Part 2: Why Democrats should like buses

By David Levinson and David King.

Today Democrats are associated with rail. The reason we hear from politically connected folks is construction jobs and unions and real estate development and property owners. Of course their more urban constituency prefers rail to roads, while higher densities fit with their urban ideal. To the extent that Democrats have an underlying principle of “Equality” and “Social Justice”, they should support Buses.

Buses are S.M.A.R.T. in Starkville Mississippi
Buses are S.M.A.R.T. in Starkville Mississippi

Why Democrats should want to prioritize improving buses

  • Buses serve more people than trains ever will0 (since they are more cost-effective1), so bus improvements benefit more people.2
  • Bus riders are much more likely to be Democrats since they have lower average incomes compared to rail users and the general population.
  • Buses generate more operating jobs than trains, as bus drivers are labor and buses don’t carry as many passengers as long trains.
  • Buses are harder to automate than trains, so driver jobs are longer lasting jobs. While there are fewer construction jobs than rail projects, those are short term anyway.
  • There are more manufacturing jobs per passenger.  Bus manufacturing is more likely to be local.
A Political Economy of Access: Infrastructure, Networks, Cities, and Institutions by David M. Levinson and David A. King
A Political Economy of Access: Infrastructure, Networks, Cities, and Institutions by David M. Levinson and David A. King

Political Parties, Three-Axes, And Public Transport

  1. Part 1: Introduction
  2. Part 2: Why Democrats Should Like Buses
  3. Part 3: Why Republicans Should Like Buses
  4. Part 4: Why Libertarians Should Like Buses
  5. Part 5: Why Greens Should Like Buses
  6. Part 6: Summary

0. [Updated 9/4/14: in the US, certainly outside of New York City, and probably including New York City. More than half of US rail use (2.4 B out of 4.2B annual unlinked passenger trips) is in New York City. Caveat, rail share is growing some nationally (which is not too surprising given the amount of investment in new rail infrastructure replacing bus service). As with anything, rail investments face diminishing returns, since the high benefit/low cost rail projects have been built.]

1. [Updated 9/4/14: See, e.g. [1] and [2] and [3] and [4] which look at capital costs per rider in the Twin Cities. Of course operating costs per rider are different, and a full train may be lower than a full bus. Neither is full most of the time. Whether lower operating costs offset higher capital costs is an empirical question and case-contingent, in most cases the marginal new rail line vs the marginal new bus line will net out to be more expensive overall with reasonable interest rate assumptions. None of this is to say rail isn’t (or is) a better investment than highways (or doing nothing/no build) at the margin, which is an argument for another day.]

2. [Updated 9/4/14: Presently, only in New York, Massachusetts, and DC does rail ridership exceed bus ridership (New Jersey is close). Of course those are the best transit markets, and rely on mostly early 20th century rail infrastructure (Boston and metropolitan New York). Nationally BTS reports bus overall has 52% of the transit market, and rail 44%. ]

8 thoughts on “Part 2: Why Democrats should like buses

  1. 1- Buses do serve more people… because most cities do not offer rail. That’s circular reasoning.

    2- Buses are not more cost-effective at all. Sure, they’re cheap to get started, but they’re extremely expensive to operate, in large part because, as you point out, buses require a LOT more labor to operate, and the majority of the cost of transit is labor, not gas, not equipment. That’s why Ottawa is converting its BRT to LRT, expecting to save tens of millions of dollars in operating costs each year.

    Granted, in American sprawl, there is just nowhere near the density to offer any decent transit service. Offering buses there rather than rail is just cutting losses on services that will NEVER be widely used anyway. Bus service there is used as a welfare program, cities are built assuming people use cars to go everywhere, buses exist only to allow people without cars to be able to get around (2-3 times slower than people with cars). And if you care about being cost effective, cycle tracks are much better investments than buses as people go faster on bikes than they do on most buses apart from expresses.

    The proper way to increase transit use has been shown in the Rosslyn-Ballston corridor in Arlingon: build a high-speed, high-capacity transit line, then concentrate developments there. Trying to serve areas designed for cars by transit is a losing proposition, the smart choice is transit investments with vast zoning changes to densify massively around the high-capacity transit lines, to concentrate jobs, stores and services around transit stations.

    Land use reforms are the most crucial factor in increasing transit use. You just have to make sure that the transit service has the capacity and quality to deal with new developments, and that bus lines do poorly.


    1. Yes, rail can carry more people per operator; but at the cost of frequency (except in the highest demand corridors). The Ottawa conversion to rail occurred because high frequency bus service built up the market.

      And, when one includes maintenance-of-way (rails, signals, switches, etc.) and station maintenance labor costs, the rail labor advantage starts to evaporate.

      Frequency is the single most important factor to attract users to public transit and where bus frequency combined with travel time and access time is better than the similar access time plus frequency plus travel time for trains, people will choose buses. Rail has its place and buses have their place; but anyone who only advocates one mode is doing all of public transit advocacy a disservice.


      1. First things first, the only ones who advocate one mode are the anti-rail, all-bus-all-the-time guys. Even the strongest pro-rail activist recognizes that there are plenty of routes that have too little ridership to justify upgrading to tail and see the need for feeder buses for rail lines.

        Frequency is important, but I’d argue that it is only as important as speed. In fact, both frequency and speed affect the same factor, which is the travel time. If there’s a bus every 5 minutes but it takes 1 hour to travel 6 miles and get where you want to go, you will not take transit, if there is a rail line near that bus route that has a 30-minute headway but that does the trip in 15 minutes, people will take rail. Since rail has higher capacity and speed, in general it will mean that many people who used to take parallel bus lines will converge on rail, thus increasing ridership, and the lower operating costs will result in more trains running than buses would for the same capacity utilization, especially in off-peak periods.

        As to operating costs, even once accounting for right-of-way maintenance (and not compensating for the fact that buses scrap roadways very quickly but that bus companies do not have to pay to repair them), rail remains much more efficient than buses. In Montréal, the operating cost per passenger in the subway (okay, rubber-tired but it is more like rail than BRT) is 1,30$, in buses, it’s 3,10$. Even including all capital expenses, the cost per ride is probably around 4$ in buses and 2$ in subways, and people go farther in subways than in buses.

        Overall, the STM (Montréal’s transit agency) spend about 3,10$ per passenger to offer the service it offers right now. In Ottawa, which has a BRT instead of a subway, the cost is 3,90$ per passenger. Again, that’s why Ottawa is converting its BRT to LRT, many in Ottawa have started saying that BRT was a bad idea and that they should have gone for LRT in the first place.


      2. I echo Simval’s comment and would like to add that, buses operating in mixed traffic at headways of more than 5 minutes are going to be affected by bus bunching. Yes, steps can be taken to reduce the bunching but it means for every 4 times you have a bus running every 5 minutes, you will have one (like when its raining or there is an accident) where you’ll have no bus for 15 minutes and then three buses in a row.


  2. How confident are you in the statement that “Buses serve more people than trains ever will (since they are more cost-effective), so bus improvements benefit more people” ?

    In the first three months of 2014, according to APTA, heavy rail, light rail and commuter rail served 45% of US transit riders, versus 51% by traditional bus or trolley bus services. The equivalent figures in 1996 were 31% for rail and 67% for buses.* Clearly, there is reason to believe that trains may carry a majority of US transit riders within the decade.

    This isn’t to dismiss your otherwise reasonable argument, just to point out that there clearly has been a significant mode shift towards rail.

    * And this isn’t just a consequence of changes in New York City, in part because of the construction of new lines and reorientation of operations patterns but also in part because of a public preference for rail. In Chicago, rail accounted for 27% of daily CTA riders in 1996 (earliest figures available); it now accounts for 47%; in Washington (WMATA), 58% to 66%; in Philadelphia (SEPTA) 47% to 51%; in Boston (MBTA) 67% to 69%; in Los Angeles (Metro) 7% to 24%; in San Francisco (Muni) X% to 29%; in Portland (TriMet) 12% to 36%; etc.


  3. “Ever will” is obviously a projection, and a bit floral as I said to Alon Levy. It went from 0 to ~20% (post Green Line) in MSP in the period you note, with full build-out of planned rail could maybe get to ~40% in decade. Still in 46 of 51 states (incl. DC) bus carries far more than rail. I don’t think the bus numbers are complete, as it ignores school bus (which is huge, but a somewhat different market, at least until high school), illegal van services (which are directly competitive, but who knows how large), and shorter private inter-city services.

    BRT systems are also being deployed, though not as much as I wish, and if we ever get deregulated to the point of private buses (including Jitney, multi-party taxis, airport shuttles, and other bus-like services), buses will explode, as they can, while rail cannot, since it takes so long to build. (“Uber-izing” public transport just as Uber changes taxi).

    I believe federal support for transportation is fading (unless the Democrats sweep both houses in 2016 ), and that has distorted investments to be capital intensive (i.e. rail).

    As we discuss some in out final piece, bus has a lot more room to grow easily (i.e. it has more low-hanging fruit).


  4. Here’s the deal, I agree with you that buses are good, but I think you need both. When the Silver Line was extended to Wiehle Avenue in Reston, Fairfax County no longer had to run buses all the way into the West Fall Church Metro station. Not only did they use the extra capacity created from having to run only as far as Wiehle Avenue, they bought and are running even more buses (a 40% increase in service). This means that folks that had only one route, now have a choice of two or three routes to pick from during rush hours and a lot of people went from having only M-F daytime service to seven day a week service that extends into the night. If the County can sustain this level of service for five years, I have no doubt the now largely empty buses outside of the peak will fill up over time.

    This is how BRT systems are sucessful in Latin America, a large portion of riders outside the core areas arrive via feeder buses that are integrated from both a fare and service standpoint. I do disagree about deregulation with you David. In fact, in Latin America they are moving towards more not less regulation of bus services due to issues with public safety, overcrowding, pollution, and liability concerns.


    1. IIRC, in Bogota, before the Transmilenio, there were already bus lines down the arterials that were open to all private bus companies. The result was complete congestion on the bus lines, there were so many buses taking the bus lines that they would jam up. Going with a centralized system (though operators are private) allowed them to avoid the congestion of the previous deregulated bus lines and sped up the system tremendously.

      There are cities in the developed world with deregulated bus systems, notably in Japan, but none of these systems seem to work very well (unless they connect to subway or train systems). The big problem with deregulated bus systems is that a proper bus network has a lot of synergies, with lines feeding one another to allow the entire area to become reachable. In a deregulated systems, these synergies are often frowned upon when it means that an unprofitable line from one company is feeding a profitable line run by another company. So deregulated bus lines tend to satisfy spot demand and big destinations but do not offer a coherent system with optimal mobility. The focus is more on optimizing lines rather than optimizing networks.

      Jarrett Walker on his Human Transit blog has a post about Nairobi’s private, deregulated transit system that showcases this. As he says, almost every route goes downtown, creating huge traffic jams in the center and making crosstown trips extremely hard to do as every trip requires to go downtown then take another line to get somewhere else, often imposing huge detours.


Comments are closed.