As everyone knows, the country is going to undergo some significant "graying" over the next few decades with the proportion of senior citizens going up substantial. That's going to pose a challenge on many levels, but one level is that building more and better transit will help us cope with the problem that seniors often aren't comfortable driving, especially in sub-optimal conditions, and it often isn't safe to have them on the road. If we don't want to see a huge proportion of the population immobilized, we're going to need more ways for people to get around.
« Bush Gets It Right | Main | Straight Talk »
Transit, Transit Everywhere
26 Jun 2008 08:41 am
Comments (145)
I think it is important to note that seniors and, say, commuters don't necessarily have perfectly aligned transit needs. So this could play out in various ways. The more hopeful possibility is that seniors, commuters, and other potential users of public transit work together to increase the total availability of public transit. The less hopeful possibility is that seniors and others end up competing for the biggest slices of a limited pie. I think the former scenario is more likely, however, at least assuming fuel prices remain high (or go higher).
On the other hand, there may be a tradeoff: mass transit generally isn't point-to-point, and it may be more difficult for older people to walk two to five blocks after getting off the bus/subway than for younger people. Something to keep in mind when designing the transit systems. (And probably something to study empirically; how many of the elderly people who can't walk far can't drive either? How big a problem is each of these factors? I don't know.)
Unfortunately the future probably looks like more of the present: seniors in nursing homes or retirement villages with a bus running to the grocery store.
Seniors love golf carts.
I agree with Matt Weiner. Not exactly OT, but one the driving forces for people with kids to get cars and move out to the 'burbs is the need for point-to-point transportation (which driving forces tends to get ignored on blogs like this, in favor of other explanations for the move to the 'burbs).
It's fine to walk to the bus, take a bus to the store (or just walk to the store, dependending on how far away the store is), get your groceries, and walk/bus back (pushing them in one of those cart thingies) when you are a young, healthy person buying supplies only for yourself. But when you are shopping for a kid (and kids take a lot of stuff) or you are not so young and healthy, it's rather onerous to use public transport. Or heck, even to use a car (which in a major city is not exactly point-to-point, as often your parking space, if you have one/can find one, is not exactly right in front of your residence), it's more convenient and cheaper in the suburbs (sorry, this is on my mind -- I'm moving to the city -- The City -- soon the fact that I'll have to pay for a parking spot such a walk from my apartment is rankling when I'm used to parking right behind my apartment for free).
Of course, c.f. DTM above, some solutions benefit all users of public transport ... more routes, more frequency of service ... but after a while, even in a city as dense as NYC, you can only have so many routes and so much frequency and still have it be mass transport as opposed to practically individualized transport (which you gain nothing from having it public).
OTOH, if you build it they will come might apply to public transport. I might drive to my work because the bus to my work only comes once an hour. Because so few people use public transport, they don't have greater frequency. But if the bus came often enough, people like me would use it ... so it would justify the frequency. But it might take a while as people need to get used to arranging their lives around public transport (you can miss a bus ... you can't miss your car).
I don't really like the top-down planning aspect here. Do we really know that seniors would especially like public transit?
As an alternative, what about zero-emission taxis in a few decades?
A broader point: If public transit can only survive through government subsidy, I think something is wrong. Maybe public transit can survive on its own, or maybe the problem is that the alternative (car driving) is so heavily subsidized.
But our spending on a program should reflect its value to us. Markets are a good way of solving that problem. Why can't markets establish mass transit without government subsidy?
mk,
Indeed we do heavily subsidize automobiles (from tax breaks for car makers to our government built road systems). Infrastructure actually is considered a valid role of government, even by many libertarians. So why are our railroads in largely private hands?
As to buses, at some level they cannot compete with cars, simply because of the time aspect -- a bus, sharing a road with a car, is always going to take longer than the car as it has to make stops, etc.
and, FWIW, markets are hardly a panacea ... at the very least, as any athlete in any sport besides Calvinball will tell you, any competition requires rules, regulations, referees, etc. And these cost $$$ and naturally involve the government (or some sort of consortium that becomes entirely equivalent to government).
I have a driver. What's the problem?
mk,
Your statement can easily be turned around to any public good:
Why can't markets establish standing armies without government subsidy?
Why can't markets establish jurisprudence without government subsidy?
Why can't markets establish police and fire protection without government subsidy?
Perhaps you should try rephrasing your question.
Unless you assume that all of those seniors are packed together in one residence, transit for seniors has exactly the same problems as transit for anyone else in the suburbs: transit sucks at moving dispersed people to dispersed destinations (and then back).
Even if you assume that seniors will be packed in some kind of senior living center, then a dedicated bus for said center (whch ten is paid for as part of the fee for living there) makes more sense.
I laid out why rail systems are a non-answer over here at the bottom of the comments; never fear though, Matt, who has a man-crush on transit systems, is still casting about for a way to sell us on systems that serve little purpose and cost tons of cash.
infrastructure actually is considered a valid role of government, even by many libertarians.
I guess I'm not sure why. What collective action problem or coordination game is the government solving?
I could see people saying, the roads should be open to everyone, and privatized toll roads might just target the richer segments, or something. Is that the issue, are there other things?
Matt, who has a man-crush on transit systems, is still casting about for a way to sell us on systems that serve little purpose and cost tons of cash.
What's it called?
MONORAIL!
Great. Can't wait for this thread to get all Mixnered up. After all, "transit" is right there in the title!
This strikes at the heart of the huge inequities in our auto-centric model of transportation--we've built our whole society around making it easy for young (but not too young), healthy, wealthy people to get around, and consequently made it almost impossible for everyone else.
For my part, although I am a public-transit booster, I think technology is going to play a big part in fixing this mess. I look for autonomous vehicles to start appearing in 15 or 20 years...and we may live to see the day when you're not even _allowed_ to drive your own car on the Interstate.
That sort of thing will help with the elderly and the disabled. The poor? Out in the cold again. It's never much fun to be poor.
Your statement can easily be turned around to any public good
I'm not sure mass transit qualifies as a public good, at least not in the classical sense.
A public good is "non-rival" and "non-excludable." In other words, 1) my consumption of it does not reduce another's ability to consume it, and 2) if the good is present at all, people can't practically be excluded from using it.
Typical examples are national defense and air (to breathe).
By contrast, mass transit seems to be rival (because it does not have practically infinite capacity) and excludable (you pay a fare to ride).
Craig,
I agree. "Automated" driving is definitely the holy grail of traffic engineering. We human are horrible drivers... Highly distracted, needlessly selfish (people who block the box, drive slow in the left lane, etc.) and prone to mistakes such as over-braking. Automation of cars would solve a hell of a lot of problems. No more drunken driving. Break-downs could be efficiently moved off to the side of the road, etc.
Is that the issue, are there other things?
A great many other things. Oil is truly magical stuff. Discovery and implementation of a gasoline powered system permitted humans to do novel things like build sprawling suburbs, take cheap flights, endlessly pave roads at low cost, and adopt economic outlooks that assume it is an attitude, not the resources and technology, that propels this great nation forward. Once demand for oil outstrips supply, a lot of those assumptions and lifestyle choices don't make as much sense. Because technical fixes have been the rule for 60 years or so, it is a popular notion that some kind of combination of new technologies will keep butts in car seats, happily motoring into the future for always and always Amen.
The hard truth is that there isn't an easy replacement for oil. No portable liquid or substance sitting around, virtually free, has the same energy output as oil. Assuming the population will continue to rise and that sprawl has gone about as far as it practically can, congestion becomes an issue in more and more places.
mk, there is no reason that free market principles can't play a big role in switching to a more realistic system or moving people and product around, but some coordination is needed right now because this is leaving crisis territory and moving into emergency territory. For seniors, for working class folks, and especially the poor.
That technical fix you don't believe in seems to be closer than you think.
I'll certainly be amused when the wailing and gnashing of teeth on the left gets going as they realize that the car economy is likeley to continue to be possible (and thus preferable to most Americans).
Good look with the hairshirt approach of being a "climate change" scold; I expect that to work about as well as Jimmy Carter's sweater.
"By contrast, mass transit seems to be rival (because it does not have practically infinite capacity) and excludable (you pay a fare to ride)."
Mass transit is but one part of a transit infrastructure. Highways, railways, waterways, airports, etc. are all part of transit infrastructure.
Transit infrastructure is beneficial to the economy as a whole, much like energy infrastructures, etc. On the whole, your definitions are quite dodgy and suspect.
"1) my consumption of it does not reduce another's ability to consume it, and 2) if the good is present at all, people can't practically be excluded from using it."
If you consume water, doesn't that mean there is less water available for everyone else? By your definition water is excluded from this list. In fact, most anything can be excluded. We can't all go to court at the same time, therefore courts are practically excluding people from using it. Overall these sorts of "classical" libertarian beliefs are signs of mental laziness. I suggest taking some time to actually think things through and take your definitions to their logical conclusions.
Re James Robertson
A little caution is required on Dr. Venters' idea. I can recall back in the early 1970s a prediction that controlled fusion and economical solar energy would be on line by the end of the Millennium. We are now some 30+ years later and controlled fusion is still a chimera and solar energy is still not economically viable. If something sounds too good to be true, it probably is. What Dr. Venter is proposing is a free lunch and as Mr. Robertsons' fellow conservative, George Will, famously states, there is no such thing as a free lunch. I certainly hope that Dr. Venters' idea proves out, as it would provide a solution to both peak oil and global warming but past experience tells me that there may turn out to be numerous flies in this ointment.
James Robertson, who is seeming increasingly to be Mixner's sock puppet, links to an article in Popular Mechanics, journal of flying cars and personal jet packs, that does not address any questions of scalability or infrastructure development lead times.
The world currently consumes 1 cubic mile of crude oil per year. The sugar feedstock that the bacteria in this article consume is one third as energy dense as oil, so we will need to produce at least two to three cubic miles of sugar to replace existing oil consumption. Does Mr. Robertson have any idea how much arable land, water, fertilizer and pesticides will be required to produce that much sugar? What will the effects of such production be on food supplies? The cellulose converting bacteria mentioned in the article don't even exist and cellulose is one half to one third less energy dense than sugar, so production will need to be commensurately larger to compensate.
Sorry to go all "Jimmy Carter's sweater" on you, but you have given no reason to think that our car economy has nothing to worry about.
I think that algal/bacterial biofuels will be important factors in our energy future, but I also think that parking the Escalade will be just as important, if not more so. If Mixner and Mr. Robertson ever presented arguments that integrated an understanding of the scalability issues that underlie their preferred solutions, they would deserve to be taken more seriously.
Mass transit is but one part of a transit infrastructure. Highways, railways, waterways, airports, etc. are all part of transit infrastructure. Transit infrastructure is beneficial to the economy as a whole, much like energy infrastructures, etc.
Agreed. So it seems the question is: how much of these widespread benefits are externalities, and how much are captured by the people who directly participate?
Because if there are few positive externalities, arguably there is less reason for the government to be involved since there is no collective action problem.
There perhaps are some externalities here. If I build a road to east balargaville, the shops there will be happy because they will get more customers. That added value is not represented in the transaction between the driver and the road owner.
How big this externality is, is difficult to say without some research.
If you consume water, doesn't that mean there is less water available for everyone else? By your definition water is excluded from this list. In fact, most anything can be excluded. We can't all go to court at the same time, therefore courts are practically excluding people from using it.
Yes, I think I pulled us off course with quibbling over the definition of public goods. The classical definition is indeed restrictive-- we agree that provision of public goods is only one of the government's proper set of roles. The point about definitions was really a sideline to the discussion.
Probably best to focus on whether there are externalities (i.e. whether there is a collective action problem).
It's interesting that there are about 22 comments here but nobody has really talked about Transit Oriented Development (TOD). This may in part reflect Matt posing the question as providing transit to riders, when the question can just as easily be stated as providing riders to transit.
TOD, of course, is a simple concept so strong it can force itself even upon the limited attention span and memories of Americans. Build a transit line and land values by stations rise, because you have created a market.
For the entire lives of many commenting here, this very natural tendency of markets to form on transportation lines has been warped by massive governmental action building roads, and using American armed might to ensure a supply of cheap gas. How well this has worked may be judged from the fact that, at the county, state, and federal level, the governments have gobbled so much of the people's money, building and supporting sprawl, that people now groan at the thought of paying for schools, health care, or indeed, any expense they are allowed to have a say about.
Nonetheless, TOD is such a strong and natural economic force that, as soon as it is certain that a rail transit line will be built, developers begin assembling parcels, new buildings that are more energy-efficient are built, and people move to live and do business close to the stations. From a regional planning perspective, this has follow-on benefits too numerous to list in a simple blog post.
Suffice it to say, you don't need to bring the transit to the seniors, you can bring the seniors to the transit. And thereby hangs a tale...
On the other hand, there may be a tradeoff: mass transit generally isn't point-to-point, and it may be more difficult for older people to walk two to five blocks after getting off the bus/subway than for younger people. Something to keep in mind when designing the transit systems. (And probably something to study empirically; how many of the elderly people who can't walk far can't drive either? How big a problem is each of these factors? I don't know.)
Hmm...I'm currently living in a country with extensive mass transit infrastructure (I forget the exact stat, but in my city no one lives more than something like 500 meters from a transit stop). This is a fairly poor country by American standards, with per capita GDP about 50% of the U.S. A zillion oldsters ride the buses and trams, and manage to get where they are going -- and this in a land where hip replacement surgery is basically unheard of: you see a lot sad old, hunched-over bodies shuffling along with crutches and canes and walkers. It's a tough life, I assume, but these are people who survived the Nazis AND Stalin. I wonder if our pampered American geezers could hack it.
Of course, this is mostly moot. This country has a traditional urban design fabric, which makes the transit/pedestrian intensive lifestyle relatively feasible. Most of the U.S. -- 95% of it west of the Allegheny Mountains -- is just not designed at present to snap this kind of infrastructure on top of it without massive upheaval. All the worse for us, as far as I can tell.
mk,
Yep, the economic reason for public transit subsidies are the positive externalities. That issue has been studied extensively, but obviously you have to do case-by-case studies to determine what subsidy (if any) is warranted for a particular project.
But the short answer is that the positive externalities to transportation infrastructure in general tend to be pretty substantial, basically because the external beneficiaries typically include whomever is involved economically with the people and goods travelling. And note this is true on either end of a given route (e.g., a good commuting route benefits the commuters' employers, but also the developers who provided the commuters' residences, and so on).
And when you get to something like public transit, you can get some additional externalities from enabling switching from other modes of transportation. For example, to the extent a new public transit service helps reduce congestion on a nearby highway by allowing some drivers on that route to switch, the remaining drivers are external beneficiaries of the public transit (and so in fact are the related external beneficiaries of the highway). Other switching externalities come from freeing up more land for other uses (since public transit tends to have more capacity per land unit), reducing fuel consumption and emissions, and so on.
Again, though, all this has to be done on a case-by-case basis. Still, that is the general answer for why public subsidies for transportation infrastructure make sense.
Another day, another bad argument for transit. Here's why:
1. The population has already been graying for decades, and simultaneously transit has been losing market share to private autos. Transit's share is now so low (about 1% of total surface transportation passenger-miles) that it may not fall much further in the near future, but it's not going to grow significantly either.
2. If there is greater demand for transit for the elderly, buses would be a far more logical solution than trains. Buses are much cheaper, can provide much greater coverage close to people's homes, and are generally more accessible by the frail and disabled than trains.
3. A better solution than buses may be an increase in municipal paratransit services (dial-a-ride vanpools and other auto-based solutions).
To some extent, Matt's vision of New Towns have already occurred in the Sun Cities and Del Webbs. Seniors move into a house on the golf course and travel around by golf cart. When they become less mobile they move to the condos near the hospital. These "cities" have their own shuttle buses to the grocery stores, pharmacies, and rec centers located within the community and to airports and shopping outside of the community.
I wonder at the desire of older Americans to be "highly mobile." The statistic that soon 1 in 4 drivers will be older than 65 is a bit misleading. Perhaps 1 in 4 people with a driver's license will be older than 65, but unless you happen to be within range of K&W Cafeteria around 3pm, I doubt 25% of the drivers actually on the road will be that old.
I also like the comment that keeping younger drivers off the roads saves lives. While it's mentioned that older drivers have degraded abilities, nothing is mentioned about keeping them off the road in order to save lives. Wonder why?
DTM,
And when you get to something like public transit, you can get some additional externalities from enabling switching from other modes of transportation. For example, to the extent a new public transit service helps reduce congestion on a nearby highway by allowing some drivers on that route to switch, the remaining drivers are external beneficiaries of the public transit (and so in fact are the related external beneficiaries of the highway). Other switching externalities come from freeing up more land for other uses (since public transit tends to have more capacity per land unit), reducing fuel consumption and emissions, and so on.
Same old nonsense.
There is no credible argument for transit over private autos on energy-efficiency or pollution grounds. Transit buses are less fuel-efficient and more polluting than cars. Most transit is buses. Transit trains provide only modest benefits in fuel-efficiency and reduced pollution over cars. Future cars will almost certainly be more fuel-efficient and less polluting than both trains and buses.
There are many other ways of addressing congestion than expanding transit. Examples include building more roads, introducing congestion pricing, promoting telecommuting, car pooling, and staggered work schedules, and increased vehicle and highway automation. Transit is in general much slower than cars, and much less flexible, comfortable and convenient, which is why the vast majority of commuters prefer to drive.
Transit trains provide only modest benefits in fuel-efficiency and reduced pollution over cars.
Weren't you trying to convince us last week that trains were actually less fuel-efficient than cars? I swear I remember you and DTM boring the rest of us to tears about this point.
Mixner,
I didn't realize you are still peddling this line: "Transit buses are less fuel-efficient and more polluting than cars."
Just checking--do you still believe that is "fundamentally" and "necessarily" the case?
DTM,
I didn't realize you are still peddling this line: "Transit buses are less fuel-efficient and more polluting than cars."
I, on the other hand, am painfully aware that you are still peddling the utterly discredited "transit buses are more fuel-efficient than cars" line.
Just checking--do you still believe that is "fundamentally" and "necessarily" the case?
Yes, I "still" believe it.
Mixner,
Well, then you might want to check out the latest version of the chart you were citing. I believe you were relying on the 2000 numbers, but the latest data is actually from 2005:
http://www.bts.gov/publications/national_transportation_statistics/html/table_04_20.html
Between 2000 and 2005, the fuel-efficiency of passenger cars by this measure improved slightly, but the fuel-efficiency of transit buses improved far more, causing transit buses to pass passenger cars.
Oops. I guess the situation in 2000 wasn't as fundamental and necessary as you were arguing.
By the way, those fuel efficiency gains in transit buses were in response to what in retrospect were relatively modest increases in diesel prices from 2000 to 2005. Care to bet on what is going to happen in response to the much greater increases in diesel prices since then?
DTM wrote:
But the short answer is that the positive externalities to transportation infrastructure in general tend to be pretty substantial, basically because the external beneficiaries typically include whomever is involved economically with the people and goods travelling. And note this is true on either end of a given route (e.g., a good commuting route benefits the commuters' employers, but also the developers who provided the commuters' residences, and so on).
Thanks, that clarifies things.
I wonder to what extent these externalities could be internalized by making developers and shopping centers parties to the transaction of road-building. But anyways, it seems clear there is some collective action problem that requires some amount of subsidy.
DTM,
Well, then you might want to check out the latest version of the chart you were citing.
We've been over this. You're still not listening. Here, I'll explain it to you again. Read carefully now:
The data in the table you cite only goes to 2005. We're now half-way through 2008. New vehicle sales over the past year or more have overwhelmingly favored smaller and more fuel-efficient cars, and that trend is likely to continue if gas prices remain high. Hybrid vehicle sales are also increasing rapidly. These trends increase the average fuel-efficiency of the nation's private auto fleet over time. Also, single-year comparisons aren't very meaningful. There's significant year-to-year fluctuation, and even for 2005 the difference is statistically insignificant. The longer-term data shows a clear fuel-efficiency advantage for passenger cars over transit motor buses.
Furthermore, other reports, using more recent government data, also indicate a clear advantage for passenger cars over transit buses. See, for example, Table 1 in this report. The data comes from government sources for 2006 through 2008. The table indicates that passenger cars are substantially more fuel-efficient than motor buses (about 20% more efficient), and more efficient even than trolley buses. This is consistent with the information in the Surface Transportation report I cited earlier.
mk,
The problem is that in most cases the external beneficiaries are diffuse. That is because the people and goods using a given chunk of infrastructure are likely to be travelling a lot of different routes for many different purposes. So, it quickly becomes a lot more trouble than it is worth trying to identify specific external beneficiaries, and instead it makes sense just to provide the subsidy through broad taxes. In fact, generally about the closest people typically come to breaking down the external beneficiaries is trying to figure out an appropriate split of the public subsidy between city, county, state, and federal sources.
Mixner,
Right, you were wrong in your predictions the last time once we got out of sample data, but this time you will surely be right!
DTM,
Right, you were wrong in your predictions the last time once we got out of sample data, but this time you will surely be right!
What are you blabbering about now? The greater fuel-efficiency of cars is not a "prediction," it's a fact.
Energy consumption in BTUs per passenger-mile:
Motor Buses: 4,365
Trolley Buses: 3,923
All Automobiles: 3,885
Toyota Prius: 1,659
As you can see, even our existing automobile fleet is more fuel-efficient than buses. And the Toyota Prius, which represents the future of auto technology, is vastly more fuel-efficient than buses.
Sigh. Insanity is doing the same thing over and over and expecting different results. But, I guess I'm not really expecting different results. Hopefully this'll just keep folks not named Mixner from being misled:
Furthermore, other reports, using more recent government data, also indicate a clear advantage for passenger cars over transit buses.
Of course, as even the Cato Institute admits:
"All of these numbers are very sensitive to load factors. Because the vehicles themselves tend to weigh far more than the passengers being carried, doubling the number of people on board any vehicle will cut the energy consumption and emissions per passenger almost in half."
In other words, this low figure for transit buses is thus, in fact, purely a reflection of the status quo of very low transit ridership in many regions. To use that to argue against improvements in transit is tautological at best.
(And of course there are a host of other problems with the figure. Transit buses run short, stop-and-go trips on city streets, while automobiles misleadingly receive an efficiency boost on a per mile basis from longer highway trips. Potential efficiency improvements in automobile technology parallel potential efficiency improvements in bus technology. Etc.)
Mixner's facile ceteris paribus analysis--buses have low ridership, and therefore low efficiency per mile, and therefore automobiles are superior--utterly fails to address the central tenets of transit oriented development:
Increase ridership and simultaneously reduce passenger miles traveled by building relatively dense, walkable communities in which far fewer vehicle trips and miles are needed, and in which transit, when necessary, is as ubiquitous and convenient as possible.
jack lecou,
In other words, this low figure for transit buses is thus, in fact, purely a reflection of the status quo of very low transit ridership in many regions.
No, it's not low ridership. You misread your own quote. It's low load factors. Load factors in transit buses and trains will always be low because of the nature of the service they provide. The need to provide frequent service and to accommodate large variations in demand across different times of the day and different segments of their routes means that most seats will be empty most of the time.
mk, just so you know: people are not really interested in rehashing the very premises of government for your benefit when discussing these issues. If you're a libertarian and don't really understand why people think the government has a role in maintaining public infrastructure, you should learn more about the philosophy of government and public works.
There's a bridge that connects my neighborhood to the adjacent neighborhood. You know why? Not because "the market" has demanded that there be a bridge, but because it makes my life pretty damn nice that I can get to the adjacent neighborhood by taking that bridge. Similarly the massive park near me. Did "the market" decide that there should be a park bisecting DC? No, but we decided that it would be pretty darn nice to have a big park. I suppose joggers could pay tolls to pay for it, and we could see if the park was market-supportable, but most of us seem to like the park regardless as to whether it would exist under mythical market forces.
The fact that Mixner had to abandon his original source, the Bureau of Transportation Statistics, and get something from Cato pretty much tells you what you need to know. And the categories don't even match up--we were once discussing "transit buses", but now Mixner is trying to substitute "motor buses".
The funny thing about that attempted switch is that back when Mixner was defending the BTS chart, he made a big deal about it being specifically transit buses--we weren't supposed to look at other categories even though the vehicles were equivalent. But now that his former source has betrayed him, he has flip-flopped and apparently now we now longer need to look at specifically transit buses.
Which is understandable, I guess, because the only transit-specific category in the Cato chart is "All Transit". And "All Transit" comes in ahead of "All Automobiles" on Cato's chart. Oops ... not a Mixner talking point, that. And funny enough, somehow that "All Transit" category didn't come over when Mixner lifted data from Cato.
Finally, I looked into Cato's sourcing and it turns out there are different primary data sources for different kinds of vehicles in their table. That is a very big no-no, of course, and obviously it makes sense to stick with a single, offical source.
But like I said, it really says all you need to know that Mixner defended the BTS chart to death one day, then abandoned it the next day when it turns out the more recent version of the same chart contradicted his assertions.
Mixner,
Just for fun ... exactly what then do you think explains the difference between the 2000 and 2005 BTS charts? In 2000 the transit bus BTU/passenger-mile was 4,147. By 2005 it was 3,393. And this isn't just a product of those particular years. Transit buses were over 4000 every year from 1994 to 2000. They then went under 3700 in 2001, under 3600 in 2002, 2003, and 2004, and finally under 3400 in 2005.
What is your explanation for this change in the statistics? And if you want to claim it isn't statistically significant, show me your calculations, because I highly doubt that is right.
DTM,
The fact that Mixner had to abandon his original source, the Bureau of Transportation Statistics, and get something from Cato pretty much tells you what you need to know.
I didn't "abandon my original source." And the fuel-efficiency figures I provide above are from government sources for 2006 through 2008.
And the categories don't even match up--we were once discussing "transit buses", but now Mixner is trying to substitute "motor buses".
Transit buses consist of motor buses and trolley buses. The vast majority are motor buses. Cars are more fuel-efficient than both types of bus. The Toyota Prius is vastly more fuel-efficient.
Which is understandable, I guess, because the only transit-specific category in the Cato chart is "All Transit". And "All Transit" comes in ahead of "All Automobiles" on Cato's chart.
"All transit" is only slightly more fuel-efficient than "All automobiles" and is no more efficient than "Passenger Cars." "All Transit" is much less fuel-efficient than the Toyota Prius, which represents the future of car technology. In fact, the Prius will probably seem like a gas-guzzler compared to the typical new car of 10 or 15 years from now.
The bottom line is that transit overall has, at best, marginal fuel-efficiency benefits over autos today. Transit will almost certainly be much less fuel-efficient than the nation's auto fleet 10 or 20 years from now. And transit buses are already much less fuel-efficient than autos.
DTM,
Finally, I looked into Cato's sourcing and it turns out there are different primary data sources for different kinds of vehicles in their table. That is a very big no-no, of course, and obviously it makes sense to stick with a single, offical source.
There is no "single, official source" because no one agency provides data for all transportation modes. If you had bothered to read your own source more carefully, you might have noticed that its highway and transit bus data come from two different sources: the Department of Transportation and the American Public Transit Association. APTA isn't even a government agency at all. It's a lobbying group for the transit industry.
Mixner,
If you didn't abandon the BTS chart, then why are you still saying things like "And transit buses are already much less fuel-efficient than autos"? The BTS chart says different, so if you were still accepting its validity you would have to stop making that claim. And for that matter, Cato's chart (which as I noted uses improper sourcing anyway) doesn't actually support your claim, since it includes no entry for transit buses.
And in any event, the BTS chart alone is sufficient to prove you were simply wrong about it being impossible for transit buses to improve in fuel efficiency. Which is why your arguments about the future are obviously silly: you are assuming that automobiles will be gaining in efficiency while transit buses are not. The BTS chart demonstrates that transit buses can in fact improve in efficiency--and from 2000 to 2005, they did so much faster than automobiles.
So, your assumptions were wrong. And that is why you have to abandon the BTS chart--being wrong is something you simply cannot tolerate.
DTM,
If you didn't abandon the BTS chart, then why are you still saying things like "And transit buses are already much less fuel-efficient than autos"?
Er, because they are?
The BTS chart says different,
No it doesn't. The BTS chart only goes to 2005, is unreliable for single-year comparisons anyway, and has other problems. I explained all this to you in my post of 1:39pm. Did you miss it?
Cato's chart (which as I noted uses improper sourcing anyway) doesn't actually support your claim, since it includes no entry for transit buses.
How many times do we have to go over this? The Cato table breaks transit buses down into two categories: motor buses and trolley buses. Automobiles are more efficient than both types of bus. Passenger cars are more efficient still. And the Toyota Prius blows away transit buses on fuel efficiency.
And in any event, the BTS chart alone is sufficient to prove you were simply wrong about it being impossible for transit buses to improve in fuel efficiency.
I never said it is "impossible" for transit buses to improve in fuel efficiency.
Which is why your arguments about the future are obviously silly: you are assuming that automobiles will be gaining in efficiency while transit buses are not.
Transit buses are likely to gain in fuel efficiency in the future, but no more than, and probably much less than, private autos. Hybrid and all-electric technology can be applied to transit buses as well as cars, but that simply yields comparable efficiency improvements in both types of vehicle. It doesn't close the gap. And PHEV technology would be difficult to apply to transit buses because of their need to operate more or less continuously for many hours at a time.
Comparing the fuel efficiency of passenger cars to mass transit is an apples-to-oranges comparison. It’s misleading to compare the fuel efficiency of public transit travel, which mostly takes place under congested urban conditions, against some blanket national auto consumption average, much of which is based on driving in uncongested conditions. Mass transit is a not a subsitite for all types of driving; it's mainly a substitute for driving in congested urban areas, which, because it involves frequently sitting in traffic and lots of stops and starts, is much less energy-efficient than driving on average.
If you want to do a apples-to-apples comparision, you have to do something like compute the average automobile energy consumption rates per city and compare that value against rail/bus energy consumption in those respective cities.
The sad part is, the DTM/Mixner argument is mostly about trivia. In most of the US, the population is too dispersed in terms of living locations and destinations for mass transit to be even marginally useful.
The discussion of efficiency would matter a lot more if that weren't the case, but it is. And no matter how efficient busses or trains become, they don't solve the basic problem: the utter lack of usefulness for the vast majority of the population.
No, it's not low ridership. You misread your own quote. It's low load factors. Load factors in transit buses and trains will always be low because of the nature of the service they provide. The need to provide frequent service and to accommodate large variations in demand across different times of the day and different segments of their routes means that most seats will be empty most of the time.
I don't see who this pathetic argument from assertion is supposed to convince.
Intuitively, increased ridership should increase average load factor. During peak times, when buses are already full, additional passengers cause extra (fully-loaded) buses to be added--driving up the average load factor. During non-peak times, no additional buses need be run, but passengers are added at the margin--driving up the average load factor.
Nationwide, load factors are all over the map. NYC, LA and San Juan, PR average over 15 passengers/mile. Portland, OR gets around 10. Salt Lake City only gets 3.6. To argue that there's no room for improvement in national average load factor is just silly.
Some of those city-to-city differences probably correlate with area served and density, but that obviously just strengthens the case for transit-oriented development.
Comparing the fuel efficiency of passenger cars to mass transit is an apples-to-oranges comparison. It’s misleading to compare the fuel efficiency of public transit travel, which mostly takes place under congested urban conditions, against some blanket national auto consumption average, much of which is based on driving in uncongested conditions. Mass transit is a not a subsitite for all types of driving; it's mainly a substitute for driving in congested urban areas, which, because it involves frequently sitting in traffic and lots of stops and starts, is much less energy-efficient than driving on average.
We've been over this before, too. Most car travel is within the same urban and suburban environments where transit buses run. Intercity travel by car is only a small fraction of the total. Also, transit bus service on commuter and express routes serving suburbs involves significant highway driving rather than the congested city driving of inner-city bus routes. Even if we assumed all transit bus miles were "city" miles, and we substituted the "city" fuel economy rating for the "combined" rating for autos, autos would still be more fuel efficient than buses. For hybrid cars, the advantage is even greater, because hybrids typically get better "city" fuel economy than "combined."
jack lecou,
Intuitively, increased ridership should increase average load factor. During peak times, when buses are already full, additional passengers cause extra (fully-loaded) buses to be added--driving up the average load factor.
You are confused. Increased ridership would improve load factors where it could be satisfied from unused capacity in existing services, but if it required capacity to be added it may reduce load factors. For example, if your current peak-time bus runs on average at 80% of capacity (average peak load factor of 0.8) and demand increases by 50%, you would need to run an additional bus, and your average load factor would fall to 0.6.
During non-peak times, no additional buses need be run, but passengers are added at the margin--driving up the average load factor.
Huh? Off-peak demand is lower than peak demand. All else being equal, that means lower load factors at off-peak times. Transit authorities can and often do attempt to compensate for this by reducing the frequency of service at off-peak times, but the lower the frequency of service, the less viable transit becomes as an alternative to cars. At off-peak times where demand falls below a certain threshold--typically, nights and weekends--there may be no service at all. That makes transit even less viable as an alternative to cars. If you need travel at times when there is no transit service, and at times where service is infrequent, you are probably going to buy a car. And if you have a car for that reason, you're probably going to use it for most of your travelling even at times when the transit service is running frequently.
"For example, if your current peak-time bus runs on average at 80% of capacity (average peak load factor of 0.8) and demand increases by 50%, you would need to run an additional bus, and your average load factor would fall to 0.6."
Yes but if your demand increases by a more reasonable 25% your load factor is now 1. A more plausible scenario is that increasing oil prices cause modest increases in mass transit ridership, something like we are seeing now increases by modest percentages, and load factors move towards greater efficiency.
Yes but if your demand increases by a more reasonable 25% your load factor is now 1.
I'm not sure why you think a 25% increase in demand is "more plausible" than a 50% one. Or some other number. It depends on the timescale under consideration and a bunch of other factors. In any case, if your current load factor is greater than 0.8 you'll need to run an additional bus to accommodate even a 25% increase in demand.
And virtually no transit route would be feasible with an average peak-time load factor anywhere close to 1 anyway. They need to maintain substantial excess capacity above the average load factor to accommodate variations in demand on different days and at different points along the route. Otherwise they'd frequently be leaving lots of angry customers waiting at the bus stop or train station because the bus/train is full.
A more plausible scenario is that increasing oil prices cause modest increases in mass transit ridership, something like we are seeing now increases by modest percentages, and load factors move towards greater efficiency.
That's not plausible at all. Transit system capacities are generally sized at the minimum needed to effectively satisfy peak demand. Even a small increase in peak demand would likely require an increase in capacity (more or bigger buses and trains), which is likely to reduce average load factors. This is why cash-strapped local governments are struggling to meet even the small increase in demand that recent gas price increases have created.
I don't think I'm the one confused here.
Say we have a bus route on which we run 20 buses for the morning and afternoon rushes, all mostly packed, call it 80%, if you like.
We also run buses every half hour or so during the day. Another 15 or so round trips. Let's say these are all at maybe 5% load factor.
Average load factor on the line is 47.86%.
Now let's say ridership increases 5%.
At 80% peak load, a 5% increase in peak ridership could be absorbed easily without adding any buses, but let's be really conservative and add an extra bus at both peaks. (The handful of extra passengers during the day we won't worry at all about, since the daytime buses are 95% empty.)
Now our average load factor is over 48.75%.
48.75% is bigger than 47.86%.
And we could easily have just let the peak load increase to 84% without adding buses. Then we'd be up to over 50%.
Any extra non-peak passengers are bonus, increasing non-peak load to, say 5.5%, for another couple tenths of a point on overall average load.
This is a contrived example, but it seems pretty generalizable. Adding passengers to underutilized non-peak runs increases load factor unconditionally. Adding passengers to more or less fully loaded rush hour runs increases load factors as long as any added buses end up at least as full as the previous overall average--easily done, since peak loads are by definition very high while average loads are generally fairly low.
Mixner,
I'm still waiting for your explanation of the trend in the BTS chart. As I noted previously, after a long period of stable numbers above 4000 during most of the 1990s through 2000, the numbers started trending down starting in 2001.
Again, if you want to claim this is statistically insignificant, provide your calculations. Otherwise, provide an explanation of how this happened.
jack lecou,
I might note that in my city, there has also recently been a "right size" program, where smaller buses have been allocated to routes where ridership is insufficient for bigger buses. That is a simple way to increase loads, and I suspect a part of the explanation for the recent gains in fuel efficiency according to the BTS measure. Of course that requires some additional vehicles (to the extent you will now leave some bigger buses idle outside peak times and swap to smaller buses), but higher diesel prices obviously make the payback period on those additional vehicles much shorter.
DTM,
I'm still waiting for your explanation of the trend in the BTS chart.
I don't know why. What do you think I need to "explain," exactly, and why?
As I noted previously, after a long period of stable numbers above 4000 during most of the 1990s through 2000, the numbers started trending down starting in 2001.
What numbers?
Again, if you want to claim this is statistically insignificant,
I didn't claim it was statistically insignificant.
If you have an actual, you know, argument to make, then make it.
James Robertson,
I agree many, indeed most, transit needs in the United States will not be fulfilled with local public transit in the conceivable future. Again, though, as fuel prices increase, the number of needs that can be most efficiently served by local public transit also increases. Of course there is no contradiction between those two propositions.
Mixner,
I'll try to make this very simple.
Look at the BTS chart I linked above. Notice the stable period in BTUs/passenger mile for transit buses for most of the 1990s through 2000. Notice the downward trend starting in 2001. Notice also that BTUs/passenger mile also trended down for passenger cars in this period. But finally notice that BTUs/passenger mile trended down faster for transit buses than passenger cars in this period.
Here is one hypothesis to explain all this: in response to the trend of higher fuel prices during roughly the same period, both transit bus authorities and passenger car users attempted to improve fuel efficiency. But transit bus authorities were able to respond more effectively in this period.
Now if you agree with that hypothesis, fine. If you have an alternative hypothesis, provide it. If think this is just statistical variation, provide your calculations.
I'm waiting.
Jack Lecou,
I think you've screwed up your math, but since you don't show your calculations it's hard to be sure. By my calculations, your scenario yields a lower load factor, not a higher one.
In any case, in general, a significant increase in peak demand is likely to lower the average load factor, because it will probably require additional capacity to satisfy.
DTM,
I might note that in my city, there has also recently been a "right size" program,
What city is that, and where may I read about this "right size" program? Do you have any links?
where smaller buses have been allocated to routes where ridership is insufficient for bigger buses. That is a simple way to increase loads,
This seems rather implausible. Give us some examples of these routes. What were/are the bus sizes and loads before and after the change?
A thread that gets trolled by Mixner and James Robertson eventually gets so bent out of shape that the very laws of the physical universe cease to apply.
According to Mixner, for example, cars will always be more efficient than rail transit. Never mind that steel wheels on steel rails with gradients under 5%, running on energy supplied by wires, are propelling 100 people with the same horsepower used by a car to carry four- cars will always be more efficient than rail, according to Mixner. And he's not suggesting this to you, he's telling you. By golly, when he makes a comment you darn well better memorize what he says, or you will be reminded later that he already told you that.
What is this mysterious power that has suspended the laws of thermodynamics and physics? Can it be that the simple desire to live in a suburban home and shop at the mall is the occult key the Egyptians and the Aztecs failed to find, thus dooming their empires to extinction? Is this the inner meaning of the monoliths in 2001- A Space Odyssey?
Or is this all just a cruel joke by a sardonic Creator, who can post four lines about transit and watch Mixner type all afternoon in the hopes of persuading 0.05% of the readers who actually read the comments?
The Creator, in all probability, is not sardonic, but simply vexed, as are we all, but Mixner still stands as good practice for the rest of us, who will meet "that guy" in any local discussion about transit. Eventually the people who have never thought much about it will make better decisions than people like Mixner who have spent so much time reading literature from Cato.
It's the miracle of democracy. Who woulda thunkit?
DTM,
It doesn't really make sense to "agree" or "disagree" with a hypothesis. There are obviously many possible explanations for the numbers in the chart. The volatility of the transit bus data suggests that it is not reliable. The fact that the source of the data is APTA rather than a government energy or transportation agency casts further suspicion. Without more information, there's no basis for any conclusions. And this is all irrelevant anyway to the central point that autos are already more fuel-efficient than transit buses, and new-technology cars like the Prius are vastly more fuel-efficient.
Wow, Mixner, that is totally awesome- "there's no basis for any conclusions"- except, of course, the conclusion you've already drawn.
Too bad you can't use your powers of omniscience for some useful purpose, like writing fortune cookie inserts that actually tell the future.
Initial scenario, 40 peak buses, 30 non-peak buses, 80% & 5% load, respectively:
(40*80+30*5)/(40+30)~=47.86
5% increase in peak riders, matched by 5% increase in peak buses:
(42*80+30*5)/(42+30)=48.75
5% increase in peak riders, absorbed by existing capacity:
(40*84+30*5)/(40+30)~=50.14
And it should be obvious that unless non-peak ridership increases by 1000%-2000%, additional riders in those hours are pure profit.
And it should be obvious that unless non-peak ridership increases by 1000%-2000%, additional riders in those hours are pure profit.
(Hmm. In case that's unclear, a 2000% increase in day and evening ridership is also profit, it might just require a few more buses.)
Jack Lecou,
You said: "Say we have a bus route on which we run 20 buses for the morning and afternoon rushes" and "We also run buses every half hour or so during the day. Another 15 or so round trips."
(20*80+15*5)/(20+15) = 47.86
You said: "a 5% increase in peak ridership ... let's be really conservative and add an extra bus at both peaks."
(20*84+15*5)/(22+15) = 47.43
jack lecou,
And it should be obvious that unless non-peak ridership increases by 1000%-2000%, additional riders in those hours are pure profit.
What's obvious is that you really don't understand how bus routes work. First, 0.8 isn't a remotely realistic average peak-time load factor. This is because demand varies dramatically between different points along a route and between different days. At some points along a route, there will likely be only a handful of people on the bus. At others--bus stops close to popular destinations such as offices, schools, shops, transfer points, etc.--the bus will likely be almost full. At rush hour, it's likely to be standing room only, in fact. The bus has to be large enough to accommodate these spikes in demand or lots of angry people are going to be left behind at bus stops because the bus is full. But that means the bus is going to have lots of empty seats at less popular points on its route, and the average load factor for the route will be low. And at off-peak times, unless the frequency of service is reduced dramatically, the load factor will be even lower.
In order to provide dependable service that is both capable of accommodating spikes in demand at peak periods and also frequent enough at non-peak times to still be a viable alternative to car travel, buses must be sized and operated such that their average load factor is low. This problem is independent of the total volume of passengers the system serves.
20 buses in the morning and afternoon equals 40 buses.
But whatever. Doesn't make a whit of difference with the math.
20 peak buses, 15 off-peak, 5% ridership increase, same scenarios otherwise:
A: (20*80+15*5)/(20+15)=47.86
B: (21*80+15*5)/(21+15)=48.75
C: (20*84+15*5)/(20+15)=50.14
80% is unreasonably large? You're probably right, but it doesn't make any difference. How about a 40% peak?
A: (40*40+30*5)/(40+30)=25
B: (42*40+30*5)/(42+30)=25.42
C: (40*42+30*5)/(40+30)=26.14
Still don't like it? Let's try 20%:
A: (40*20+30*5)/(20+30)=19
B: (42*20+30*5)/(22+30)=19.04
C: (40*21+30*5)/(20+30)=19.8
...Astonishing, I know.
This problem is independent of the total volume of passengers the system serves.
It most certainly is not. In addition to the amazing mathematical phenomenon I have illustrated above, the law of large numbers is a route planner's friend.
As ridership volume increases, not only does the flow become more reliably predictable, but more tools, like express routes, become available to smooth out spikes along the route and throughout the day.
jack lecou,
20 buses in the morning and afternoon equals 40 buses.
You didn't say that. You said: "a bus route on which we run 20 buses for the morning and afternoon rushes." If you can't articulate your scenario clearly, that's not my problem.
20 peak buses, 15 off-peak, 5% ridership increase, same scenarios otherwise:
A: (20*80+15*5)/(20+15)=47.86
B: (21*80+15*5)/(21+15)=48.75
Your B is wrong. You said "add an extra bus at both peaks." The daily capacity increases by 2 buses. So the denominator is (22+15), not (21+15). The calculation is therefore:
B: (21*80+15*5)/(22+15)=47.43
Particularly bizarre that you would object to a 1000% increase in off peak ridership. This would have not only the effect of bringing off-peak loads more in line with peak loads, it would also necessitate roughly doubling the number of off-peak buses - increasing the frequency of buses throughout the day and thus the utility and reliability of the system.
Like I said, pure profit.
jack lecou,
80% is unreasonably large? You're probably right, but it doesn't make any difference. How about a 40% peak?
You seem not to have understood the problem I described in my post of 9:18pm. The limiting factor is not average peak-time load factor, but peak peak-time load factor. The bus needs to be big enough to accommodate not merely the average peak-time demand along its route, but the peak peak-time demand. Otherwise, at those peak locations along the route, more people will be waiting to board the bus than can be accommodated, and the driver will have to turn them away because the bus is full. Of course, in practise this sometimes happens anyway. But transit planners try to avoid it by making sure buses are large enough to accommodate the number of passengers seeking to ride on the bus at the busiest point along its route at the busiest time of day. And that means that average peak-time load factor will be lower. So the mere fact that there is unused capacity above the average peak-time load factor doesn't mean that increased demand could be satisfied without running more buses. And running more buses is likely to lower the average load factor.
ERM | June 26, 2008 12:36 PM:
Of course, this is mostly moot. This country has a traditional urban design fabric, which makes the transit/pedestrian intensive lifestyle relatively feasible. Most of the U.S. -- 95% of it west of the Allegheny Mountains -- is just not designed at present to snap this kind of infrastructure on top of it without massive upheaval. All the worse for us, as far as I can tell.
Where I live, in the outer suburbs of Northeast Ohio, it could be added quite quickly.
What is needed is dropping the "one size fits all" mindset that afflicts both sides of the debate and take a realistic look at the population and what would be required to make the option of a pedestrian/transit lifestyle feasible.
To have the option widely available, there is no need for everyone in the US to "move into densely populated large cities". Rather, dedicated transport corridors ... electric rail or electric light rail or electric trolley bus, depending on available patronage for that route ... with stops, and zoning and infrastructure support for walkable development in the quarter mile around the dedicated stop ... that'd do it.
That gives a walkable zone of about 1/5 of a square mile around each stop. If these stops are five miles apart, on average, that creates a corridor 3 miles wide where everyone lives within 3 miles of a stop, well within the "golf cart" range, or cycle range, or range of a short shuttle bus loop, or range of a on demand "transit cab". Chopping off the outer arc of the corridor and treating it as a corridor 3.2mi wide, each station serves 16 sq. mi, with 0.19 in the walkable zone at the core. If the infill development in the walkable zone is four times the average suburban density, that means 0.76/16.57 or 4.6% of the population can live in the walkable core ... and if the demand is greater than that, simply establishing a crossing route that connects to the stop on the dedicate corridor running to additional "suburban villages" can easily double or triple that share.
If instead of looking on suburbs as a moral outrage, we look on them as the raw material that we happen to have on hand, there are design solutions to the problems we face.
jack lecou,
Particularly bizarre that you would object to a 1000% increase in off peak ridership.
I don't object to it at all. I'm saying that even if ridership increases by as little as, say, 5% (to use your number), it is very unlikely that all those new riders will show up only at non-peak times when there's plenty of room for them. A lot of them are going to want to ride at rush hour, when the buses are already likely to be running close to capacity at busy points along their routes like offices and schools and shops. So you'll probably have to run more buses to handle the extra demand.
You didn't say that. You said: "a bus route on which we run 20 buses for the morning and afternoon rushes." If you can't articulate your scenario clearly, that's not my problem.
An 'each' may have been erroneously pruned somewhere during editing. We could focus on that, or the actual argument, now clarified. Which will it be?
Your B is wrong. You said "add an extra bus at both peaks." The daily capacity increases by 2 buses. So the denominator is (22+15), not (21+15). The calculation is therefore:B: (21*80+15*5)/(22+15)=47.43
Umm, well, if you want to be petty, I'll point out that increasing the peak buses by 2 increases the denominator, but also increases the buses in the numerator while decreasing the peak load:
B: (22*76.4+15*5)/(22+15)~=47.43
Which is still a drop, but not much. Not all that surprising, really, since we added 10% more buses to cope with only 5% more riders...
(Obviously adding 2 buses to 40 is 5%, which is why I chose 20 buses at each peak in the first place. It also illustrates one of the many ways that larger volumes can more efficient.)
Anyway, how about addressing the actual argument now, rather than arguing with an obvious typo?
Anyway, how about addressing the actual argument now, rather than arguing with an obvious typo?
I think I've addressed it at length in my posts of 9:18pm, 10:31pm and 10:40pm. If you still don't understand the problem with your analysis, never mind.
Mixner,
You write:
"There are obviously many possible explanations for the numbers in the chart."
But back when you only had the 2000 numbers, you thought you knew the explanation for the numbers. Curious that now it has become so mysterious.
"The volatility of the transit bus data suggests that it is not reliable."
This sounds like a statistical argument. If so, show your calculations. Otherwise, it is begging the question to dismiss the change that you predicted couldn't happen as mere "volatility".
"The fact that the source of the data is APTA rather than a government energy or transportation agency casts further suspicion."
And yet back when you thought this chart said what you wanted it to say, this wasn't a problem.
"Without more information, there's no basis for any conclusions."
And yet you were drawing all sorts of conclusions when you had less information. Again, quite curious.
"And this is all irrelevant anyway to the central point that autos are already more fuel-efficient than transit buses."
Except the latest version of the chart you once were citing contradicts this claim. Other than that, right, it is irrelevant.
"What city is that . . . ."
Yeah, I don't think I am going to tell you that. That comment wasn't directed to you in any event.
You seem not to have understood the problem I described in my post of 9:18pm. The limiting factor is not average peak-time load factor, but peak peak-time load factor. The bus needs to be big enough to accommodate not merely the average peak-time demand along its route, but the peak peak-time demand.
My goodness! What a devastating argument! This insight makes it clear to me that buses are obviously inferior to cars. I sure feel dumb now.
If only someone had thought of a solution to this problem and it had been posted earlier in this thread. That would have saved us the embarrassment of being schooled by Mixner.
Oh...wait a minute...
...there has also recently been a "right size" program, where smaller buses have been allocated to routes where ridership is insufficient for bigger buses. That is a simple way to increase loads...you will now leave some bigger buses idle outside peak times and swap to smaller buses...
What's the phrase I'm looking for? Oh yeah:
Try again!
DTM,
But back when you only had the 2000 numbers, you thought you knew the explanation for the numbers.
No, I never claimed to have an "explanation" for the numbers in that chart. The chart is irrelevant to the central point that cars are more efficient than buses.
Yeah, I don't think I am going to tell you that.
Since you're obviously just making up this "right size program" nonsense, that isn't terribly surprising.
DMonteith,
I'm pretending to believe that you're trying to make a serious argument here.
Replacing a bus that is already full at peak locations on its route with a smaller bus isn't going to work.
Buses are, of course, a red herring ... certainly, regular street buses are an efficient way to stretch the reach of dedicated transport corridors, but since federal policy has been to starve public transport compared to the subsidies to loss making road and air transport, we do not anything approaching the fuel efficiency in our public bus fleet that we could have, if it was national policy to have a fuel efficient public bus fleet.
Add in the obvious fact that its the marginal fuel consumption, not the average fuel consumption, that determines the impact of adding patronage to buses, and the marginal fuel consumption is lowest in precisely those areas where the capacity on the road is driven by the need to keep the bus route system open as a transport network ... which is where the average fuel consumption per passenger is the highest.
And of course, it has been national policy, within the policy of starving public transport relative to loss-making road and air transport, to starve rail relative to buses ... while it is electric rail that offers the truly low marginal energy consumption per additional passenger carried.
And under investment in dedicated transport corridors means a loss of patronage drivers for buses, which means lower patronage, which means a higher average fuel consumption by the buses per passenger.
Replacing a bus that is already full at peak locations on its route with a smaller bus isn't going to work.
Oh. My. Fucking. God.
I knew you were stupid, but I think you've just proven that you're the Mount Everest of stupid.
The smaller buses run during off peak times and the larger ones run during peak times so that the average load factor for the entire fleet stays high and efficiency is maximized. Tell me if I'm going too fast for you here, though I'm not sure if it's possible to simplify it any more.
I'm pretending to believe that you're trying to make a serious argument here.
It's not even my argument. It's a transit policy that I've seen applied to subways (Boston, London, Miami) whereby the later evening trains consist of considerably fewer cars (run at greater intervals) than rush hour trains. It's a mind numbingly simple solution to the problem that you've been resting your case on. I'm just pointing it out because you seem to have missed it the first time it came around upthread. Now you've missed it the second time.
Third time's a charm?
Since you're obviously just making up this "right size program" nonsense, that isn't terribly surprising.
Not his city as far as I know, but mine:
http://tdp.portauthority.org/paac/portals/1/pdfs/PeerReview.pdf
I think the meme that "transit system" is sort of a black box is as prevalent as "city" connotates squalor and crime. Not particularly useful (but profitable for some) images can unfairly generalize what these things mean in a lot of people's minds.
Some cities and transit systems do it better than others, and progress in arranging transit systems is no deader than innovations in hybrid technology, GPUs, stuff you can do to food, or porn.
I know that some people cannot accept that, and that is your religious belief, but others hold the defensible view that they can. Investing in transit systems where they work is certainly no goofier than building a sprawling metropolis in the the middle of a desert.
The smaller buses run during off peak times and the larger ones run during peak times so that the average load factor for the entire fleet stays high and efficiency is maximized.
The obvious problem with this scheme is that it greatly increases capital costs. Buying two or more fleets of buses for a route and idling each fleet for a large part of the day is obviously much more expensive than buying a single fleet for use during all service periods. It is especially inefficient for the large-bus fleet, which would presumably be idle at all times except rush-hour periods on weekdays. Storage and maintenance costs would also likely increase greatly. Smaller vehicles also tend to have higher seat-mile costs than larger ones, which would likely wipe out part or all of the fuel-efficiency savings from the higher average load factor. There may also be mobility issues. Larger buses are more accessible by disabled and elderly people, who are a key market for transit buses. Some substitution of smaller buses for larger ones at off-peak times may be justified on some routes, but if this were a cost-effective and practical way of significantly increasing load factors on general transit bus routes it would already be in widespread use.
Has none of this occurred to you, moron?
You seem not to have understood the problem I described in my post of 9:18pm. The limiting factor is not average peak-time load factor, but peak peak-time load factor. The bus needs to be big enough to accommodate not merely the average peak-time demand along its route, but the peak peak-time demand.
I am, of course, well aware of this. And it is obviously irrelevant to your point (or, if anything, weakens it).
It doesn't matter what the particular shape of the load curve is along a particular route at a particular time of day. Go ahead and make one up. You'll notice it has an average...80%, 40%, 20%. Something. Plug that in above. Weird, huh?
You see, peak load levels are by definition higher than off peak load levels. So by adding buses, as long as they're operating at peak-like load levels--or indeed, merely above average load levels--will increase the system-wide average. That's not rocket science.
(Admittedly, I'm assuming that the increase is not pathologically chaotic, the entire 5% for the month falling on a random Tuesday or something. That would be a problem. But in general, we'd expect the increase to be more or less spatially and temporally uniform throughout the system.)
And notice that the fact that load varies along the route does not exactly strengthen your case. More riders not only increases system-wide average load, it also increases the possibilities for route improvement. For example, suppose we have stops:
A-b-c-D-e-F-g-h-I-j-K-L
Where capital letters denote stops with significantly higher traffic than lower case stops.
Now, if our ridership increases enough, we'll have to add buses. It might be that we'll now have enough riders and buses to justify pulling a bunch of buses off the small stops entirely, and making an express route.
Not only will service improve, but buses on both the local and express halves of the route will see their load curves flatten out, allowing them to increase average load without being in danger of exceeding maximum load.
Or suppose we add so many buses, that buses now come every 5 minutes during rush hour, instead of every 20.
Then, not only is service radically improved, but also we can afford to reduce our load factor safety margins. Exceeding capacity every now and then now has much reduced consequences for people waiting at stops.
robotic,
Not his city as far as I know, but mine:
http://tdp.portauthority.org/paac/portals/1/pdfs/PeerReview.pdf
I see nothing in that document about a "right size program," or indeed any discussion at all about changing transit vehicle sizes. Perhaps you could direct me to that discussion if you think there is one.
Some substitution of smaller buses for larger ones at off-peak times may be justified on some routes, but if this were a cost-effective and practical way of significantly increasing load factors on general transit bus routes it would already be in widespread use.
Except, as I think was already pointed out, but in any case should be obvious, this may be an efficient response to higher fuel costs. We don't see this much yet because until recently capital and labor costs were certainly dominant. Now we may be entering a regime in which fuel costs are more dominant, and load rate maximization is a higher priority.
jack lecou,
You see, peak load levels are by definition higher than off peak load levels. So by adding buses, as long as they're operating at peak-like load levels--or indeed, merely above average load levels--will increase the system-wide average. That's not rocket science.
Huh? The whole point is that they likely would not be operating at peak-like load levels. Increases in demand do not generally come in convenient bus-sized increments. They come in arbitrary-sized increments. If you have to add extra buses because your existing capacity is slightly too small to accommodate the increase in demand, your average load factor is most likely going to go down.
And notice that the fact that load varies along the route does not exactly strengthen your case. More riders not only increases system-wide average load, it also increases the possibilities for route improvement.
It's hard to make any clear sense of this. If by "average load" you mean "average load factor," the claim is simply false. If you mean "average number of passengers per bus," it's still false. The effect of more riders on both average load factor and average vehicle occupancy depends on all sorts of variables. You simply can't predict the effect without more information.
Now, if our ridership increases enough, we'll have to add buses. It might be that we'll now have enough riders and buses to justify pulling a bunch of buses off the small stops entirely, and making an express route. Not only will service improve, but buses on both the local and express halves of the route will see their load curves flatten out, allowing them to increase average load without being in danger of exceeding maximum load.
Yet another bizarre claim. If you're pulling buses off the local stops you're obviously cutting service to the people who use those stops. And the effect on "load curves" would depend on the precise relationship between demand and the changes in capacity provided by your new/changed routes.
Or suppose we add so many buses, that buses now come every 5 minutes during rush hour, instead of every 20. Then, not only is service radically improved, but also we can afford to reduce our load factor safety margins.
But you've also quadrupled the number of buses and bus drivers you need. And massively increased all the other costs associated with buying, operating, maintaining and storing your bus fleet, since you now have four times as many buses.
Is there a point to all this? You now seem to be just throwing out random ideas, "Suppose we do this...." "Suppose we do that...." without thinking them through or relating them to any clear goal.
Buying two or more fleets of buses for a route and idling each fleet for a large part of the day is obviously much more expensive than buying a single fleet for use during all service periods.
But lifetime fleet costs are unaffected, Einstein.
Running one bus all day every day means that that bus must be maintained and replaced more often than it would if it ran only part of each day. Total capital depreciation is conserved. Up front costs would be higher but are easily amortized. The only significant further cost would be storage/ parking space but I doubt that's a deal breaker.
And I presume that you agree that gains from maximizing average load would be significant since you've been arguing all along that low average loads are busing's achilles heel. Surely even small improvements in the huge problem of average load would be more than enough to cover the cost of parking and a marginal increase in finance charges, no? If not, then this whole average load thing must not be as big a deal as you claim.
So are you wrong now, or were you wrong a few hours ago?
But lifetime fleet costs are unaffected, Einstein.
No they're not, missy.
Running one bus all day every day means that that bus must be maintained and replaced more often than it would if it ran only part of each day. Total capital depreciation is conserved. Up front costs would be higher but are easily amortized.
Nonsense. Maintenance and capital depreciation costs depend on the age of a vehicle as well as its number of hours in operation. The return on capital for large buses especially would be greatly reduced by idling them for all but a few hours of the day. Your fleet would also much more likely be rendered obsolete before its capital costs were amortized if you kept it twice as long but ran each bus only half as often. And the upfront capital costs of buying twice or more as many buses to service the route would incur a huge additional penalty in financing costs.
The only significant further cost would be storage/ parking space but I doubt that's a deal breaker.
It would be yet another huge expense on top of the additional capital and maintenance costs. There would also likely be additional fleet management costs. The Los Angeles Metro Local bus system operates thousands of transit buses. Doubling or tripling the number of buses in that system would create a logistical, storage and maintenance nightmare.
And I presume that you agree that gains from maximizing average load would be significant
No, I do not agree. You wouldn't be "maximizing average load," anyway. At best, you'd merely be increasing load factor. The magnitude of that increase would depend on the pattern of variation in demand throughout the day, the ratio of the vehicle sizes in your two bus fleets, the effect of changing bus sizes on demand, and all sorts of other variables. Again, to the extent that operating differently sized vehicles at different times of the day is a practical and cost-effective way of accommodating variations in demand, transit operators are already using it. The fact that there is so little of it demonstrates that it rarely makes sense.
Huh? The whole point is that they likely would not be operating at peak-like load levels. Increases in demand do not generally come in convenient bus-sized increments. They come in arbitrary-sized increments. If you have to add extra buses because your existing capacity is slightly too small to accommodate the increase in demand, your average load factor is most likely going to go down.
My reply is below.
Let's hope it settles this silly aspect of the argument once and for all.
Mixner - I will be happy to re-run this with any load factor and initial bus fleet assumptions you care to give me. The results will be the same (obviously for very small initial number of buses, load factor does increase only in fits and starts):
INC = percent increase in ridership, uniform over system
PB = # of peak buses (# offpeak buses is 47 on the whole run)
PLF = average load factor of peak-time buses (whenever it exceeds 30, we add a bus)
NPLF = average load factor of non-peak buses
ALF = system-wide average load factor
INC PB PLF NPLF ALF 0% 47 30.0% 5.0% 17.5% 1% 48 29.67% 5.05% 17.49% 2% 48 29.96% 5.1% 17.66% 3% 49 29.64% 5.15% 17.65% 4% 49 29.93% 5.2% 17.82% 5% 50 29.61% 5.25% 17.81% 6% 50 29.89% 5.3% 17.98% 7% 51 29.58% 5.35% 17.96% 8% 51 29.86% 5.4% 18.13% 9% 52 29.56% 5.45% 18.11% 10% 52 29.83% 5.5% 18.28% 11% 53 29.53% 5.55% 18.26% 12% 53 29.8% 5.6% 18.42% 13% 54 29.51% 5.65% 18.4% 14% 54 29.77% 5.7% 18.57% 15% 55 29.48% 5.75% 18.55% 16% 55 29.74% 5.8% 18.71% 17% 55 29.99% 5.85% 18.87% 18% 56 29.71% 5.9% 18.85% 19% 56 29.96% 5.95% 19.01% 20% 57 29.68% 6.0% 18.98% 21% 57 29.93% 6.05% 19.14% 22% 58 29.66% 6.1% 19.11% 23% 58 29.9% 6.15% 19.27% 24% 59 29.63% 6.2% 19.24% 25% 59 29.87% 6.25% 19.4% 26% 60 29.61% 6.3% 19.37% 27% 60 29.85% 6.35% 19.52% 28% 61 29.59% 6.4% 19.5% 29% 61 29.82% 6.45% 19.65% 30% 62 29.56% 6.5% 19.62% 31% 62 29.79% 6.55% 19.77% 32% 63 29.54% 6.6% 19.74% 33% 63 29.77% 6.65% 19.89% 34% 63 29.99% 6.7% 20.04% 35% 64 29.74% 6.75% 20.01% 36% 64 29.96% 6.8% 20.15% 37% 65 29.72% 6.85% 20.12% 38% 65 29.94% 6.9% 20.27% 39% 66 29.7% 6.95% 20.23% 40% 66 29.91% 7.0% 20.38% 41% 67 29.67% 7.05% 20.35% 42% 67 29.88% 7.1% 20.49% 43% 68 29.65% 7.15% 20.46% 44% 68 29.86% 7.2% 20.6% 45% 69 29.63% 7.25% 20.56% 46% 69 29.83% 7.3% 20.7% 47% 70 29.61% 7.35% 20.67% 48% 70 29.81% 7.4% 20.81% 49% 71 29.59% 7.45% 20.77% 50% 71 29.79% 7.5% 20.91% 51% 71 29.99% 7.55% 21.05% 52% 72 29.77% 7.6% 21.01% 53% 72 29.96% 7.65% 21.15% 54% 73 29.75% 7.7% 21.11% 55% 73 29.94% 7.75% 21.25% 56% 74 29.72% 7.8% 21.21% 57% 74 29.91% 7.85% 21.34% 58% 75 29.7% 7.9% 21.3% 59% 75 29.89% 7.95% 21.44% 60% 76 29.68% 8.0% 21.4% 61% 76 29.87% 8.05% 21.53% 62% 77 29.66% 8.1% 21.49% 63% 77 29.85% 8.15% 21.62% 64% 78 29.65% 8.2% 21.58% 65% 78 29.83% 8.25% 21.71% 66% 79 29.63% 8.3% 21.67% 67% 79 29.81% 8.35% 21.8% 68% 79 29.98% 8.4% 21.93% 69% 80 29.79% 8.45% 21.89% 70% 80 29.96% 8.5% 22.02% 71% 81 29.77% 8.55% 21.98% 72% 81 29.94% 8.6% 22.1% 73% 82 29.75% 8.65% 22.06% 74% 82 29.92% 8.7% 22.19% 75% 83 29.73% 8.75% 22.14% 76% 83 29.9% 8.8% 22.27% 77% 84 29.71% 8.85% 22.23% 78% 84 29.88% 8.9% 22.35% 79% 85 29.69% 8.95% 22.31% 80% 85 29.86% 9.0% 22.43% 81% 86 29.68% 9.05% 22.39% 82% 86 29.84% 9.1% 22.51% 83% 87 29.66% 9.15% 22.47% 84% 87 29.82% 9.2% 22.59% 85% 87 29.98% 9.25% 22.71% 86% 88 29.8% 9.3% 22.66% 87% 88 29.96% 9.35% 22.79% 88% 89 29.78% 9.4% 22.74% 89% 89 29.94% 9.45% 22.86% 90% 90 29.77% 9.5% 22.81% 91% 90 29.92% 9.55% 22.93% 92% 91 29.75% 9.6% 22.89% 93% 91 29.9% 9.65% 23.01% 94% 92 29.73% 9.7% 22.96% 95% 92 29.89% 9.75% 23.08% 96% 93 29.72% 9.8% 23.03% 97% 93 29.87% 9.85% 23.15% 98% 94 29.7% 9.9% 23.1% 99% 94 29.85% 9.95% 23.22% 100% 94 30.0% 10.0% 23.33%
(Here's hoping the pre tag works in more than just preview...)
I feel like I'm in the Monty Python argument sketch.
Your disagreement with any one of my points would carry much more weight if you didn't disagree with everything anyone's ever posted on this blog.
You seem to be in such a convulsively contrarian state that you're not even bothering to check your arguments for logical consistency. To wit:
Maintenance and capital depreciation costs depend on the age of a vehicle as well as its number of hours in operation...Your fleet would also much more likely be rendered obsolete before its capital costs were amortized if you kept it twice as long but ran each bus only half as often.
Which is it? Is my fleet going to fall apart just as quickly as it would otherwise, or is it going to last so long as to lapse into obsolescence before decommissioning? And besides, what exactly is an obsolete bus? One with only 16 megs of RAM? Also, if I kept the fleet twice as long wouldn't I be able to spread the larger up front cost over the same longer time frame, at the cost of a marginally higher interest rate (i.e. a "huge" couple of percentage points)? Last time I checked, that's just what amortization means. Also, wouldn't a larger order from the manufacturer and a larger contract for parts lead to bigger discounts per vehicle that would at least partially defray the finance penalty?
Doubling or tripling the number of buses...
Who said anything about tripling? You seem to think that Jack Lecou is advocating that but you'd need to take that up with him. I'm not convinced that even a doubling would be necessary because there are presumably at least some routes that exhibit less bipolar or less predictable use patterns that would not benefit from this regime.
But, whatever. I eagerly await your utterly predictable rejection of everything enumerated above. I'm tempted to make a strong claim that the sky is blue just to see you rebut my foolish delusion.
"Even if you assume that seniors will be packed in some kind of senior living center, then a dedicated bus for said center (which ten is paid for as part of the fee for living there) makes more sense."
Hm, I live in a university town with a relatively large number of "retirement villages", or appartment complexes where old people live (some like to live close to young people, apartments are rented mostly by students). The local transit autority has some buses and routes dedicated to the needs of the seniors, who would rather go to a strip mall or the shopping mall (we have but one shopping mall) than to the university campus -- where all other transit lines converge.
Since the authority invested in a fleet of buses running on LNG, it basically removed other operators -- it runs routes that were formerly served by the university buses or buses of individual apartment complexes. Buses for seniors are smaller and equipped for wheelchairs.
When you are converting vehicles to natural gas or electric, buses are more economic than cars. And a transit authority can be more efficient than individual operators (at least in a relatively rural area, where union labor is not all that expensive).
Talking about electric, it makes more sense to have electric buses than cars. Buses have predictable routes, so there is no problem with limited range, they can accommodate flywheels, or you even can have trolleys.
Some remarks about "two fleets of buses".
One. Off-peak buses would be vans, 10-12 passanger vehicles, so they would cost considerably less than buses with 30 sitting and 70 standing places. We are increasing the capital costs perhaps by 25%.
Two. Wear and tear of big buses would go down, so the number of break jobs etc.
Three. Fuel costs would be down, as this is the whole topic. Suppose that each day we drive 100 miles at 20 mpg rather than 5, then we use 5 gallons rather than 20, so we save 15, nearly 4000 gallons per year. Suppose that a van costs 40 thousand dollars, and will be amortized in 4 years.
My numbers are out of thin air, but the guesses are reasonable.
About obsolete buses: if a bus is in good shape, you can replace engine etc. at the cost that is much lower than the cost of a new bus.
If you pick through Mixner's reply to the right-sizing idea, it comes down to the argument that if it really could work, it would already have been done.
The basic problem with this argument is that obviously there is a tradeoff involved: adding more buses to serve the same number of passenger-miles will require some sort of additional sunk cost, but in turn the benefit will be savings on fuel during operations. Given that tradeoff, obviously as fuel costs rise, those sunk costs won't be increasing, but the benefits will. So, there is no reason to assume that if adding a smaller bus would make sense at today's fuel prices, it would already have been done at the much lower fuel prices in the past.
The bottomline is that Mixner will never admit he is wrong, so it isn't really worth trying to get him to do that. But for the rest of us, I think the BTS chart is actually pretty informative, and encouraging.
Specifically, it appears to be true that during a period of declining fuel prices, transit buses were being used in a fuel-inefficient manner, specifically by running large buses at low loads. But conversely, this represented an area in which the fuel efficiency of public transit could be improved. And that is apparently exactly what happened between 2000 and 2005--transit authorities managed to respond quickly to increasing fuel prices with fuel efficiency increases. And I strongly suspect that when we are looking at the 2006-2010 continuation of this series, we will see even more such gains, which again is good news.
But, of course, Mixner doesn't want there to be any good news regarding public transit. So, he will deny all this is even possible, even thought it has already been happening.
Mr. Mixners' claim that busses are more polluting then cars is probably true for diesel powered buses. However, natural gas powered busses, such as have been purchased by WMATA are far less polluting then either. WMATA now operates some 400+ natural gas powered vehicles, approximately 30% of their fleet.
As an interim step toward both reducing oil consumption, pollution, and carbon dioxide emissions, replacement of coal burning power plants with natural gas and diesel powered transit vehicles with natural gas powered vehicles seems to be an approach with immediate payoff in that it does not require any new technical advances.
SLC,
That is correct--CNG (compressed natural gas) is a proven technology and good from all of the environmental, operating cost, and domestic production standpoints. Plus they have also recenty developed CNG-hybrids, which may be a very competitive solution for quite a while.
I see nothing in that document about a "right size program," or indeed any discussion at all about changing transit vehicle sizes. Perhaps you could direct me to that discussion if you think there is one.
It was referred to throughout the peer review (a nice overview of transit systems in medium sized cities, btw), but I suppose I should have been more specific. In 2006 the Port Authority eliminated some routes, cut back on others, increased some in the core, and adjusted the fare system to increase the efficiency of the system as a whole to bring it in line with similar systems that had made similar right sizing moves in recent years. The process continues and next year they will introduce a new framework of using the buses to feed the trolley to cut down on the time they idle (and waste fuel and time) in downtown traffic. You wanted evidence of an evolving system, and there it lies.
jack lecou | June 27, 2008 2:57 AM
Huh? The whole point is that they likely would not be operating at peak-like load levels. Increases in demand do not generally come in convenient bus-sized increments. They come in arbitrary-sized increments. If you have to add extra buses because your existing capacity is slightly too small to accommodate the increase in demand, your average load factor is most likely going to go down.
My reply is below.
Let's hope it settles this silly aspect of the argument once and for all.
Obviously it won't ... the point of the silly aspects of the arguments here is to post a plausible-at-first-glance objection and then to make the discussion thread so long that someone taking a cursory glance gets the idea that there is some substance to the objection.
Those willing to work through these kinds of questions in detail are a small minority and most would already firm supporters of at least leveling the transport subsidy playing field and giving public transport equal treatment with car and air transport.
And of course it was never more than plausible sounding. When running a fleet of peak-load sized buses, they are either near capacity during peaks, or well below. If they are near capacity during peaks, then they have better fuel efficiency than the average commuter. Since adding more buses allows improving frequency during the peak, it increases the QOS and increases patronage, on top of the baseline increase in demand, and adding patronage when the buses are operating at better fuel-efficiency than cars will obviously be a fuel savings winner.
And when buses are so far below capacity that their average fuel efficiency is less than that of cars, then patronage can increase without increasing the number of buses, so the marginal fuel cost of adding patronage is less than the marginal fuel savings of not driving, so shifting passengers from cars to buses is a fuel savings winner in the balance of the service day as well.
And there are substantial opportunities for improved energy efficiency for the typical city bus, since high efficiency diesels with any form of stop/start energy recovery will do much better on a per-seat basis than the current norm, established during the days of ultra-cheap oil when buses were to provide the equity-based transport alternative for those who could not use the dominant car-based public/private transport system.
Unlike cars, a program to increase the energy efficiency of buses can be pursued that quickly improves the fuel efficiency of the entire fleet. A combination of a steeply discounted interest rate for the replacement of existing buses with fuel efficient buses, and purchase of fuel inefficient buses at a "fair trade in value" to scrap them and take them out of the fuel consumption picture, would be a very effective program for cutting operating costs of public transport operators, as well as providing a boost to the economy over the next two years, which going on the last two economic downturns will be about the length of the unemployment-recession.
It's hard to make any clear sense of this. If by "average load" you mean "average load factor," the claim is simply false. If you mean "average number of passengers per bus," it's still false. The effect of more riders on both average load factor and average vehicle occupancy depends on all sorts of variables. You simply can't predict the effect without more information.
I mean average load factor, though obviously average number of passengers per bus is equivalent mathematically, assuming our vehicles all have the same capacity.
As for the effect of ridership on average load factor on a route, see the post with all the numbers above.
Yet another bizarre claim. If you're pulling buses off the local stops you're obviously cutting service to the people who use those stops. And the effect on "load curves" would depend on the precise relationship between demand and the changes in capacity provided by your new/changed routes.
Whether this pays off or not depends on the relative popularity of the different stops, and the overall ridership, certainly.
The whole point is that higher ridership may make something like this a good idea on a line where it wasn't before. (And you'd be waiting for the point where it could improve overall quality of service: Note that not only do express users get a benefit, local-bus users may get some on-time-percentage and trip-time benefits from reduced traffic at stops and more even loads.)
As for the effect on load curves, we're replacing this traffic pattern:
A-b-c-D-e-F-g-h-I-j-K-L
with this one for locals:
a-b-c-d-e-f-g-h-i-j-k-l
and this one for expresses:
A-D-F-I-K-L
Obviously the devil's in the details, but overall both should now tend to have less spread between maximum load factor and minimum load factor, so they can target a higher average load factor.
But you've also quadrupled the number of buses and bus drivers you need. And massively increased all the other costs associated with buying, operating, maintaining and storing your bus fleet, since you now have four times as many buses.
No, we haven't quadrupled the number of buses.
We've quadrupled the number of passengers, and also the number of runs and revenue hours. All of which is a good thing. The exact number of additional drivers and buses required will depend on the length of the route, frequency of service and so forth.
BOTTOM LINE: Ridership improves load factors, both directly, and potentially in a lot of other subtle ways.
BruceMcF:
Le sigh. I know. But I can hope.
jack lecou,
Let's hope it settles this silly aspect of the argument once and for all. ....
How many more times do I have to go over this before you will pay attention? You cannot make assumptions about the need for increased capacity in response to an increase in demand from average peak-time load factor. The limiting factor is peak peak time load factor. The limiting factor is the number of people seeking to ride the bus at the busiest point of its route at the busiest time of day, not the average number of people at the busiest time of day.
You can keep repeating your silly calculations as many times as you like, but as long as you keep ignoring the need to provide sufficient capacity to satisfy peak demand, they are utterly worthless.
piotr,
One. Off-peak buses would be vans, 10-12 passanger vehicles, so they would cost considerably less than buses with 30 sitting and 70 standing places. We are increasing the capital costs perhaps by 25%.
Brilliant. You've obviously been channelling DTM. So how are you going to accommodate spikes in off-peak demand greater than 10-12 passengers? For that matter, how are you going to accommodate days or routes where even the average off-peak demand is greater than 10-12 passengers? You'll have to run two or more of your vans for each regular bus you're replacing. That means twice as many bus drivers to pay, higher maintenace and repair costs, greater congestion and other costs. If the route previously required 20 buses, you'll now need 40 at least. If the labor cost for a driver is $100,000 a year, that's an additional $2,000,000 a year just in additional labor costs for drivers. For each route in your system. Are you sure you'll be saving $2 million worth of fuel per route per year?
Another obvious problem with your absurd proposal is mobility. Have you ever actually been on a transit bus? They are generally large vehicles with easy entry and exit through wide automatic doors, a central aisle to facilitate access to seating, and a special area at the front of the bus with priority easy-access seating for the elderly and disabled. Many have a special area for wheelchairs. Many have a special storage area for bikes. Many have special storage areas for luggage or other bulky cargo. They're designed this way to accommodate special-needs passengers, as well as passengers carrying bags or other items, to facilitate access. I can't wait to see granny with her walker, or the single mom with two kids and bags of shopping in tow, try to get in and out of your 12-passenger minivan.
robotic,
It was referred to throughout the peer review (a nice overview of transit systems in medium sized cities, btw), but I suppose I should have been more specific. In 2006 the Port Authority eliminated some routes, cut back on others, increased some in the core, and adjusted the fare system to increase the efficiency of the system as a whole to bring it in line with similar systems that had made similar right sizing moves in recent years.
No, there is no reference to it whatsoever. If you still claim otherwise, cite the page number and quote the text where the document discusses bus sizing and proposes reducing bus sizes to increase load factors. Do you even read your own sources?
DTM,
Given that tradeoff, obviously as fuel costs rise, those sunk costs won't be increasing, but the benefits will.
Yes, that's right. All these new smaller buses you're proposing to use in place of the present large ones are just going to magically appear out of thin air. The transit authorities won't actually to have to pay for them. And the costs arising from idling the existing fleets of large buses for all but a few hours each weekday aren't real costs either.
And I'm still waiting for details of this alleged "right size program" in your city, DTM.
No, there is no reference to it whatsoever.
Yes there is, but it isn't explicitly called "right sizing". My apologies. An earlier version of this document discussed it more clearly. However, in the introduction on page 1-3 it references the right-sizing:
"Also relevant to the quantitative analysis are the service changes implemented by Port
Authority since the 2006 NTD was published; in particular a system-wide service cut that
reduced the total service hours by 15%. This service change was implemented in June
2007 and recent measures show that the 15% reduction in service resulted in a 3%
reduction in ridership. Although the results of this service change have not been fully
documented and recorded in the NTD, where appropriate, the peer analysis includes a
separate measure that includes a 15% reduction in service hours and 3% reduction in
ridership. In other areas, we discuss the likely impact of the service change on the
parameter and interpreted findings."
I wrote the preceding to give you an idea of how the service reductions and route changes have played out.
Posted by jack lecou | June 27, 2008>
Le sigh. I know. But I can hope.
I see no grounds at all for hope.
In arguing against the workability of an approach that implies a fleet of large buses sitting idle during much of the off-peak, he raises the "issue" of:
So how are you going to accommodate spikes in off-peak demand greater than 10-12 passengers? For that matter, how are you going to accommodate days or routes where even the average off-peak demand is greater than 10-12 passengers?
Obviously, if new patronage drives some routes out of the range of an off-peak maxi-taxi ... you just run a regular sized bus on that route. After all, the whole strategy implies that the spare capacity it sitting there, available to be used.
Or the objection that somehow the operator of the bus route loses the ability to track patronage at various parts of a bus route, saying:
You cannot make assumptions about the need for increased capacity in response to an increase in demand from average peak-time load factor. The limiting factor is peak peak time load factor. The limiting factor is the number of people seeking to ride the bus at the busiest point of its route at the busiest time of day, not the average number of people at the busiest time of day.... as if the route system is defined by some deity and the operator of the bus service just has to live with whatever the distribution of patronage on the bus route may happen to be.
So, no, there doesn't seem to be a basis for hoping, there.
Mixner,
You write: "All these new smaller buses you're proposing to use in place of the present large ones are just going to magically appear out of thin air. The transit authorities won't actually to have to pay for them."
Yes, clearly I failed to account for the costs of adding more buses. If only I had written something like:
"[O]bviously there is a tradeoff involved: adding more buses to serve the same number of passenger-miles will require some sort of additional sunk cost, but in turn the benefit will be savings on fuel during operations."
But since I totally didn't write anything like that, you nailed me. Outstanding piece of work on your part, very perceptive, you should be proud.
SLC,
Mr. Mixners' claim that busses are more polluting then cars is probably true for diesel powered buses. However, natural gas powered busses, such as have been purchased by WMATA are far less polluting then either. WMATA now operates some 400+ natural gas powered vehicles, approximately 30% of their fleet.
Most transit buses are diesel, and transit buses overall are more polluting than cars, and far more polluting than new-technology cars like the Toyota Prius. As transit authorities gradually replace their bus fleets with newer-technology vehicles, buses overall are likely to become less polluting than they are now. But of course, the same is true for cars as conventional vehicles are replaced with newer-technology ones. The achilles heal for transit buses is and always will be load factors. Under any real-world comparison of comparable engine/fuel technologies between cars and buses, cars will always come out ahead because transit buses will always have to run on average with so many empty seats.
robotic,
Yes there is, but it isn't explicitly called "right sizing".
No, there isn't. The text you quote contains no mention of bus size whatsoever, let alone a proposal to reduce bus size in order to increase load factor.
Posted by jack lecou | June 27, 2008 10:36 AM
Le sigh. I know. But I can hope.
Note in the following that "any real world" can be taken to mean:
* buses are restricted to a small mode split, and
* there are no anchoring patronage drivers and
* population is low density on average and
* there is no clustering infill development
... since in either current or historical instances where enough of the above do not hold, the "always and everywhere" claim does not hold.
The achilles heal for transit buses is and always will be load factors. Under any real-world comparison of comparable engine/fuel technologies between cars and buses, cars will always come out ahead because transit buses will always have to run on average with so many empty seats.
And, yeah, sure, that means that the "always" in the claim is what you would call a "lie", but ce'st la vie.
BruceMcF,
Most of what you write is so ridiculous I put it in the "too stupid to bother with" category. This isn't really an exception, but I'll bite anyway:
Obviously, if new patronage drives some routes out of the range of an off-peak maxi-taxi ... you just run a regular sized bus on that route. After all, the whole strategy implies that the spare capacity it sitting there, available to be used.
The problem isn't "new patronage," it's variation in demand among the existing "patronage." If you seriously believe that there are bus routes currently served by regular-sized buses with a maximum off-peak demand that rarely or never exceeds 10-12 passengers per bus, then identify them. Since there are literally thousands of transit bus routes across the country, you should have no trouble coming up with lots of examples, if you believe that replacing regular buses with 12-passenger vans is a serious proposal for increasing average load factors.
And once you've done that, once you've shown that there is even a single bus route in the country that satisfies the necessary demand criteria, then explain how you have determined that the savings in fuel costs would exceed the additional capital costs and other costs arising from the replacement. And once you've done that, show me how you have determined that the mobility issues I described, and all other practical considerations that transit planners must take into account in determining the size of the buses its runs, would not kill your proposal anyway.
DTM,
Yes, clearly I failed to account for the costs of adding more buses. If only I had written something like: "[O]bviously there is a tradeoff involved: adding more buses to serve the same number of passenger-miles will require some sort of additional sunk cost, but in turn the benefit will be savings on fuel during operations."
But you also wrote, "obviously as fuel costs rise, those sunk costs won't be increasing, but the benefits will." You're contradicting yourself. If the number of routes for which replacing large buses with smaller ones is economically beneficial rises as fuel costs rise, then obviously the "sunk costs" will be increasing, because you'll have to keep buying more and more smaller buses to replace larger ones.
But then, since neither you nor anyone else has presented a remotely serious case that there is even a single transit bus route in the entire country where a "right size program" of this kind would be economically beneficial at all, let alone that there are enough such routes for such a program to have a meaningful effect on the overall fuel-efficiency of the nation's transit bus system, this is all academic anyway.
Mixner,
You wrote above:
"Load factors in transit buses and trains will always be low because of the nature of the service they provide. The need to provide frequent service and to accommodate large variations in demand across different times of the day and different segments of their routes means that most seats will be empty most of the time."
Please demonstrate the truth of all these assertions, using relevant and comprehensive route data and studies.
Or is this another Mixner-says-whatever-he-wants-but-other-people-need-to-provide-treatises situation?
DTM,
Please demonstrate the truth of all these assertions,
Sorry, you don't get to demand evidence from others until you're willing to provide evidence to substantiate your own ridiculous claims.
Still waiting for details of this alleged "right size program" in your city, DTM.
Mixner,
First, I'm still waiting for your explanation of the trend in the BTS charts.
Second, of course the sunk costs are multiplied per case of implementation, as are the benefits. My point was that the benefits per case of implementation will increase with fuel price increases, but the sunk costs per case of implementation will not.
Now frankly, I don't think even you are stupid enough not to understand the point I was making. So, I think you are just arguing in bad faith. But I don't get why--who do you think is going to reward your bad faith? You know better, and I know better, and I doubt anyone else is paying the slightest bit of attention at this point.
How many more times do I have to go over this before you will pay attention? You cannot make assumptions about the need for increased capacity in response to an increase in demand from average peak-time load factor. The limiting factor is peak peak time load factor. The limiting factor is the number of people seeking to ride the bus at the busiest point of its route at the busiest time of day, not the average number of people at the busiest time of day.
How many more times? Well, I hope you'll stop repeating the silly canard. Maybe you’ll get it this time:
This is an utterly spurious objection. I have explained this at least once before, and I shall do so again, slowly, and with pictures. (The snark is free.)
Mathematically, it does not matter what the underlying distribution of load factors on segments is. That distribution does not change under a uniform increase in ridership. If we know we typically do not exceed capacity when running at an X% average load factor, and actual average load factor is less than X%, we're not exceeding capacity. That's all we need to know.
For example, suppose we have a route in the morning running between an outer ring and the downtown core. We might expect the load factor on a typical bus to rise gradually, perhaps more quickly as density increases, then peak as it crosses the inner ring, and taper off rapidly as commuters disembark. Graphically, maybe something like this:
######################
* *
* *
* * *
* * *
* * * *
* * * *
* * * * * *
* * * * * *
Where each column of ‘*’s is the load factor of a typical bus at that point along the route, and the line of ‘#’s is the maximum capacity of the bus. We keep our typical maximum load factor below that, because there is some variance we need to allow for.
If we call each ‘*’ 10%, then obviously our average load factor is 50%.
Now let’s increase ridership by 50%:
* *
* *
##########*##*########
* * *
* * *
* * *
* * * *
* * * *
* * * *
* * * * * *
* * * * * *
* * * * * *
Oh noes! Even our typical bus is over the limit!
I wonder what the average variance is now though... Oh, weird, 75%. That’s exactly 50% higher than the previous average load factor. Almost like the 50% increase is linear...
What an amazing coincidence. When the underlying distribution doesn’t change, targeting a 50% average load factor is a mathematically perfect stand-in for not having a peak which exceeds maximum capacity!
And what happens to load factors on our typical bus if we add 50% more buses picking up passengers before we get there?
######################
* *
* *
* * *
* * *
* * * *
* * * *
* * * * * *
* * * * * *
Oh. That looks familiar...
Re Mixner
In terms of the carbon footprint, their is no possibility of a gasoline powered vehicle of any type having a smaller carbon output then a natural gas powered vehicle of the same type. Natural gas just has a smaller percentage of carbon in its molecule then does gasoline. No technology is going to overcome this problem. The only way to lower the carbon output is to change the fuel.
By the way, Mr. Mixner hasn't told us how the current highway infrastructure is going to handle the extra vehicles. In Virginia, Mr. Mixners' right wing Rethuglican pals in the State Legislature adamantly refuse to raise the gas tax to fund new infrastructure construction.
DTM,
First, I'm still waiting for your explanation of the trend in the BTS charts.
Then you're going to be waiting for a long time. Since I haven't claimed to be able to "explain" that trend, I'm not sure why you're still waiting. You did claim that there is "right size program" in your "city." I'm still waiting for you to provide shred of evidence to support that assertion, or even that any such program exists anywhere in the country.
But then, since it's obvious you're just making the whole thing up, your continued evasion isn't terribly surprising. As long as you continue to present your wishful thinking and guesses as real-world facts, you're going to continue to be called on it.
Second, of course the sunk costs are multiplied per case of implementation, as are the benefits.
Yes, so your claim that "those sunk costs won't be increasing, but the benefits will" is obviously false, isn't it? Your posts are full of this kind of contradiction.
jack lecou,
I don't know if it's the amount of time you devote to constructing your meaningless made-up scenarios and tables and graphs that's preventing you from understanding the conceptual problem with your argument, or something else, but you still seem completely confused. You write:
Mathematically, it does not matter what the underlying distribution of load factors on segments is. That distribution does not change under a uniform increase in ridership. If we know we typically do not exceed capacity when running at an X% average load factor, and actual average load factor is less than X%, we're not exceeding capacity. That's all we need to know.
This is just completely wrong. First, you obviously cannot assume that the distribution of demand would be preserved. If the new riders were primarily commuters, for example, then the additional demand would be concentrated in peak-time services. But even if it were uniform, your claim that "all we need to know" is that current demand does not exceed current capacity in order to predict the effect of satisfying the new demand on load factor is just utter nonsense. The effect on load factor will depend on the ratio of new capacity to new demand, and you simply don't know what that would be without more information. That's the problem with your latest "scenario," too. You have simply assumed that the increase in demand and supply would be the same, and that there would be no change in demand distribution. You can't make those assumptions. You don't know what the relationship between new demand and new capacity would be in any real-world situation. Since increases in demand do not come in convenient bus-sized increments the effect on load factor of meeting new demand is not predictable without additional information.
In terms of the carbon footprint, their is no possibility of a gasoline powered vehicle of any type having a smaller carbon output then a natural gas powered vehicle of the same type. Natural gas just has a smaller percentage of carbon in its molecule then does gasoline.
Wrong. The carbon output obviously depends on the size of the engine, among other things. A hybrid bus with a small gasoline engine might well produce less carbon that a CNG bus of the same size and power.
But this is irrelevant to the point, anyway. I didn't claim that conventional cars are less polluting that CNG buses. I said that cars in general are less polluting than buses in general. New technology, including CNG, can be used in both buses and cars. Cars will always come out ahead of transit buses in comparisons of vehicles using the same technology, because transit buses so often have to carry around so much dead weight in the form of empty seats, due to their need to provide frequent service and accomodate dramatic variations in demand.
First, you obviously cannot assume that the distribution of demand would be preserved. If the new riders were primarily commuters, for example, then the additional demand would be concentrated in peak-time services.
Good lord. The stupid, it burns!
Lecou is talking about a spatial distribution, you insufferable dolt! Changes in temporal distribution are dealt with by changes in frequency and/or splitting service into express and local buses that travel the same route. Now unless patterns of land use change on an hourly basis, I suggest you reconsider just what the hell you're doing here.
You're certainly not enhancing your reputation.
Again, just so you know, our universe consists of dimension of both space and time.
This is just completely wrong. First, you obviously cannot assume that the distribution of demand would be preserved.
It's a good first order approximation. More people in general start taking the bus. They tend to get on at the station nearest their house. The population distribution hasn't changed. So...
(In any case, it's not really a problem. As I've mentioned previously: If the increase is predominantly with more riders getting on or off at previously less used stations, this smooths the spikes in load factor. On the other hand, if already busy stations get the extra business, it increases the chances that we'll be able to effectively rearrange service with an express route or something. Win either way.)
The pathological case would be a large (as percent of overall ridership) increase in riders all at one stop, making our load factor graph much spikier. That could be a problem, load factor wise, but it's hardly a realistic scenario.
If the new riders were primarily commuters, for example, then the additional demand would be concentrated in peak-time services.
This is, in fact, exactly the scenario considered in my last post. It deals only with a peak (morning) time commuter service. In fact, the idea that there is a large increase in ridership on peak-time buses is sort of integral to the whole concept. Or didn't you notice?
(I'm getting the feeling you haven't really thought it through... Big surprise.)
But even if it were uniform, your claim that "all we need to know" is that current demand does not exceed current capacity in order to predict the effect of satisfying the new demand on load factor is just utter nonsense.
This is not what I said at all.
I said that if the load factor profile on a route at a particular time of day is such that it's peak is, say, 120% (safely below an absolute maximum capacity of say, 140%), and average load factor along the length of the route is, say 30%, then adding buses and passengers in roughly equal measure (with new passengers frequenting stations in the same proportions as old passengers) will keep the average load factor at 30%. Since the underlying distribution of stop usage is approximately the same, we can also conclude that peak load remains at a safe 120%.
The effect on load factor will depend on the ratio of new capacity to new demand, and you simply don't know what that would be without more information.
Since new exogenous demand is precisely what is driving our addition of new capacity, this is laughably false.
(There are second order feedbacks of course - more buses improves service, increases demand, we add more buses, but that's straightforward enough.)
That's the problem with your latest "scenario," too. You have simply assumed that the increase in demand and supply would be the same.
...I don't know whether to laugh or cry...
and that there would be no change in demand distribution. You can't make those assumptions. You don't know what the relationship between new demand and new capacity would be in any real-world situation.
Real word situation: I am operating the transit system for a large city. There is a panic about carcinogens in automobile steering wheels or something. 10% more commuters start taking the bus. I see that buses are becoming more crowded, I add buses until peak load factors are acceptable again. 10%. Magic.
Since increases in demand do not come in convenient bus-sized increments the effect on load factor of meeting new demand is not predictable without additional information.
If I have 47 (peak-time) buses, a 10% increase in demand is 4.7 buses worth of passengers. Not a very convenient increment, I guess.
Buses only come in integers, so I add 5 more buses. The new buses have have an average load factor about 94% that of the other peak buses. Adding 5 near-peak buses increases overall average load factor in the system.
If I have 523 peak buses and demand goes up 7%, I add 37 buses with an average capacity about 98% peak. Overall load factor goes up.
There's a whole table full of this stuff up thread. Check it out.
DMonteith,
Your refusal to respond to anything I write without endless personal insults really is tedious, but if that's the way you want to play it....
Good lord. The stupid, it burns!
You stupid, idiotic, moronic imbecile!
Lecou is talking about a spatial distribution, you insufferable dolt!
He did not identify the nature of the distribution he was referring to, and the assumption is not justified whether the type of distribution in question is spatial or temporal, you monumental cretin!
Changes in temporal distribution are dealt with by changes in frequency and/or splitting service into express and local buses that travel the same route.
This statement is a complete nonsequitur. The issue is not how changes in temporal distribution "are dealt with." You obviously don't know how they "are dealt with." The issue is the effect of increased ridership on average load factor, and that effect depends on, among other things, the temporal distribution of the new demand.
You're certainly not enhancing your reputation.
You're certainly not doing anything to dispel the notion that you are an ignorant fool.
jack lecou,
It's a good first order approximation. More people in general start taking the bus. They tend to get on at the station nearest their house. The population distribution hasn't changed. So...
No, it's an unjustified assumption. You simply don't know what the distribution of the new demand would be. It could be mainly workers. It could be mainly students. It could be mainly shoppers. It could be mainly retirees. You just don't know. It depends on the particular case. The distribution of new demand across different destinations and different times of the day is likely to be very different for each of those groups. And the nature of that distribution, among other things, would influence the effect of the new riders on average load factor.
This is not what I said at all. I said that if the load factor profile on a route at a particular time of day is such that it's peak is, say, 120% (safely below an absolute maximum capacity of say, 140%), and average load factor along the length of the route is, say 30%, then adding buses and passengers in roughly equal measure (with new passengers frequenting stations in the same proportions as old passengers) will keep the average load factor at 30%.
Yes, I understand you meant that. And I'm pointing out in reply that you cannot assume, from knowledge of current peak demand and current average load factor, that the new demand would be satisfied by "adding buses and passengers in roughly equal measure." Without more information, you have no idea what the ratio of additional capacity to additional riders would be.
If I have 47 (peak-time) buses, a 10% increase in demand is 4.7 buses worth of passengers. Not a very convenient increment, I guess. Buses only come in integers, so I add 5 more buses. The new buses have have an average load factor about 94% that of the other peak buses. Adding 5 near-peak buses increases overall average load factor in the system.
You really ought to stop pretending that you can demonstrate your assertion with these ridiculous contrived scenarios, especially since you are so bad at formulating them in the first place. First, as you have stated your scenario above, "4.7 buses worth" would be a 10% increase in capacity, not demand. We don't know how large an increase in demand it would be because you haven't stated the current demand. Therefore, we have no basis for calculating the change in load factor.
Since you like made-up scenarios so much, I'll make it real simple for you. We can dispense with considerations of multiple buses, stops and times of day and pare it down to the simplest case of a bus service that runs a single bus once a day between two points. Initially, the demand is 90% of the capacity of the bus. The average load factor is 0.9. Demand increases by 20% of the bus's capacity, so demand is now 110% of capacity. An additional daily bus is added to the service to handle the extra demand. Voila! Ridership has increased, but average load factor has dropped from 0.9 to 0.55. This isn't really hard to understand. I'm not sure why you're having so much trouble.
Claim:
He did not identify the nature of the distribution he was referring to...
Falsification:
...each column of ‘*’s is the load factor of a typical bus at that point along the route...
We might expect the load factor on a typical bus to rise gradually, perhaps more quickly as density increases, then peak as it crosses the inner ring, and taper off rapidly as commuters disembark...
For example, suppose we have stops...A-b-c-D-e-F-g-h-I-j-K-L...Where capital letters denote stops with significantly higher traffic than lower case stops.
Claim:
...and the assumption is not justified whether the type of distribution in question is spatial or temporal...
Justification:
*crickets chirping*
Claim:
The issue is the effect of increased ridership on average load factor, and that effect depends on, among other things, the temporal distribution of the new demand.
Refutation:
...targeting a 50% average load factor is a mathematically perfect stand-in for not having a peak which exceeds maximum capacity!
Note:
Earlier you claimed that "The limiting factor is the number of people seeking to ride the bus at the busiest point of its route at the busiest time of day, not the average number of people at the busiest time of day".
Question:
Is the issue average load factor or peak peak capacity? You seem to be arguing with yourself here.
And, finally:
Your refusal to respond to anything I write without endless personal insults really is tedious, but if that's the way you want to play it....You stupid, idiotic, moronic imbecile!
Progress! You acknowledged the validity of my position and responded forthrightly within the rhetorical framework and logical premises of the pertinent subset of our discussion! Now if we can just bring this spirit of cooperation from ad hominem snark to the merits of the arguments at hand we might get somewhere!
Not holding my breath though.
DMonteith,
Falsification: ...each column of ‘*’s is the load factor of a typical bus at that point along the route...
Falsification of your "falsification": You're quoting from a different part of the post. I was responding to the text I quoted immediately above my response, which contains no mention of the nature of the distribution but merely asserts that the "distribution does not change under a uniform increase in ridership."
Justification: *crickets chirping*
I have no burden of justification. If you wish to claim that there is a relevant difference between spatial and temporal distribution with respect to the issue in question, the burden is on you to identify that difference and explain why it's relevant. Good luck.
Refutation: ...targeting a 50% average load factor is a mathematically perfect stand-in for not having a peak which exceeds maximum capacity!
It's hard to know why you think this statement "refutes" the statement of mine you claim to be refuting. Do you have, you know, an actual argument to offer?
Is the issue average load factor or peak peak capacity?
A number of issues have been raised and are being discussed. The issue that the statement of mine you quote here addresses is the limit on accommodation of greater peak-time demand without increased capacity.
You acknowledged the validity of my position
I think you misread the statement you quoted. It describes you as a "stupid, idiotic, moronic imbecile."
Initially, the demand is 90% of the capacity of the bus. The average load factor is 0.9. Demand increases by 20% of the bus's capacity, so demand is now 110% of capacity. An additional daily bus is added to the service to handle the extra demand. Voila! Ridership has increased, but average load factor has dropped from 0.9 to 0.55. This isn't really hard to understand. I'm not sure why you're having so much trouble.
This is the fundamental cause of your confusion, I think.
Your example proves that ridership can reduce load factor, but only in a bus service that operates exactly one bus, which therefore cannot have peak and non-peak service.
Move up to even just 3 or 4 bus trips total, a couple peak, a couple non-peak, and your argument evaporates. It just gets worse from there.
(And note that even in your pathological case, load factor only really decreases a vanishingly small fraction of the time: Pick a random load factor for your first bus, between 1 and 100. Now pick a random increase, between, say, 1 and 1000. See the problem?)
Of course, real bus services do not operate just one bus. During peak times, a real bus line will have half a dozen or more buses passing each stop every hour. And new passengers are not tacked onto the back of the emptiest bus, they dribble out at stops (at different rates throughout the day) and are picked up by the first bus that comes past. There are at least two more-or-less predictable peaks during the day when buses are load-rate limited, and the rest of the day buses operate below capacity.
If we generalize a little, we get, well, any of my examples above. And we find that increases in ridership generally improve load factors.
jack lecou,
Your example proves that ridership can reduce load factor, but only in a bus service that operates exactly one bus,
I assume you mean increased ridership. My "example" certainly does prove that increased ridership can reduce load factor. But it obviously does not prove that this can be done "only in a bus service that operates exactly one bus." The latter claim is just obviously false, as a simple tweak of the example makes clear: Initially, the service runs two buses, each with an average load factor of 0.9. Meaning the average load factor for the service is also 0.9. Demand on the route then increases by 0.3 of the capacity of a single bus. Demand is now 105% of capacity, so a third bus is added. The average load factor for the service falls from 0.9 to 0.7.
Of course, real bus services do not operate just one bus.
Yes, obviously. The point of my "example" was to show that it's possible to contrive situations in which a change in ridership yields a higher load factor, a lower load factor, or no change to the load factor. Which is why your endless made-up tables and graphs and "examples" are such an utter waste of time.
The fundamental point you still seem incapable of grasping is that the effect of an increase in ridership on load factor depends on its effect on the ratio of riders to capacity, and you simply can't predict that without additional information about the specific circumstances of the case.
Mixner,
The problem with your constantly arguing in bad faith is it hardly encourages people to comply with your requests for information.
For example, ordinarily I would happily identify my locale and one of the relevant routes that have been switched to recently-acquired smaller buses (that turns out not to be too difficult, because one of those routes literally goes right past my house, so I now see the new smaller buses with my very own eyes on a regular basis, with the route number in handy bright lettering).
The problem is that you are asking the question in bad faith. Specifically, we both know that if I actually supplied this information, you would then start arguing that anecdotal evidence isn't data, that this is not a typical case, yada-yada-yada.
So, this is yet another typical Mixner exchange: you make all sorts of assertions about ridership on bus routes without any supporting data, but if anyone dares contradict your assertions, they must provide a detailed treatise to your specifications ... and if they do, you will ignore their response anyway because you weren't making the request in good faith.
So, no, I'm not going to do that. And that is because you have given me every reason not to.
DTM, you are such a ridiculous poser. Did you seriously think you would get away with this "right size program" nonsense? It's not just that you make stuff up, it's that you're so transparently lying. Keep going, and I'll keep calling you on it.
My "example" certainly does prove that increased ridership can reduce load factor. But it obviously does not prove that this can be done "only in a bus service that operates exactly one bus." The latter claim is just obviously false, as a simple tweak of the example makes clear: Initially, the service runs two buses, each with an average load factor of 0.9. Meaning the average load factor for the service is also 0.9. Demand on the route then increases by 0.3 of the capacity of a single bus. Demand is now 105% of capacity, so a third bus is added. The average load factor for the service falls from 0.9 to 0.7.
I should not have said exactly one bus. It's not literally one bus, but it is very low numbers of buses. (And for the record, I never said that increases in ridership could NEVER decrease load factors, only that they increase them in the vast majority of circumstances.)
To illustrate:
(The number of initial buses runs along the top, percent increase in ridership over maximum capacity is down the side. The numbers in the grid are the corresponding system load factors after the increase. For example, your scenario can be found in the '2' column, on the '15%' line.)
2 3 4 5 6 7 8 9 10
0% 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0 90.0
5% 95.0 95.0 95.0 95.0 95.0 95.0 95.0 95.0 95.0
10% 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0
15% 70.0 78.75 84.0 87.5 90.0 91.88 93.33 94.5 95.45
20% 73.33 82.5 88.0 91.67 94.29 96.25 97.78 99.0 100.0
25% 76.67 86.25 92.0 95.83 98.57 89.44 92.0 94.09 95.83
30% 80.0 90.0 96.0 100.0 90.0 93.33 96.0 98.18 100.0
35% 83.33 93.75 100.0 89.29 93.75 97.22 100.0 93.75 96.15
40% 86.67 97.5 86.67 92.86 97.5 91.0 94.55 97.5 100.0
45% 90.0 81.0 90.0 96.43 90.0 94.5 98.18 93.46 96.43
50% 93.33 84.0 93.33 100.0 93.33 98.0 93.33 96.92 100.0
55% 96.67 87.0 96.67 90.63 96.67 92.27 96.67 93.21 96.67
60% 100.0 90.0 100.0 93.75 100.0 95.45 100.0 96.43 100.0
65% 77.5 93.0 88.57 96.88 93.0 98.64 95.38 99.64 96.88
70% 80.0 96.0 91.43 100.0 96.0 93.33 98.46 96.0 100.0
75% 82.5 99.0 94.29 91.67 99.0 96.25 94.29 99.0 97.06
80% 85.0 85.0 97.14 94.44 92.73 99.17 97.14 95.63 100.0
85% 87.5 87.5 100.0 97.22 95.45 94.23 100.0 98.44 97.22
90% 90.0 90.0 90.0 100.0 98.18 96.92 96.0 95.29 100.0
95% 92.5 92.5 92.5 92.5 92.5 99.62 98.67 97.94 97.37
100% 95.0 95.0 95.0 95.0 95.0 95.0 95.0 95.0 100.0
Notice how, except for the "2 buses" column, the majority of the new load factors are greater then 90. By time we reach the 6 bus column, they ALL are.
Your examples only work because with very small numbers of buses (less than 5), the increases in capacity are "lumpy", and particular combinations may result in a decrease in load factor.
Even with 3 buses most increases increase load factor. With 3 more, the effect disappears entirely. In fact, if we were to extend the chart even further out, we would see that with tens or hundreds of buses, the tendency is for any ridership increase above 10% to push load factors to 99.999% immediately.
This should not be surprising.
(NOTE: From your example, I gathered that your 90% is not the "maximum practical load factor", in your system, but rather 100% is. Although not realistic, that's fine for simplicity. And note that we could easily just treat the numbers like 90% above as "% of maximum practical average operational load factor" rather than "% of capacity", and all the math works out the same.)
Another note: I did the increases down the left in increments of 5% so that I could include Mixner's case and still get all the way to 100% without filling too many pages.
However, it occurs to me that this had the unfortunate effect of eliding the absolute worst case increase for a 90% bus system: 10.1% [well, theoretically, 10%+delta]. I wasn't really trying to hide anything, and it doesn't change the bottom line.
For the record: In this scenario, the 7 and 8 columns do dip slightly below 90 with a 10.1% increase, but with 9 or more buses, ridership always increases load factor.
jack lecou,
Yet another utterly irrelevant table. Put the spreadsheet away for five minutes and try thinking. Since your table provides no results for scenarios other than those with an initial load factor of 0.9 it obviously doesn't tell us anything about how the load factor would change for those other scenarios. Nor does it tell us anything about changes in load factor for scenarios in which the additional buses have a different capacity than the original bus. Or scenarios in which the additional demand is accommodated by replacing the original bus with a bigger bus, or two or more smaller ones. So even for this one contrived type of service (point-to-point, no stops, no daily variation in demand), your table tells us nothing meaningful about how load factors would change under any scenario of change in demand and capacity except the particular one I described.
I don't know how many times I have to tell you this: you can't support your claim that increased ridership would usually produce increased load factor with contrived examples. You can't support it at all because you simply don't know. The effect of a change in ridership on average load factor in any real-world situation simply cannot be predicted without additional information about how that additional ridership would be accommodated. You could spend the next week at your computer generating thousands of tables for thousands of combinations, and they still wouldn't tell us how load factors would change in real-world bus systems.
If you still don't understand this, never mind. If you want to keep on constructing your meaningless tables and charts, knock yourself out. I won't be looking at any more of them.
Since your table provides no results for scenarios other than those with an initial load factor of 0.9 it obviously doesn't tell us anything about how the load factor would change for those other scenarios.
I'll give you a hint: if the initial load factor is less than 50, all possible increases in ridership increase load factors. Even on a single bus.
As initial load increases from there, a small number of the scenarios with the fewest numbers of buses begin to show small dips.
The situation is very worst as we approach 100%. Like the 90% above, but 95 or 99 would be somewhat worse. We might have to get out to a couple dozen buses before ALL increases in ridership increase load factor.
Of course, at that point it's silly. Load factor is already pretty much maxed out, so if it dips down to 99.8%, big deal. It's just oscillating tightly up against 100%.
Nor does it tell us anything about changes in load factor for scenarios in which the additional buses have a different capacity than the original bus. Or scenarios in which the additional demand is accommodated by replacing the original bus with a bigger bus, or two or more smaller ones.
Explain to me exactly how any of those possibilites would HELP your case.
Your entire case was based on the idea that adding less than a busload of people to a very small system leaves a difficult to manage remainder. It falls apart once we have a few more buses to even out the lumpiness, but it falls apart even faster if we have a stock of more appropriately sized vehicles to add or substitute. (Besides, you just spent half the thread arguing that maintaining a multi-size vehicle fleet was economically impossible.)
The only possible way this helps you is if you've got just one small bus, and your ridership increases a little, and all you have to add is a very large bus.
You're down to the corner case of the corner case.
So even for this one contrived type of service (point-to-point, no stops, no daily variation in demand)
I'll just let you think about how the examples we just went through are not contrived at all, but are in fact perfectly good models for routes with arbitrary numbers of stops and spatial variation in demand.
As for daily variation, well if we have reliable peaks in temporal demand (morning/afternoon rush), then of course the situation gets even better. We can have peak and non-peak service. I've already done the charts on that.
your table tells us nothing meaningful about how load factors would change under any scenario of change in demand and capacity except the particular one I described.
And all other ones it models...
Look Jack,
He just doesn't get it. He's just now complained that you haven't provided graphs that explore the entire space of possibilities and in the next breath stated that even if you were to provide them he wouldn't look at them.
Given the logic he's following, he probably has major problems with the gas laws because they fail to describe the behavior any particular molecule at any specific point in time. Likewise, the averaging out of subatomic quantum fluctuations in systems larger than nanometer scale must just keep him awake at night. Also, a single day drop (or gain) in stock prices is for Mixner proof positive those claiming that a bull (or bear) market exists are fools. He points to the noise in data and says "look, there's no signal here!". A couple of weeks ago he told me that April's 10% drop in oil prices was evidence that my contention that oil prices would keep rising was dubious.
You've thoroughly discredited him, even if he doesn't know it.
jack lecou,
I'll give you a hint: if the initial load factor is less than 50, all possible increases in ridership increase load factors. Even on a single bus.
Nonsense. If the new bus was larger than the current one the load factor could obviously decrease even in scenarios where the initial load factor was less than 0.5.
The situation is very worst as we approach 100%. Like the 90% above, but 95 or 99 would be somewhat worse. We might have to get out to a couple dozen buses before ALL increases in ridership increase load factor.
Yes, the higher the initial load factor, the more likely an increase in demand would require additional capacity, and the more likely the load factor would fall. Since you don't know what the initial load factor would be in any real-world service of this type, and you don't know how much the service would need to grow to accommodate increased demand, you can't possibly predict how many buses would need to be added. A 12-fold increase in demand for a single service isn't terribly plausible. You seem to think that all possible initial passenger loads, from a single passenger to a single passenger less than full capacity, and all possible increases in demand, are equally plausible. But that's nonsense. You are so obsessed with blindly cranking out meaningless numbers that you seem incapable of understanding the conceptual problems with your argument.
Explain to me exactly how any of those possibilites would HELP your case.
Er, can you really not do the math? In my original scenario, the load factor fell from 0.9 to 0.55. If the additional bus had 50% higher capacity than the current bus instead of being the same size, it would fall even more, from 0.9 to 0.44.
Your entire case was based on the idea that adding less than a busload of people to a very small system leaves a difficult to manage remainder.
No, my entire case is built on the undeniable fact that, without knowing the details of the situation, you simply can't predict how much capacity would be added in any real-world scenario to accommodate an increase in demand. You therefore have no basis for predicting the change in average load factor.
I'll just let you think about how the examples we just went through are not contrived at all, but are in fact perfectly good models for routes with arbitrary numbers of stops and spatial variation in demand.
Nonsense. Every one of your "examples" is totally contrived. Every one of them makes assumptions about the magnitude and distribution of new demand and the magnitude and distribution of added capacity, that obviously may not be true in any real-world situation.
Nonsense. If the new bus was larger than the current one the load factor could obviously decrease even in scenarios where the initial load factor was less than 0.5.
Yeah, I mentioned that already. You run a bus line with a single small bus and then get some new passengers, but all you've got in the garage to add is a huge double decker, right?
And I'm the one with the non real world examples?
Yes, the higher the initial load factor, the more likely an increase in demand would require additional capacity, and the more likely the load factor would fall.
No, you're missing the point. That ONLY happens with very small numbers of buses.
And EVEN THEN, if you have say, 3 buses and start at 98%, you can, if you tweak the increase just right, make load factor lower...all the way down to the mid 90s. That's technically a decrease, but is this really the point you're trying to make?
Since you don't know what the initial load factor would be in any real-world service of this type, and you don't know how much the service would need to grow to accommodate increased demand, you can't possibly predict how many buses would need to be added.
But we do know what the initial load factors would be in a real world service: Very high (near practical maximum) on the morning and afternoon peaks, very low throughout the rest of the day. We model this with separate peak and non-peak services. The charts are all there upthread.
A 12-fold increase in demand for a single service isn't terribly plausible. You seem to think that all possible initial passenger loads, from a single passenger to a single passenger less than full capacity, and all possible increases in demand, are equally plausible. But that's nonsense.
Indeed. I've merely been trying to be thorough. Both realistic and non-realistic numbers have been covered, I did not say they were all equally plausible.
In realistic scenarios (as in almost all possible scenarios) increased ridership increases load factors.
Er, can you really not do the math? In my original scenario, the load factor fell from 0.9 to 0.55. If the additional bus had 50% higher capacity than the current bus instead of being the same size, it would fall even more, from 0.9 to 0.44.
Um. In your original example, demand increased to 110% of capacity of the original bus. If you have a bus around with 150% capacity, why not just replace the first bus. Load factor 73%...
Since that's equivalent to having 2 buses, and adding a third, it's already in the chart.
And yeah, that's still a decrease, but for the umpteenth time, it only happens with very small numbers of buses, where even a single new bus is a relatively big capacity increase. With even a modest number of runs on the route, this sort of lumpiness disappears.
No, my entire case is built on the undeniable fact that, without knowing the details of the situation, you simply can't predict how much capacity would be added in any real-world scenario to accommodate an increase in demand. You therefore have no basis for predicting the change in average load factor.
As I have been pointing out, your case is built on nothing but handwaving. The only scenarios in which increases in ridership decrease load factor are examples like yours--with very small numbers of buses--or a scenario in which a very large increase in ridership (as a percent of the entire line) ALL starts getting on our peak service at the same stop just before the peak load point. Depending on the numbers, it's POSSIBLE that situation would lower our average load factor a bit.
But neither your 2-bus scenarios nor that one are very realistic, and in any other set of circumstances, ridership increases allow load factor increases.
Nonsense. Every one of your "examples" is totally contrived. Every one of them makes assumptions about the magnitude and distribution of new demand and the magnitude and distribution of added capacity, that obviously may not be true in any real-world situation.
I have pointed out several times how the models are a great deal more general than you seem to understand, and the results are robust under almost any possible set of circumstances.
All you can come up with now are obscure corner cases like Bob's bus service (now with a second bus due to customer demand!), and this vague handwaving about how it's too complicated for you to understand the general case. (And, of course, goalpost moving, but we won't get into that because, hilariously, you still haven't actually moved them far enough.)
DMonteith-
Thanks. It's just so dissatisfying when someone is wrong on the internets...
Hook up a turbine to the silly little boy and you'd power a small city for a year. He's as close as you'll get to a perpetual motion hot air generator.
jack lecou,
Yeah, I mentioned that already.
No, you falsely claimed that "if the initial load factor is less than 50, all possible increases in ridership increase load factors." That's just wrong.
No, you're missing the point. That ONLY happens with very small numbers of buses.
No, it doesn't. It could happen with any number of buses, depending on the other variables in the particular situation.
But we do know what the initial load factors would be in a real world service: Very high (near practical maximum) on the morning and afternoon peaks, very low throughout the rest of the day. We model this with separate peak and non-peak services.
No, you did not "model" it. You simply described a single contrived scenario of a 50% increase in ridership that preserved the existing demand distribution and was accommodated by a 50% increase in capacity. And even under that contrived scenario, the load factor didn't increase. It remained constant.
Indeed. I've merely been trying to be thorough. Both realistic and non-realistic numbers have been covered, I did not say they were all equally plausible.
Your tables simply list different combinations, with no weighting for plausibility. You treat every theoretical combination of increase in demand and capacity as equally plausible. You assume the distribution of new demand will exactly match the distribution of existing demand. None of these assumptions is justified.
Um. In your original example, demand increased to 110% of capacity of the original bus. If you have a bus around with 150% capacity, why not just replace the first bus.
Um. Because you may be anticipating further increases in demand above 150% of original capacity. Because you want to increase the frequency of service. Because the first bus has special seats for disabled riders and the new one doesn't. Because the larger bus is also used on another route and is not available until later in the day. Any number of possible reasons. That's why, as I keep telling you, you cannot predict the change in capacity without addditional information.
And yeah, that's still a decrease, but for the umpteenth time, it only happens with very small numbers of buses,
And for the umpteenth time in response, any real-world bus service of this type may only ever involve a small number of buses. You can't assume every combination is equally likely to be present in the real world, remember?
As I have been pointing out, your case is built on nothing but handwaving. The only scenarios in which increases in ridership decrease load factor are examples like yours--with very small numbers of buses--or a scenario in which a very large increase in ridership (as a percent of the entire line) ALL starts getting on our peak service at the same stop just before the peak load point.
More nonsense. The effect of an increase in ridership on load factor is independent of the total volume of passengers served, because it depends on the ratio of new capacity to new volume, and you simply cannot predict how that ratio would change without more information.
The simplest refutation of your mindless "let's list every theoretical combination and count the ones that increase load factor" argument is the fact that it is contradicted by real-world transit systems. I don't know if the government maintains statistics on bus transit passenger volumes and load factors, but it does for Amtrak. And the Amtrak figures completely contradict your stupid assertion. Total ridership has increased by about 10% over the past 5 years, but average load factor has not increased at all. Year-to-year load factors have both increased and decreased across increasing passenger volume.
You can't assume every combination is equally likely to be present in the real world, remember?
Another Classic Mixnerism. Remember, he'll pout and moan unless you accept the precise combination of pseudo-truths about past, present and future spun out of his silly little head.
Still, a World of Warcraft server somewhere is a little less tedious when he's gassing up these threads.
Comments closed July 10, 2008.
| 
Heads in the Sand
If we don't want to see a huge proportion of the population immobilized, we're going to need more ways for people to get around.
Good point. The senior population, along with other folks simply priced out of the system, would be a severe economic burden. An older relative of mine, long an opponent to anything publicly funded, changed his tune about transit after surgery threatened his sight.
One of the problems with the arguements against transit is that it assumes that everybody will be able, physically and economically, to drive some sort of fancy new miracle vehicle without regard to ability to drive or to pay for it.
Posted by RoboticGhost | June 26, 2008 9:03 AM