2010 Transit Data Update

The Antiplanner has reposted the consolidated spreadsheet for the 2010 National Transit Database. The revision of a file I posted last month fixes an error in the calculation of the total number of seats and standing room provided by each transit agency and mode of travel.

More important, the revised file includes some calculations, including BTUs and CO2 emissions per passenger mile, seats and standing room per vehicle, the average number of passengers per vehicle (passenger miles divided by vehicle revenue miles), and operating subsidies per trip and passenger mile. Many more calculations can be made using this spreadsheet and you are welcome to download it and do them.

The Federal Transit Administration added a new kind of transit this year: demand-taxis (id code DT). This is a demand-responsive system that uses private taxis in place of the wheelchair-accessible buses used by many transit agencies. This actually saves money as the average demand-responsive bus costs taxpayers about $30 a ride while the average taxi costs about $17 a ride.

As in previous years, the Antiplanner also added another mode code: SC for streetcars, which tend to carry far fewer people than modern light rail. While the FTA classifies all rail lines that run in streets as light rail, I sort out most vintage trollies and modern streetcars. This isn’t always possible: the report for Portland, for example, lumps its streetcar and light-rail data together.

For the first time, the database separates maintenance (which the FTA calls “capital improvements of existing service”) from true capital improvements (which the FTA calls “capital improvements of expanded service”). Taxpayers spent more than $6.2 billion on true capital improvements in 2010, including $2.9 billion on light rail, $1.7 billion on heavy rail, and $1.1 billion on commuter rail.

Here are some things we can conclude from these data.

1. Transit doesn’t particularly save energy. According to the Department of Energy’s Transportation Energy Data Book, the average car uses about 3,500 BTUs per passenger mile (see page 2-15). By comparison, the average transit vehicle used about 3,440 BTUs per passenger mile. Getting people into cars like the Prius (1,700 BTUs per passenger mile) will do more to save energy and reduce pollution than expanding transit systems.

2. Light rail is even less energy efficient than transit as a whole. Light rail uses, on average, more than 3,600 BTUs per passenger mile. The most energy-efficient form of transit is van pools, followed by publics (shared taxis), heavy rail, and commuter rail. Automated guideways, demand-response, and ferry boats are the least energy efficient.

3. While commuter rail is, on average, more energy efficient than most other transit, the numbers are skewed by New York City’s commuter-rail system. Portland’s commuter-rail line uses nearly 6,000 BTUs per passenger mile, while Salt Lake’s uses well over 5,000. Of course, New York’s numbers also skew averages for heavy rail, bus, and transit as a whole, but not light rail because there is so little light rail in the New York metro area.

4. When taken as a whole, the transit systems for the vast majority of urban areas use far more energy per passenger mile than driving. Only New York, Chicago, Detroit, Atlanta, San Francisco-Oakland, and Honolulu use significantly less energy per passenger mile than driving. While Minneapolis-St. Paul’s and Portland’s systems use slightly less energy than driving. Most others use much more: Seattle’s, for example, uses 4,900, Dallas-Ft. Worth’s uses 6,300, and Phoenix’s uses nearly 6,000 BTUs per passenger mile. This does not count the energy costs of constructing rail lines in these cities, which can be huge.

5. Once again, San Jose takes the prize for having the most pathetic major transit system in the nation. San Jose light-rail cars carry an average of just 16.6 passengers at a time, meaning they are only about 11 percent full. Denver closely follows with just 17.5 passengers per car. Despite the objective failure of its light-rail system and the absence of any need for a heavy-rail line, San Jose is moving ahead with construction of a BART line to Fremont.

6. One of the best, and by some measures the best, transit systems in America is in Honolulu, whose buses use just over 2,200 BTUs per passenger mile (compared with New York’s second-best of close to 2,600 BTUs per pm) and whose buses carry an average of 14.6 riders, compared with a national average for buses of just 10.7. Too bad Honolulu is set on spoiling the system by building an ugly, expensive, elevated rail line.

The problem with transit is that it pretty much has to run nearly empty, on average, if it is going to operate during most hours of the day and to remote suburbs where few people use it. The main use is during rush hour, and that is mainly in one direction even though vehicles have to make round trips to serve commuters. So while Washington MetroRail, for example, may seem packed during rush hour, it officially clocks in at just 13.4 percent full. Some commuter lines, both bus and rail, deal with this by operating only during peak hours. But if you want to serve more than a handful of commuters, you are going to end up with a system that is costly both in dollars and energy.

Share

31 thoughts on “2010 Transit Data Update

  1. JimKarlock

    Perhaps the solution is to operate mass transit ONLY during rush hours and provide subsidized taxi or jitney fares to the needy during other hours. Obviously the non-needy can pay their own way since they are NON- needy.

    Thanks
    JK

  2. Jardinero1

    I would agree with Jim above. In Latin America, Jitneys are widespread, full, and require no subsidy. In the US, transit authorities have an institutional bias against Jitneys in their own systems and work hard to outlaw private Jitney service because they fear, rightly, it will undermine their big busses and rail lines.

  3. Jardinero1

    As an aside, my wife will vouch for the Honolulu bus system. As a teenager, she would spend summers with her Grandparents in Kailua and could ride anywhere on the island for a dime! She went everywhere.

  4. C. P. Zilliacus

    Jim Karlock wrote:

    Perhaps the solution is to operate mass transit ONLY during rush hours and provide subsidized taxi or jitney fares to the needy during other hours. Obviously the non-needy can pay their own way since they are NON- needy.

    To some extent, this is already the case.

    Outside of a few urban areas in the U.S. (New York City, New Jersey, Philadelphia, Boston and Chicago), commuter rail is mostly a peak-direction service only. Same is frequently the case with commuter bus service.

    But when billions of dollars are spent on new passenger rail lines (be it streetcars, light rail or heavy rail), then the government funders of the rail project (including, but not limited to, the Federal Transit Administration) probably want the line used much more than during weekday rush hours. How else to justify “transit-oriented” land use patterns?

    Remember also that these rail transit projects are (usually) intended to provide “an alternative to the private automobile,” which is hard to do if the transit line only runs during weekday peak commute periods.

  5. msetty

    Actually, there are some credible studies that show each passenger mile traveled by transit eliminates roughly 2 auto VMT, a point The Antiplanner has never addressed. On average, that would make transit about 3 times as efficient than driving. I’d supply these references right now but I’m out the door at the moment. Perhaps tonight I’ll post them.

  6. Dan

    In addition to what Mr Setty says, when deployed in concert with denser land uses, VMT decreases even more. Heck, we can even relax the land uses first to naturally – by market forces – increase density and reduce VMT (oh, I know: bad). With continuing rise in fuel prices demand for transit will be market-driven. Everyone is happy, as people…erm…The Market decides.

    DS

  7. Sandy Teal

    Dan and msetty make a lot of sense. It seems intuitive that people who have to travel on inconvenient transit will travel a lot less then people who can travel in convenient and comfortable automobiles.

  8. Jardinero1

    Market forces create denser land uses very readily. The most apt comparison is between Houston and Portland.

    The population density of lightly planned Houston is 3600 psm and the population density of Portland is 4200 psm. The interesting part is what happens in Houston in those areas where there are not even deed restrictions to constrain development. In Montrose you have a population of about 11,200 psm http://www.city-data.com/zips/77006.html and in Gulfton you have a density which is estimated to be above 21,000 psm. Both areas are very vibrant though in extremely different ways.

    As a libertarian and free marketeer, I hate to admit that both areas would have benefited from form based zoning. They likely would have the same densities but be more habitable in the long run.

  9. FrancisKing

    Antiplanner wrote:

    “The Federal Transit Administration added a new kind of transit this year: demand-taxis (id code DT). This is a demand-responsive system that uses private taxis in place of the wheelchair-accessible buses used by many transit agencies. This actually saves money as the average demand-responsive bus costs taxpayers about $30 a ride while the average taxi costs about $17 a ride.”

    This is also true in the UK, during busy periods. Does it work so well when only one person want to use the service?

  10. FrancisKing

    Antiplanner wrote:

    “The problem with transit is that it pretty much has to run nearly empty, on average, if it is going to operate during most hours of the day and to remote suburbs where few people use it. The main use is during rush hour, and that is mainly in one direction even though vehicles have to make round trips to serve commuters.”

    This, however, is not true. Local municipality staff can drive (mini?) buses during the rush hour at both ends of the day, and do office work between those times. This then works with a steady flow of buses throughout the day provided by the bus company, providing a peak response when required by demand. The problem is that traditionally all buses have been worked at a steady rate throughout the day – so not enough buses during rush hour, and too many buses at other times.

  11. JimKarlock

    msetty: Actually, there are some credible studies that show each passenger mile traveled by transit eliminates roughly 2 auto VMT
    JK: Well, duhh!
    When you spend an hour on transit making a trip that would take 10 min by car, you simply don’t have time to travel very far. That also means you can’t travel to a better job. Or to a lower cost home. Transit wastes time which is money.

    msetty: that would make transit about 3 times as efficient than driving.
    JK: Interesting claim since transit costs several times the cost of driving and is much slower.

    Thanks
    JK

  12. C. P. Zilliacus

    MSetty posted:

    Actually, there are some credible studies that show each passenger mile traveled by transit eliminates roughly 2 auto VMT, a point The Antiplanner has never addressed. On average, that would make transit about 3 times as efficient than driving. I’d supply these references right now but I’m out the door at the moment. Perhaps tonight I’ll post them.

    But why is reducing VMT with transit a desirable goal?

    Is reducing VMT a desirable goal at all?

    Especially when (at least in the United States) the generators of that VMT (in other words, highway users) fund very large percentages of transit operating subsidies and even larger percentages of transit capital subsidies are funded through motor fuel tax payments, tolls and other fees and taxes levied on highway users?

    As a colleague and friend of mine said a long time ago, transit loses money on every customer and does not make it up in volume.

  13. Sandy Teal

    I hope the Antiplanner works to publicize his carbon based analysis of transit vs. automobile. Probably 90% of the public believes that transit is very carbon friendly.

  14. the highwayman

    Sandy Teal said: I hope the Autoplanner works to publicize his carbon based analysis of transit vs. automobile. Probably 90% of the public believes that transit is very carbon friendly.

    THWM: O’Toole’s “job” is to sell gas & cars. So with O’Toole, numbers don’t count!

  15. Dan

    Jardinero: both for the increase in land rents around transit, and the additional supply of land that is no longer needed to park in two places and drive on (use your favorite number for sqft needed for each car, but regardless of that number, it will go down with fewer VMT (less demand for vehicle trips)).

    DS

  16. C. P. Zilliacus

    Dan wrote:

    Emissions, public health, land rents…just off the top of my head in 3 seconds.

    Emissions (and health impacts of same) have largely been dealt with through much better vehicle emission controls and (in some cases) reformulated fuels (both gasoline and Diesel fuel).

    By “land rents” do you mean land use?

  17. Dan

    CPZ:

    Emissions (and health impacts of same) have largely been dealt with

    No [1, 2, 3].

    By “land rents” do you mean land use?

    No, economic rent or Ricardian rent. ["...Economic rent is due partly to differences of productivity, but chiefly to advantages of location..."]

    DS

  18. larryscheib

    I’m confused. Antiplanner clusters bus transport with rail which appears to me is an attempt to hide a bias. Bus service tends to service less dense areas (I’ve ridden many express “ghost” buses while commuting in Phoenix) and can skew comparisons dramatically. Comparing the more dense based rail with auto better illustrates the advantages of mass transit over the auto, 3500 btu vs 2600 btu. There are also many other considerations when comparing transportation modes. For instance I don’t have to worry about paying for parking my Bombadier train set at my office each day, or insurance coverage in case a Prius rams me head-on into its oblivion. Plus the rail foot print is much smaller than roads per passenger transport rate. Further, rail systems are only as good as their design and full integration to an urban setting. Phoenix did a poor job in design hence creating an eternity to move from point A to point B. The Seattle line from downtown to the University district will take a fraction of the time over a similar distanced section of the Phoenix line. Also the more developed out and integrated a system, say Melbourne Au, the more efficient and greater its return.

  19. Andrew

    All of the analysis offered here is insane.

    First of all, Randall is only looking at CARS, not light trucks (pick-ups, SUV’s, minivans, etc.). Most VEHICLES on the road are not CARS. Light trucks have half the efficiency of cars. And Toyota Pius’ don’t get the mileage that is claimed in my own driving experience with them.

    Second, we all know that the vast majority of trips in VEHICLES for purposes of getting to work are single occupancy. Carpooling has been falling off a cliff for decades. So forget the whole “pay no attention to the man behind the curtain” routine by pretending that 0.6 other people are riding to work with everyone. Transit is typically not used for grocery shopping, driving to Church, going out to dinner, or taking the kids to a party or soccer game. So lets forget all those trips were mom is taking the two munchkins to ballet lessons. None of those trips matter for whatthis discussion is.

    Once we free ourselves from the BS of obfuscating the entire topic, we can then compare private single occupancy vehicles at well over 6000 BTU per mile on average with Antiplanner’s preferred crappy transit mode of buses at 4200 BTU per passenger mile, and various rail modes at around 2600 BTU per passenger mile ignoring a handful of toy streetcars.

    Of course this doesn’t make the point Randall wanted, so its not surprising the analysis was not offered in this way.

  20. the highwayman

    JerK: Those ARE NOT O’Toole’s numbers — they are the federal government’s.

    THWM: It doesn’t matter if 50 cars or 5000 cars drive past your house a day, the street is there by default!

  21. Dan

    Andrew, I’m pretty sure I’ve pointed out the fleet BTU before, and the mendacity about ridership. And that only 1/4 of trips are for work. Thank you for reminding us that facts rarely get in the way of wishes.

    DS

  22. Sandy Teal

    If transit is not orders of magnitude more efficient than cars, then there is no reason to invest in transit for the purpose of addressing global carbon emission issues. Huge government investments in marginal improvements is not going to do anything meaningful according to the carbon climate literature.

  23. C. P. Zilliacus

    Dan wrote:

    Emissions (and health impacts of same) have largely been dealt with

    No [1, 2, 3].

    A paper linked by reference 3 above (physical page 34) says (emphasis added):

    However, approximately 1%, 3%, 3%, and 30% of residents in San Bernardino, Riverside, Tulare, and Kern Counties, respectively, are estimated to be exposed to annual PM2.5 concentrations above 15 ?g/m3 under the NAAQS attainment scenario. It is important to recognize that the 4-5 ?g/m3 reductions in annual PM2.5 to achieve NAAQS attainment represent a dramatic improvement in air quality relative to background levels, and a dramatic reduction in population exposure to harmful levels. Furthermore, since the daily PM2.5 standard is more stringent than the annual standard, it is quite possible that the emission control plans adopted to attain the daily PM2.5 standard may result in greater reduction in annual PM2.5 than estimated in this study.

    It makes the point that I was trying to make above.

    Those improvements in air quality will continue as the fleet continues to get better in terms of emissions.

  24. Dan

    CPZ,

    1) if emissions have largely been dealt with, then there would be much, much fewer health problems than found in all the links I provided. Granted, the given health problems largely occur in the weak, poor, old, and children, so they don’t get the play that health problems of other cohorts get, and as such maybe they don’t count. YMMV.

    2) The section you quote from quantifies the % of residents exposed to air pollution you assert was ‘dealt with’, then in Section III details some issues with those exposures, e.g.:

    Ozone and fine particles (PM2.5) have long been associated with adverse health effects, and a growing body of health science literature enables us to quantify how changes in air quality translate into changes in the number of adverse health effects in a population.

    and then Section V quantifies how many people have negative outcomes based on the quantified pollutant, including for example:

    Table V-1. PM2.5-related health effects in the South Coast Air Basin
    Table V-2. PM2.5-related economic values in the South Coast Air Basin.

    So surely the effects are lessened, but as the two given tables immediately above and the subsequent tables show, ‘largely dealt with’ is a stretch at best, as the paper details. ~US$21B in economic losses from PM 2.5 isn’t ‘largely dealt with’. ~400k work loss days from PM 2.5 isn’t ‘largely dealt with’. The costs from other pollutants is an exercise left to the reader.

    Regards,

    DS

  25. the highwayman

    Andrew said:
    All of the analysis offered here is insane.
    First of all, Randall is only looking at CARS, not light trucks (pick-ups, SUV’s, minivans, etc.). Most VEHICLES on the road are not CARS. Light trucks have half the efficiency of cars. And Toyota Pius’ don’t get the mileage that is claimed in my own driving experience with them.
    Second, we all know that the vast majority of trips in VEHICLES for purposes of getting to work are single occupancy. Carpooling has been falling off a cliff for decades. So forget the whole “pay no attention to the man behind the curtain” routine by pretending that 0.6 other people are riding to work with everyone. Transit is typically not used for grocery shopping, driving to Church, going out to dinner, or taking the kids to a party or soccer game. So lets forget all those trips were mom is taking the two munchkins to ballet lessons. None of those trips matter for whatthis discussion is.
    Once we free ourselves from the BS of obfuscating the entire topic, we can then compare private single occupancy vehicles at well over 6000 BTU per mile on average with Antiplanner’s preferred crappy transit mode of buses at 4200 BTU per passenger mile, and various rail modes at around 2600 BTU per passenger mile ignoring a handful of toy streetcars.
    Of course this doesn’t make the point Randall wanted, so its not surprising the analysis was not offered in this way.

    THWM: Funny you say that, the times that I do drive it is mostly in a Toyota Highlander.

  26. the highwayman

    Sandy Teal said:
    If transit is not orders of magnitude more efficient than cars, then there is no reason to invest in transit for the purpose of addressing global carbon emission issues. Huge government investments in marginal improvements is not going to do anything meaningful according to the carbon climate literature.

    THWM: Most auto use is of the SOV type also for the sake of example Portland OR, won’t brake even till it has 200 miles of tram line again.

Leave a Reply