments to justify grossly inefficient service could well end up serving nobody well, providing a system which is perpetually starved for support and thus for resources. This is what is seen in San Francisco. MUNI is virtually useless in many neighborhoods, because one can literally beat it with a brisk walk from door-to-door.

  14. Progressive Capitalist November 16, 2012 at 8:40 am #

    Since Title VI, ADA, and EJ concerns all require coverage, there needs to be a paradigm shift at FTA in order for many transit providers to truly transform their systems and attract more new riders.
    Ridership is obviously driven more by route frequency, stop attractiveness, and system connectivity. However, as gentrification occurs in urban cores, Title VI impacts are only likely to be magnified. So how do the defenders of market urbanism (which is truly the larger paradigm here) address social equity?

  15. Alex November 16, 2012 at 9:11 am #

    What’s so grating about the line of argument used in the Freakonomics post is that no one is arguing that the existing US transit systems are, in the aggregate, environmentally efficient. In fact I would guess that most people who argue for improved transit on environmental grounds (who is that, by the way?) are saying exactly the opposite. Most environmentalists I know (my local Sierra Club is an example, not sure about the national organization) argue for better land use policies first, and tout improved transit efficiency as a side effect. So Freakonomics is actually introducing a straw man argument that is unfortunately persuasive to people who are not normally inside the debate about land use, transportation, and the environment.

  16. Joe Busman November 16, 2012 at 12:39 pm #

    There are a few rebuttal points I think you missed. First of all, people don’t board at the very beginning of a route and ride all the way to the end. Invariably, the bus will not be fully packed from start to finish. Second, riders don’t schedule their bus trips like airplanes, so buses must run on a regular schedule and sometimes everyone chooses to ride at 1 PM or 2 PM, it’s not like buses know when and can simply not run at 1 PM if they knew everyone planned on riding at 2 PM. Due to simple demand logistics, even airplanes have many empty seats. Third, the size of a bus is not a big cost factor, so if ridership falls midday, it does not make sense to switch out all the buses to short buses or even vans. If a car is 1/5th full (i.e., only a driver) it is not considered a waste of resources but if a bus is 1/5th full it is???

  17. Matthew November 16, 2012 at 1:54 pm #

    I did send Andrew a rebuttal but it has yet to appear … if it ever will.

  18. EngineerScotty November 16, 2012 at 1:57 pm #

    An ADA-related dead horse I’ll flog once again:
    Why should transit agencies be required to provide paratransit services (those cities who lack a public transit agency, generally lack paratransit).
    What if public works agencies–state DOTs, municipalities, and the like–were instead on the hook for this?
    Obviously, in cities with robust public transit systems, provision of paratransit could be contracted out to the transit agency, much as many urban districts contract with the transit agency to transport students, or private employers will buy passes for their staff. But paying for it would fall upon the municipality, not the transit agency.

  19. Didesh November 20, 2012 at 1:44 am #

    Coach hire does not just have to be for special occasions either, because the low price of coach hire from Cheap Coach Hire, means you can hire a coach for a big shopping trip, a school holiday treat or a special Sunday drive.

  20. Wanderer November 21, 2012 at 12:08 pm #

    On the relationship between coverage and environmental efficiency: The low ridership lines help feed the efficient, high ridership lines, thereby making them more efficient. For roads, we recognize an elaborate hierarchy feeding each other, and don’t say, “If it’s under 10,000 ADT, it’s toast!”
    Having a strong network of transit lines in time and space makes people more willing to reduce their car ownership. That of course leads to more transit trips and fewer car trips.
    I would also think that being able to create locations served by multiple transit lines would be more encouraging of more transit use, less car use, and more transit-oriented development