My tour of Germany, France, and the Netherlands in July brought me to numerous situations where trams are used to great effect in handling high volumes of passengers moving in exclusive rights-of-way. (I cannot emphasize too often that these are usually more like light rail than like US streetcars or Australian trams, which are often compromised by having to share a traffic lane.)
I spoke to many transit experts on the trip, and only in Munich did I hear a planner state the view that it may make sense to convert buses into trams solely because trams can attract more riders. (Note that Prof. Patrick Condon also wants to replace buses with trams, but he’s thinking about a development and urbanist outcome. He’s also arguing, as my Munich contact would not, that it may be good for transit to be slower.) This planner recently wrote to me to share the following data, showing that German cities with trams have higher per capita ridership than cities that run only with buses. It’s too big to copy here, but it’s a simple Word file:
Download Cities with or without Tram
The result is, that cities with a tram system have in average 50% more passengers in the total system (tram and bus) than cities with a bus only system.
Cities like Wiesbaden with an extensive bus lane system are among them.
OK. But if you’re going to throw these kind of data around as though they’re decisive, please remember:
Correlation is not causation, nor does it tell you which way causation goes.
In this case, the correlation may be more easily explained by a causation that goes the other way, i.e. that ridership is “causing” trams. The best reason to convert a bus to a tram, or to build trams instead of a bus line, is because you need a higher capacity — in riders / driver — than you can handle on buses. It’s perfectly logical that high-demand corridors have trams rather than buses, but that doesn’t mean the trams caused the ridership. It may be the opposite or a complex mix.
Photo: MaxM via Wikipedia
Which happens to be a perfect explanation for the difference in success levels for high speed rail systems.
All successful high speed rail systems were capacity expansions for crowded low speed rail systems. All failing systems were constructed under the impression that speed causes ridership.
Right. Another point is that things like express trains (in NYC, in particular) were not made to get people to their destinations more quickly. They were simply made to move more people, i.e. increase capacity.
As far as Germany goes, I can tell you that the citizens of last city on the list, Jena, really and truly loves their tram system. They even have the “Partybahn” – an old-fashioned tram with a bar inside, that runs every friday afternoon, and is also used for charters. They also have a transit gift shop with paper model trams, ties, mugs, the works. And this isn’t a very touristy town, though it does have a major university. Which is to say that there is some HOT emotional stuff going on in Jena for their beloved Strassenbahn. Though maybe some of this is because they are apparently the smallest city in Germany with such a system.
Then we can all agree that Ottawa’s Transitway is the result of high transit ridership in Ottawa rather than the other way around like the entire BRT promotion industry wants everyone to believe.
Because if there’s one thing the stats in Ottawa do show, it’s that the Transitway didn’t increase ridership.
Now I wonder what cities in the world have busway systems modeled on Ottawa that were “sold” in part or in whole on the concept on the basis of Ottawa’s high transit ridership without anyone telling the prospective cities that Ottawa already had high transit ridership before the Transitway was built. Do any such cities come to mind, I wonder?
Honestly, this isn’t an either-or, folks. It’s likely some of both.
Many transit advocates readily accept the theory of induced demand (“if you build it, they will come”) when it comes to freeway construction–noting that a big problem with building new freeways to “eliminate congestion” is that the new freeway frequently attracts additional motorists. You even get lots of FOD (Freeway Oriented Development) around the line. That said–new freeways are proposed when the existing roads (freeway or not) become crowded.
And so it is with (high quality) transit as well. Building new quality services, or adding capacity to existing ones, will attract additional ridership–even though it is often done as a response to existing capacity issues.
The two twin fallacies are:
1) Pretending that induced demand does NOT exist–the “existing service handles the load well enough; we can just add an additional bus or two during rush hour” argument, from folks who would really rather not expand the capacity;
2) Building transit only to satisfy induced demand, in the lack of some other explicit goal–many TOD boondoggles are the result of this. There are legitimate reasons to pursue this course–reduce auto usage, improve operational efficiency, improve service reliability/perforamnce, improve environmental outcomes, and expand the system’s political support base. These reasons need to be clearly explained and articulated.
In regards to Danny’s comments–I wouldn’t go quite so far–many successful transit systems are successful due to extensive induced demand, by virtue of providing a more attractive service. Certainly, switching to a higher capacity mode on an existing overcrowded corridor is a slam-dunk, but there are numerous examples of successful rapid transit lines which attract far more riders than the POBS they replaced. There are also, as you note, plenty of white elephants around as well–which is why a good answer to #2 is needed.
The dataset presented is missing two key things: total service area for each operator, and some indication of service level. Overall, the calculated million riders per 100,000 (another way of stating annual rides per capita) in most cases is probably exaggerated to some degree.
Some indication of average loads per vehicle kilometer would also be useful. Then it could be compared to the extensive rail and bus occupancy data that Leroy Demery and I have developed at http://www.publictransit.us.
The Aug. 2010 issue of “Tramways & Urban Transit” has a short piece on the Cologne rail and tram system. To quote the article, after World War II, “…a policy not only to persist with tramways but to gradually free them from sharing road space would ultimately create one of Europe’s best light rail systems.”
I had the pleasure of living and working in Cologne for three months in 1972, four years after the first underground section was opened. Although I had a motorcycle, I found it generally more convenient to ride transit from my apartment south of downtown to my warehouse job to the north. The trip involved a short change of vehicles, either bus to tram or tram to tram, and I could either take the surface ring route or the subway.
My experience there supports everything that Jarrett has to say about the subject. I was subsequently appointed to the Citizens Advisory Committee for Portland’s first light rail line, which was then expected to be either an express busway (by TriMet and the City of Portland) or an HOV lane (by the Oregon Department of Transportation.) Based on my German experience (note that I experienced the system as a resident, not a tourist, something Jarrett has commented on elsewhere) I fought hard to convince the engineers and planners to consider light rail.
My own support for light rail was based on the system I had seen, but the argument was also made regarding rail’s inherent attractiveness over buses. This latter argument was disputed by the bureaucrats, and did not seem to sway the decision to build Portland’s first light rail line, which I believe was ultimately made on the basis of projected long-term operating costs.
It is good to read that Cologne is continuing to upgrade their system with additional subway segments and other improvements. It is unfortunate that US transit systems seem unable to graps the essentials of the Cologne (European) model and are instead hung up on mode arguments, point-to-point advocacy, and land development potential, among others. As I see it, ridership is what builds public support for transit, not mode.
The unknown planner from Munich compiling this data about passengers in cities with or without a tram system is me. Indeed the “disease” to build trams to attract more riders is widely spread in Germany, and we believe it does work.
The newest line that went into operation in December 2009 in Munich (Tram 23) attracts already over 50% more passengers than the bus line operating there before.
In Bielefeld, a city also on the list with a rather low total public transit ridership, had an increase of ridership of 127% when the light rail line 4 opened there, compared to the bus service on the same corridor before. Despite the fact that the tram operates less frequent than the bus operating there before and is not so incredible faster as you might think.
The reason why I looked at whole cities and made an average of cities with – in average – a comparable size on both “sides” and did not compare single transit corridors is, that I wanted to eliminate the effects of tram lines built only in high demand corridors. So I wanted to look at the effects of tram as a part of a total system on the ridership.
When you have cities of a similar size and density usually you should expect a similar ridership of public transit. But obviously, in average, it is not like that. So I sent Jarrett the data as one indicator for this “tram bonus”, as we call it, that is able to get people out of their cars, like a bus system would never be able to do it.
Even the successful BRT systems in different cities do usually just channel a demand that already exist, a high demand of captive riders. Who ever has a car in those cities still uses it.
We believe in the ability of a tram system to attract more people, people who would never set a foot into a bus, and we invest a lot of money in that. And: no matter if we are right or wrong with this “belief”, it does work indeed. We see the numbers, we see the ticket sales and it makes sense, even from the economic perspective. It does not make sense for any bus line, but you can expect to increase ridership in a dimension of 50% to 100% on a line, compared to the bus service. If that makes sense for you, then do it.
And it does not not only work like that in Munich or Bielefeld or Karlsruhe or Freiburg or… And when I can trust my colleagues from France, they have a similar expirience. The same is true for Spain. BRT usually is a huge success in third world countries, without enough funding for a real transit system and enough riders without car willing to accept any improvement. Considering the growing economic problems of some regions in the US, it might be an option for cities there.
And, by the way, we still invest heavily into the bus system – with bus prioritization (already at over 200 of Munich’s traffic lights and growing), new vehicles and new, specially designed bus stops and in 2004 an extensive restructuring of our bus system. And we use ,since 1987, low floor buses as a worldwide pioneer at that time.
Modern low floor buses were developed by Munich’s public transit authority together with a bus manufacturer. Our ridership in the bus system also increases. But never as it does when we convert a bus line into a light rail/tram line. So I think we know what a bus system can do and what not.
Thanks for the additiaional thoughts on this!
I certainly wouldn’t use the term “disease.” If converting buses to trams without making other mobility improvements increases ridership enough to make it worthwhile, then of course you should do it. My role was to register some skepticism about the ability to compare data across cities to nail down such a conclusion.
On the other hand, the bus/tram difference in a city often aligns with many other differences in presentation and priority, and these make it difficult to sort out the effects. For example, Munich puts lots of emphasis on its map of just the tram network, which shows all the trams but not the “buses that are as fast, frequent, and reliable as trams”. See the PDF link here:
This is interesting because the trams are mostly radial and the high-end buses (50-series) are mostly orbital. So often, looking at the tram map, you’ll get the impression that you have to go into the city and back out to get where you’re going, when in fact a high-quality orbital bus would get you there faster, with the same average speed, frequency, nad reliability.
None of which is saying that Munich’s isn’t a great system that many cities can learn from!
Correlation can still be a good tool for unearthing possible mechanisms of causation.
I would agree with “TransitPlannerMunich” that experience with the conversion of a bus route to tram, in a city with a mature system of bus and rail, can provide insight into whether passengers prefer rail over bus.
The implementation of the “Yellow” light rail line in Portland Oregon also showed that a rail line that is shorter and less frequent than the bus service that it replaced can attract higher ridership, no matter how irrational this may seem.
Canada should offer up data on the bus-to-rail issue in two cities this decade: Calgary and Ottawa.
Calgary’s West LRT project ( http://westlrt.ca ) currently under construction will replace a “themed” BRT service and some direct-to-downtown bus services with light rail in the western parts of the city. This is Calgary’s largest ever infrastructure project and its length brings it to the western fringes of the city in one go, so it will completely alter the transit network in that part of the city. It is to open by the 2013 New Year and it will be interesting to see how ridership changes.
Ottawa will be replacing part of its bus Transitway with light rail sometime later this decade. The plan ( http://ottawalightrail.ca ) at the moment calls for a tunnel through downtown Ottawa. The section of the Transitway network being converted goes through almost entirely non-residential land uses, so there is little prospect for a significant increase in walk-on ridership. Anyone accessing the line will have to transfer from a bus. If ever there was a test for the “transfers reduce ridership” argument, this will be it.
How does the Yellow compare with the old 5/Interstate in terms of the other performance parameters–speed and reliability? While the Yellow is kinda slow for a train with exclusive ROW, it does get signal priority along Interstate, and doesn’t (usually) get stuck in traffic.
you picked the tram network map of Munich and complained about the bus lines that are missing there. Well, the buses are missing there because it is a tram network map.
The buses are part of the schematic city network map – Innenraumnetz (subway, commuter rail, tram and metro bus system):
You might also like the geographic map with all lines in the city, including all bus lines:
I had the pleasure to work in Karlsruhe for a while, with one of the best transit systems in the world for a city of that size. And a ridership of over 100 mil. people anually on a population of about 280,000 people. The head of Karlsruhe’s transit authority at that time, Dieter Ludwig, liked to say that when you need to run articulated buses with a headway of 10 minutes on a line, you have enough potential to convert the bus line to a tram/light rail line.
Not because you already had so many passengers that you could not handle them anymore by bus, but because you could attract so many more passengers to justify, fill and financially operate a tram line. Just because it is then a tram line and not a bus line. And believe it or not: it worked!
Certainly my data is just an indication and not a proof of the improtance to build an attractive rail system to attract customers. In the German publication “Nahverkehr” (9/2009, page 34-41) you will find an article about the success of french tram systems in comparison to bus only systems. The result is that in all cases the passenger numbers in the total system were rising in a dimension between 50% and 100% in those cities introducing a tram compared to bus only systems.
The reason why in most German cities you have a tram system and in some not is mainly a political one. To abandon the tram was not one of reason or logic. The decision against a tram was basing on the emotional fact that trams were seen as obstacles for the motorization and is old fashioned, in some cities. And that kind of “logic” is still in the head of some politicians.
It should be easy to test this hypothesis. Shut down a tram line for 3-6 months and replace it with buses at identical frequency running along the same alignment.
Does the patronage stay the same, increase or decrease? Repeat the experiment at different times of the year, on different lines and in different cities and look at the data.
If low floor trams and buses increase patronage, then this would suggest that the vehicle does have at least some effect on ridership.
Interesting read. Thanks for sharing the info.
Some reasons for the correlation may be subtler. I think it was Steve Munro who pointed out on his blog that riding the rails is simply more comfortable: the acceleration profile of electrically-powered vehicles is a lot smoother, and the straight line and predictable curves of rail don’t jostle you around as much as a bus. Rails don’t have potholes. For seniors or people with young children this is more than just a modal perk. Perhaps we are focusing too much on hard measures such as frequency, capacity, etc.
Also, trams are predictable, and I mean that on many levels. Witness the 160 bus-only systems in the US that are cutting service. Atlanta will axe 40 of its 131 bus lines, according to this week’s Economist. Ripping up the rails is a lot harder, both physically and politically. As a cyclist in Toronto, streetcar tracks pose a number of dangers to me, but overall I like cycling next to streetcars than buses. Streetcars and bikes complement one another in their abilities to slow down and calm auto traffic, whereas buses and cyclists are always jockeying for position in the right lane; with a streetcar, I know exactly where it’s going and can safely ride much closer to it.
Danny: All failing [high-speed rail] systems were constructed under the impression that speed causes ridership.
I disagree. When users have to choose between two competeing modes, travel time is probably the biggest single factor. If you improve the speed of particular mode choice, then it will gain market share. (I’ve seen studies which state that a 1% decrease in journey time leads to a revenue increase of 0.9-1%).
@Scotty: Portland’s former line 5 was definitely affected by traffic due to the portion on I-5 to Vancouver, Washington, so agreed, the Yellow Line isn’t a perfectly controlled experiment.
This fall, Sept. 13-26, the Portland Streetcar will be replaced by bus service due to construction. While it is too short a period for ridership to stabilize, this will be an uncontrolled experiment in short term rider preference between buses and US style “modern” streetcar service.
What are the failing high-speed rail system? I’ve never heard of one.
Even the Taiwan HSR is getting better and its failure was mostly due to design errors and debt servicing.
And speed caused a modal shift from air (and maybe road) to rail. France TGVs saved SNCF’s passenger Intercity division.
Did you take in consideration that, when a tram line open, bus routes are rerouted to fed the new line and therefore increase the number of passenger on the studied corridor?
I have the same question as samussas regarding possible network structure changes when a tram or light rail line is introduced. I’m pretty sure Portland changed the bus system to feed the yellow line when it opened.
I’ve written a DHT article which expands on this one. (And I must confess to being even more verbose than Jarrett…)
Yes. It is important to take into consideration if really new passengers are generated. This is why I found it neccessary to look at complete transit systems.
In case of the mentioned article on French cities compared before and after their bus only system was changed to a tram and bus system the conclusion was that within a few years the public transit ridership in the whole system of those cities rose by 50% to 100%.
In Munich we also look at the total network in an area that is affected by a new tram line. And when, in case of Munich’s new tram line 23, suddenly the ridership in the total area goes up dramatically (tram and bus lines combined) then certainly those passengers were not there before in public transit and are new riders.
Also often we have a reduced frequency when a tram is introduced compared to the bus line before, cause the tram has a much larger capacity. So the change of frequency cannot attract new riders, cause with a new tram system often it does not get more frequent. But even that might change fast – like in the case of Tram 17 in Munich, where ridership exploded after opening, so that now even a 5 minutes headway is not sufficient anymore.
Why do people prefer light rail or trams?
It is hard to say. Average speed in the Munich system is similar (tram 20 km/h, bus 18 km/h), both have priority at traffic lights, both have modern low floor vehicles and in the case of the bus system mostly articulated buses with boarding on all doors and a proof-of-payment philosophy (also on the standard buses, only after 9 PM you have to show your ticket to the driver). Tickets are sold at some stops and by ticket machines on board of all buses.
Still people see a rail system as a high value system. They see it as an alternative to the use of their car, what a bus system definitely is not. Even in Munich you have in average a different strucure of people in buses (more students, older people, high amount of captive riders) than on the metro or tram system.
No matter why rail is more accepted, we cannot ignore that fact. And if we want to increase ridership we must expand the rail system.
By the way: Often when tracks have to be replaced we operate a tram line for several months with buses. Ridership is down then on those lines. That is also just an indicator and not a proof. But one of many.
By the way: I am somehow amused to have such discussions, cause in Europe usually transit planner agree since 20 years that trams will attract new riders and increase public transit ridership in the whole system.
@ John in Portland context: TriMet usually puts a trunk+feeder system in place long before a light rail line opens, so that not much has to change when the line opens apart from replacing the trunk bus with rail. On the Yellow Line in North Portland, the basic grid structure of east-west lines has been there since 1982. In the west, Tri-Met installed a trunk-feeder structure at Cedar Hills (now Sunset TC) in 1979, and this structure ran with buses until westside light rail opened in 1998. In Milwaukie, the basic structure has been there almost as long. It went in around 1984 if I remember right. So Milwaukie LRT will replace the 33 north of Milwaukie but will not require many other changes.
Ridership may increase on the tram if the potential demand is choice. But I doubt it will help if its mostly or entirely dependent riders.
Moreover, the effects on the entire transportation system should be examined to see if tram would improve service. Highly constrained street alignments might make the tram slower than a bus, in which case, it’s certainly not worth doing. Because the operating costs will increase due to more vehicles needed keep constant service frequency.
Also should millions be spent to attract riders who could drive for service that would be slower than a bus so that dependent riders must suffer?
I know I keep harping on this, but it’s fundamental on environmental justice grounds. Converting bus to tram must either
A) decrease unit costs while maintaining service (the benefit), or
B) increase benefit–i.e. shorter door to door trip time
Or preferably both. I think it is sometimes arrogance on the part of some planners or bureaucrats to plan systems to increase riders that ultimately makes transit worse for existing riders.
If a new tram would make transit worse, then the planners did something fundamentally wrong.
In Europe I am not aware of a single case where that happened.
Also I do not understand why a tram should be slower than a bus. If that is the case then you should change the planners.
All new tram systems and lines I know have happy riders and exploding ridership.
I remember that when the U-Bahn was being built in Munich in the 1970’s and 1980’s, the initial plan was that all tram service would eventually be terminated, being replaced either by U-Bahn lines or by buses. I also remember at some point in the 1980’s that there were studies done that showed that when tram lines were replaced by buses around 50% of the ridership permanently disappeared. This was why eventually the bustitutions were stopped, and the commitment to trams was renewed.
Are there any studies or reports or even raw data from the time of those bustitutions, or perhaps more recently from tram expansions which objectively document the tram bonus?
Trams are often slower than buses, it is the nature of the modality. I need not explain that here, eh? It is not necessarily based on the planners competence.
Any assumption ridership will explode because there is a tram, per se, is foolhardy. There’s got to be consideration of socio-economics. The US is certainly different than Germany.
“Trams are often slower than buses”?! You really need to explain that here. That’s the first time I ever heard that asumtion.
In mixed traffic, busses have an advantage of being able to maneuver around obstacles.
In dedicated lanes, though, I can’t think of any reason why (or for the inverse), especially when traction is not part of the equation. Electric motors are generally capable of smoother, faster acceleration than are combustion engines–but the tire/concrete interface has a higher friction coefficient than the steel/steel interface–making acceleration faster, but decreasing energy efficiency at steady speeds.
And of course, one also must control for the capacity and design of the vehicle’s powertrain. Portland Streetcar uses underpowered/governor-limited trams, for instance, to present a smoother acceleration profile to riders. Given the mixed-traffic nature of the line, this probably doesn’t affect performance much, but it would make the Skoda rolling stock less desirable for a faster line.
That trams can be slower than buses has been detailed on this blog profusely by now, so I’m not going to reiterate needlessly!
The thrust of the post revolves around the extra capacity trams provide over buses and the relationship with rider attraction and demand.
The most important factors for line performance (speed, reliability) are station spacing (and stop policy), vehicle priority, and ROW exclusivity–in other words, how often does the vehicle need to stop (to pick up passengers, waiting at intersections, or behind other traffic).
Even things like fare collection policy and platform design are arguably more important than either the type of the powertrain or the type of the wheels.
Trams slower than buses? I know that phenomenon only from Ukraine, due to a lack of maintenance, but if it is a North American phenomenon also, then you do have a problem. But I doubt it that it could be like that.
I remember one day when a German consultant gave a presentation about public transit in Lviv, a western ukrainian city. A quite nice city by the way, as I can assure you from personal expirience. He told the audience how a lecturer of a local university was convinced that trams are ideal for the poor who are satisfied with slow, bad service, while buses are for some kind of better service. It was the joke of the day during that conference, everyone was laughing out loud. Sad that there are other parts of the world were it is like that. Cause regarding Lviv it is understandable, when you look at the lack of funding and maintenance.
A well designed tram system should be faster, and more reliable than a bus.
Well, I don’t see any evidence of that in the US, TPM. Moreover, many European cities were built-up prior to the Industrial Revolution. They do not have the highly productive economy of fast-moving and high-volume goods and services. The street layout of many European cities is so small scale, they never developed the high volume vehicle and truck traffic that is common in the US and its economy, allowing trams with dedicated lanes to thrive. It’s certainly not possible to have fully dedicated tram services on the arterials here in the US where there are currently only 2 lanes of auto traffic and 2 lanes of parking. So a traditional trolley is the only surface rail option. Therefore, its operation will be quite hindered and likely slower than bus service.
Certainly, there are many European cities that do have large industrial economies, but most of them have metros, too, which generates higher ridership anyway and somewhat obviates the need to expedite the trams for speed purposes.
Germany is the tiny nation in Central Europe with an export of well above 1,000 billion dollars annually. Even the balance of trade with an amount of over 200 billion dollars surplus compared to the import) is bigger than the total export of the US in a year, so far as I know. At least until last year our export was larger than the export of China.
But sure, Germany has not such a mature economy with such a high productivity like the US, so we can let toy trams operate in our toy cities. And that even with a nice speed…
TPM, certainly the US economy is pathetic now, and we don’t manufacture much either. You misrepresented what I said. (Germany is a high density nation with many cities, so it should have a large output.) But the socio-economic discrepancy with the US and the fact the fact that European cities like those in Germany have a different development history make legitimate comparisons very difficult.
The traffic loads and ROWs might make tram service no worse or better than bus there, but not in the US so one shouldn’t assume similar results. There was an earlier post on this blog about that. And again the wealth discrepancy in Germany does appear to be high compared against the US. The Germans who build and can afford to buy their cars might skew the bus/rail preference quite a bit since many would be choice riders.
What TransitPlannerMunich is referring to as a tram has a lot more to do with what we would call light rail than what we call a tram. Think MAX, not the Streetcar.
Sure, I understand that, but in that case, they would not really be in a contest with local bus service if they operate at similar stop density. Based on the original post, the comparison is still basically local bus versus local rail, even if the there are exclusive lanes for the rail.
Trams do have one advantage over buses: they don’t have to pull into traffic when they’re leaving a stop. That can delay a bus by 15-30 seconds. Per stop, which adds up. So the occasional ability to maneuver around obstacles is traded for the necessity to yield to cars. And in Europe, much of the time roads are narrower, and you don’t really have the room for a bus to get around an obstacle anyway.
Also, I’ve said this like 3 times before, but I’ll say it again: the Portland streetcar is almost pathetically slow. The vehicles do 20-25 mph (32-40 kph) typically and top out at 30 mph (48 kph). Most European trams can do 50 kph on local streets and up to 70 on private reservations.
Two words. Curb extensions. It’s perfectly viable to build curb extensions so busses stop in the main traffic lane, rather than pulling to the curb, and pulling out. (And in many cities, including Portland, busses pulling out of a stop legally have the ROW over cars in the lane; although enforcement of this is sporadic and violation is common).
Having curb extensions also has the advantage of slowing down auto traffic, which makes the bus more competitive overall. Of course, it annoys motorists to no end, who hate getting stuck behind stopped busses.
OTOH, if you look at the MAX tracks in the Transit Mall, what do you see? The trains switch back and forth between the left and right lanes–for one block, the train is against the curb so it can load/unload passengers, and busses can pass it; for the next 3 blocks or so, it’s in the left lane so it (and other busses) can pass busses loading at the curb.
A point Jarrett has long tried to make: Don’t confuse common transit agency practice with technical limitation. Just because busses operate a certain way, doesn’t mean they have to.
The fact that the streets in many European cities are narrow (something which is also true of some older US cities such as Boston) seems to be something which would hinder dedicated-lane surface transit, not help it. The reason dedicated surface transit lanes are rare in US cities is a lack of political will–so many shopkeepers and homeowners are scared to death of any loss of capacity for auto mobility or parking. Even in Portland, where the political consensus is decidedly pro-transit, downtown merchants fight any attempt to reduce car lanes or parking stalls tooth and nail.
Again, much of the confusion in the conversation with TransitPlannerMunich is that when he says "tram," North Americans think of the slow American streetcar in mixed traffic. The Munich trams are more like what North Americans call light rail, almost all exclusive ROW and with stop spacing of (I'm guessing) 400m or so in the core and even wider further out.
TransitPlannerMunich, the reason you hear claims that trams can be slower than buses is I've made that claim as it applies to American and Australian mixed-traffic streetcars, not as it applies to light rail or standard German trams. When it comes to travel time, streetcar in mixed traffic have all the same limitations as a bus in mixed traffic, plus one other big one: the inability to maneuver around obstacles that occur in their lane.
One thing people keep seeming to forget is that in Germany most tram infrastructure hasn’t replaced buses, but replaced older trams. Therefore the discussion about whether there’s a tram bonus in the US based on German evidence is flawed. The “tram bonus” in Germany has been achieved through retaining high levels of transit use, not in regenerating them. Who wouldn’t pick an upgrade from old tram to mostly-dedicated modern tramway over an unsegregated bus service? If a bonus does exist in the US it’s about trams being able to attract new passengers who were not previously using the transit system, which could be a completely different thing.
Could the explanation for the tram bonus be that we view making tram connections differently to making bus connections? So if I have to make a journey via a highly segregated light rail line and a unsegregated or semi-segregated streetcar my positive view of the former bleeds into my impression of the latter because I percieve them to be part of the same journey. If this is the case I will choose to ride an objectively lower quality streetcar because it’s only part of my otherwise positive rail journey. By contrast if I have to connect to a bus, perhaps I regard the two modes as two distinct journeys, a positive rail journey and a negative bus journey? So the German cities that scrapped their tram systems lost riders because they destroyed the illusion of a seamless journey.
If this is the case (and I will accept it’s just idle speculation) then I’d not expect to see a tram bonus in US streetcar cities unless the streetcar system appears to provide seamless connections with a higher quality light rail system.
Good point. But in the case of French cities you find cities where tram service was discontinued during the 50ies and re-introduced in the 90ies. And ridership in the total system usually doubled.
In Munich (and other German cities) we do have sections where the tram is mixed with other traffic. I just look on such a street from my window. Still service usually runs on time.
One thing that improved service quality a lot was priority at traffic lights – the tram sends a radio signal from different position when approaching a traffic light so that in most cases it can pass without any delay.
And it is true: tram stops should not be too close, so that the tram can reach a nice speed (up to 50 to 70 kph between the stops) so that you have an average travel speed of 20 kph. I even think bus stops should not be too close. 350 to 400 meters are a nice balance.
In Germany there are many examples of regional rail branch lines that reopened after decades of disuse. Those often see double or triple the ridership of the buses they replaced.
I know one example of a separate-lane LRT achieving lower speeds than the bus. This is the Central Subway under construction in San Francisco, a project that nearly all the local transit activists hate and that Muni is only building to put a line on a map that goes to Chinatown.
Elsewhere, LRT generally has speed advantage over buses, especially when both have separate lanes. At high levels of demand, LRT would run longer trains, whereas BRT would run multiple buses, which would get in each other’s way. The ability to run long trains also means that at equal operating cost, LRT would have more doors per passenger, reducing dwell times. But both effects are small.
The ridership advantage of well-done LRT has nothing to do with subways. Karlsruhe, Geneva, and Zurich have no subways; their LRT services still generate some of the highest transit mode shares in the first world. Calgary doesn’t have any transit other than LRT spine lines and connecting buses; its transit mode share is lower, but is rising very quickly, and the C-Train’s operating costs put Tokyo Metro to shame.
European modern “trams” could operate like american LRT, but their geography scope is the one of a streetcar (line length)
Munich has 11 trams route for 75 km of track.
Sacramento LRT has 2 line for 60km of track
San Diego Trolleys has 3 line for 83 km of track
those speak enough of the difference between the Europe and US:
Simply you are trying to compare 2 different systems, and the reason of success of one, which TransitPlannerMunich humbly recognizes is not well understood, doesn’t necessarily apply to the other. Indeed, usually well below prediction ridership of US LRT line, are in striking contrast of the usually well above prediction ridership of the European one:
At least speaking of France, the instant success of trams in Nantes, Grenoble, Strasbourg has took people per surprise.
That says one should note usually the tram in a European city don’t come as a “naked” product, it come with a whole urban vision:
Introduction of the trams in Strasbourg has been synonym of complete redesign of vehicular traffic, and basically you can’t cross the city by car anymore – period !
At least for France, massive huge of P&R has been part of the equation.
Could it work in US: if you have been in Strasbourg you understand why people like to go to shop and play there: Could people will be willing to do the same effort for Sacramento DT?
Also, bus system is more often than not deeply redesigned to support the increase in ridership, attracting by itself more ridership: a virtuous circle.
As an example, I could mention the bus route 1 in Grenoble, put at 70% in bus lane only: result has been an increase in ridership of 35%, on this line alone.
But, if you go to Oxford, UK: ridership has increased by 80% too: No trams there, but a urban politic similar to city with trams.
And I should also cite the own Vancouver success story which is the 99 B express bus route introduction:
predicted traffic in 96: 9,000.
quickly 30% above prediction
today traffic (before Canada line) is around 45,000 on a line shorter than the original one !
Allon, to say that the Geneva or Zurich LRT “Generate some of the highest transit mode shares in the first world” is a bit misleading:
If there is no subway there, they are part of an overall system, like the S-bahn in Zurich…
And taking a part of it to explain the success of the whole is a pretty risky business: see http://voony.wordpress.com/2010/04/26/thezurichmodel/
As a complement, to support my previous comment with eventually a less contentious example than the trams.
Pedestrianization of street has been virtually an universal success in Europe. You could draw the radically opposite conclusion in US.
So there is some choice an urban planner could take as granted in Europe while thing could be still reasonably be argued in a North American environment… And when it come to the tram, it could be the “tram bonus”.
Isn’t the limit for accelaration speeds ride comfort rather than technology knowadays, that is passengers should for example be able to stand easily while the tram accelerates? Which brings to my point
There is good speed and bad speed. Exclusive rights of ways, signal priorities, modern vehicles, separate express routes if needed, etc. are good ways to increase speed. Grade separation, increasing stop spacing, connections, and aggressive driving styles can be and often are bad speed improvements because the speed gains do not outweight passenger inconvenience generated by them.
I don’t know of a single service industry actually selling to consumers where serious grown up people believe service quality can be measured by few simple easy to collect metrics. Apart form transit that is. Yes, speed and frequency are very important, but they are not all that there is to service quality.
What the success of trams tells me is that ride quality and comfort that Bradley Wentworth points to above is seriously underrated. Look at the competition: Car industry for the last couple decades has not been focused on building faster cars. They are building bigger cars, with more space, more comfortable seats, more natural light, better airconditioning, smoother suspension, automatic transmission, cup holders, personal digital devices, etc.
One fun statistical finding is that if you look at the average speeds of trams systems there seems to be if any a negative correlation to ridership. One reason for this is that good systems sometimes sacrafice speed to reduce walking distances to stops, that is have for example 500m instead of 800m stop spacing. Once again, look at cars, drivers spend a lot of time to get just a bit closer to their destination.
All of the above of course does not matter if you are just trying to serve captive riders, they’ll use the system no matter how bad it is by definition. But you can’t get kind of modal shares succesfull European cities by just serving captive riders.
When you want to get not only captive riders but grown up people with a choice, you need to treat people not like convicts or school children, so do not treat them like “captive riders”. By the way: even captive riders have often the possibility to change, they grow up and buy a car or go by bike if you treat them badly.
When you expect that everyone has to board only at the front door, presenting the ticket to the driver as if you always suspect that he would be a fare dodger it can be humiliating (and slows down boarding time). If you expect that people can be educated to wait for a bus coming every 30 minutes and inform well ahead it will not work. I spent some time of my life in Calgary – the C-train was nice, but I remember that at that time you had to call a certain number indicated at the bus stop to get information when the next bus will arrive, cause there was no timetable on display. And that at -35 degrees celsius at night.
And you cannot expect that someone memorizes a comples network of 50 or 100 bus lines.
A bus system should be as intuitive as possible and as easy understandable as possible. (That is one of the reasons for the success of tram systems). A timetable and a system map at every stop is a plus. A frequency of at least every 10 minutes even off peak hours is vital. At least in a city like Munich a bus just going every 20 minutes is more like something you need to do to serve a neigborhood cause the city wants it. But you get just captive riders.
No one wants to think about public transit and go into its details. The best information material you can have is that information that you do not need to give, cause the system is simple enough. When you have every 5 or 10 minutes a vehicle on a line and when you change consequently no real waiting time, you do not need to supply complicated trip planners. Most people do not want to go into the internet before they go to the movies to see when to go and when to go back.
People chose a mode of transport cause it makes life for them easier. And not more complicated. And I guess not even in Portland people would ride by bus to save the environment. Except maybe a dozen who are caught then riding their bike instead waiting for the bus.
@TransitPlanner. Most of this last comment, though, is about issues that are not intrinsic to the nature of buses vs trams. They're just about the cultural history of the two technologies.
I agree completely about "captive riders." See here: https://www.humantransit.org/2010/01/unhelpful-word-watch-captive-rider.html
Lets just do the experiment. Which tram line should get shut down?
Voony: you know, I actually had your blog post in mind when I wrote my comment. In response, I’d say that,
– While the spine of Zurich’s transit system is the S-Bahn, there’s also a lot of independent tram ridership. Look at the numbers: the metro area has a 40% transit mode share, but its weekday S-Bahn ridership is only 350,000 against a population of 1.3 million.
– I brought up Calgary specifically as an example of a successful system using LRT as a spine, as the mainline rail system doesn’t go to enough places and is hobbled by American-style regulations. Calgary’s transit mode share is only 16% (amazing by North American standards, crap by European ones), but it’s increasing by a few percent every decade.
Just a new example from Jena, Germany:
Since the extension of tram line 3 to Göschwitz the passenger numbers rose impressively.
As an example they give for the stops one the extension, that were served before by a bus line (linie 13), the number of daily 321 passengers boarding at those stops during bus operation. Since the tram was extended there, the number of boarding passengers increased within 6 months to 2433 passengers daily on those mentioned stops. The old bus line 13 had also a direct access to the center of the city, with more than 8 trips offered during peak hours.
Just look at before and after data from Boston, MA. Check out the E/39 and A/57.
Also (ibidem), ripping of the tracks may be CORRELATED with unfair curbing of service. The T’s own documents showed that the 57 was running an operating surplus in the late 1990’s, but that didn’t stop them from cutting night-time service, to the detriment of those residents who depend on that bus.
There’s plenty of evidence linking light rail service in a corridor to higher ridership. The big question is whether BRT can be implemented in such a way as to yield the same gains as LRT. I’d give that a qualified yes, as only some BRT projects (and none that I’ve seen so far in the States) come close to matching LRT infrastructure (stations), speed, and ride comfort.
Hey Jarrett, come on.
I don’t think that it “may be good for transit to be slower.” What nut would say that?
I say that the race for speed for the sake of speed, absent rational land use, is unsustainable and counter productive.
Why pour all your money into speed without thinking what are we gaining?
I think that its clearly better to think synthetically about the optimum relationship between land use and transit. If you got that right the average (shorter) trip might take less time (and require less resources and produce less GHG) in a system served by lots of trams than one with no trams and only skytrain and diesel buses.
And remember…i dont say get rid of the backbone. I only ask that you think about when you have enough backbone and start spending your money on something that fits the needs of the 22nd century.
What is so hard about that?
Well, your published work does explicitly compare an intrinsically faster service (SkyTrain) with an intrinsically slower one (streetcars) and you didn't quarrel when I called my review of the subject "Is Speed Obsolete?"
I understand that you are proposing that a transit service be selected based on its affinity with a land use outcome that you prefer, and that this differs from the way any transit planner will frame the question even if he shares your land use goal. The transit planner will first notice the mobility difference implied by building a streetcar instead of SkyTrain to UBC, which is that for a number of high-demand markets to UBC transit will indeed be slower.
So while I realise you don't mean to have said "transit should be slower," you have actually proposed a slower service over a faster one, and suggested making this choice not just for Broadway but as an example of a proposed general principle. Slower speeds are a direct consequence of a specific course of action that you've proposed.
Also should millions be spent to attract riders who could drive for service that would be slower than a bus so that dependent riders must suffer?
I take strong exception to this as someone who rode both the 39 bus and the E line (Jamaica Plain, Mass.). I can tell you that compared to riding the E trolley, it was the 39 where the suffering occurred. The bus would be packed well over the FTA’s recommended 150% seated capacity max load. The T’s fleet at the time was the RTS II which can easily haul over 200%. I say “easily,” but from the rider’s POV it’s hell. The air would get bad because the air system can’t keep up and you have diesel fumes backing up into the bus (what with it idling at stops because dwell time is lousy as the bus is crowded) and then the swaying of the bus just completes it so you’re sick to the point of having a pounding headache.
On the trolley you could have 3x the passengers and never care. You glide rather than ride. The only bad air systems were on the original Boeings but those have all been retro’d. Again, it’s not that the RTS II had a “bad” air system (actually, it’s quite nice), it’s just that it wasn’t designed to and can’t handle habitual 175% to 215% loads.
I know ’cause I didn’t just ride ’em … I drove ’em too.
It’s amazing what people who haven’t ridden big city transit systems day in and day out believe to be “suffering”.
PS: About 10 years back some “geniuses” at the T overruled the City, which wanted rail on the Silver Line (Washington St Corridor), and put in “BRT”. Well, the ridership was better than the cruddy 49 bus that preceded it, but it was a MISERABLE ride. Low floor CNG-powered artics, standing room only, people rolling around in the aisles and falling all over each other as the bus lurched around. It may have been faster than the infamously slow B trolley, but you didn’t get smashed into a pancake (and people rode the heck out of that B, slow or not).
It’s happening just now and quite close to you, but not with new tram, but with new metro line – I’m talking about ongoing Prague’s Metro A expansion to Motol.
How is the Metro A expansion making things worse?
From all what I know about the Metro A project in Prague it does make sense. Actually a lot of sense.
It’s great to hear that you finally got a chance to visit Munich!
As for the streetcars causing ridership–you may be right that it’s the causation is somewhere else in Germany. But, I would argue that in the United States streetcars would almost undoubtedly boost ridership due to class and race issues, as well as more than a generation of many 100% non-transit-users. People see buses as low status or only for (poor) minorities or as confusing and inaccessible.
In that way, I think streetcars can act as (to use a drug reference) “gateway” transit modes. The more people start riding streetcars who’ve never ridden transit, the more likely they are to branch out and try other forms of transit, especially if they realize a bus CAN get them somewhere even faster. You see this in New York or London where people new to transit start with the subway or tube because it’s perceived as better. Eventually some realize bus lines offer even better service in some cases. But they would never have started riding the bus if a rail line hadn’t gotten them used to using transit.
To Patrick Condon,
I’ve bought your book. Does it say Skytrain is the logical choice for the Broadway corridor?
Where’s the logic in Barsta pushing for a large sum of money to be spent replacing the local trolley service with a local streetcar service, which, for anyone familiar with Broadway, will slow things down……and ignores the crucial discussion on the mass transit and rapid transit needs of the corridor?
And from page 50 – “Here suffice it to say that slower average speed in a system that resists congestion and is compatible with urban uses is probably a good thing.”
@tomtakt. Indeed, there may be "class and race issues" involved in streetcar ridership outcomes in the US. My question, as always, is: How much long-term infrastructure do you want to build based on class and race issues? Don't you risk coding some unpleasant features of the current culture into the permanent infrastructure in a way that could perpetuate that unpleasantness? For that matter, how much long-term infrastrucutre do you want to build based on any contemporary cultural issue? Again, I point to the Interstate highway system as a great example of long-term infrastructure, built to serve the values of one generation and culture, that is turning out not to be very helpful to the generation and culture emerging now. I'm sure that off the record, we could probably name some less-than-effective transit capital projects whose design was influenced by the politics of class and race.
The whole basic contention of my Field guide to Transit Quarrels is just this: Culture changes; geometry doesn't. Emotionally derived human longings will affect our infrastructure, as they should. But at the same time: when you're building very long-term infrastructure, don't we have an obligation to push back against cultural pressures of the day, so as to ensure that we build something that makes sense as geometry? Streetcars in mixed traffic — as most American streetcars are — have a geometry problem that affects the mobility they can offer: they get stuck behind traffic obstacles that a bus could go around. No amount of culture enthusiasm, or culturally interpreted ridership evidence, is going to change that geometry.
David in Ottawa,
“The section of the Transitway network being converted goes through almost entirely non-residential land uses, so there is little prospect for a significant increase in walk-on ridership. Anyone accessing the line will have to transfer from a bus. If ever there was a test for the “transfers reduce ridership” argument, this will be it.”
I agree. That’s why the whole Transitway should be LRT. After my letter to the editor of the Citizen I published this January, I’m planning to write another letter soon in time for the October election. It is disappointing that the current plan in Ottawa calls for the Southeast Transitway from Hurdman to South Keys to be preserved as busway.
I agree, you make some valid arguments. I’ve never been to your fine country of Germany, but they are valid arguments, and the Tram 23 looks good from what I’ve heard of. My only streetcar experience comes from vacations to Toronto.
The problems were best summarized by series of blog posts by Martin Šubrt – in Czech. I don’t have enough time to translate it all, so I’ll write some main points without going to details:
0) the need for this extension was (and continues to be) much less severe than that of other ones. The most pressing one as of now is new line D that was killed by studying it to death (last time by wondering if some other technology wouldn’t be better than that in use on current lines; the line would connect southern suburbs with center of the city and it would sort of parallel the most heavily used part of line C [100 m trains in 115 s headways, at capacity])
1) the project was born from populism of local politicians, who used FUD tactics to kill upgrade of parallel railway line 120 to S-Bahn standards. The line connects biggest nearby city of Kladno, so it has considerable ridership potential, but it’s in bad shape, single-tracked and unelectrified, so it’s current share is low
2) the political FUD was based on assumption that the upgraded line wouldn’t serve well the Prague’s part of Praha 6 – untrue, the projected stop arrangement was better than that of metro and initial headways were planned in 10-15 min range. Current practice in Prague’s S-bahn is to minimize headways on lines that serve as city transit, so it’s likely that the headways would be shorter as of now.
3) the then-mayor of Praha 6 got elected as mayor of complete Prague in 2002. Since then, he started work to fulfill his old promises about Metro to the Airport. Since it’s ridership forecasts were too low to justify metro construction. To change it, the proposed line was altered to go near hospital in Motol, that wasn’t exactly on the way
4) as a result of previous point and lesser development patterns, station spacing and station entrance placement prevent reduction of surface transit to retain current mobility, so after the line is built, either total operating expenses spent will increase, or mobility will be harmed. All of this threatened co-financing from EU/EC.
5) the same applies to other phases with addition of long and indirect path, so the extension past Motol is unlikely to happen (the current transit is faster and the demand simply isn’t there). It’s no wonder that EU/EC refused co-financing and the planning of the extension was dropped for now
6) in spite of previous point, the Motol terminus wasn’t designed accordingly:
1. the only way to make peak turning of trains possible is combo of driverless and manned turnaround
2. keeping of reasonable speed of approaching trains forces pushing of turnout further out of station, further lengthening turnaround times
3. the tail tracks will be in double-track tunnel in curve with minimum-width service platform accessible only through tracks, so:
– it’s not possible to turn around by swithing drivers (when the second driver waits at position of new front cab)
– curve causes wide gaps and height differences, increasing risk of work-related injuries
– the access to service platforms will be lengthy and dangerous because the personnel is exposed to train traffic and they can’t step back, especially in non-standard cases
4. troubling placement of backup train
* it’s placed on the same track as the turning train, so if the latter breaks down, backup can’t kick in
* there will be no other tracks apart from “future” through tracks of another extension
All of this was to be temporary arrangement, but currently it looks the DPP will be stuck with it for long years, because of another extension uselessness and shifting of attention to line D and other projects
(pretty much of this point is translation of this forum post
I can’t think of any positive stuff that would outweigh these drawbacks. Unfortunately, Czech transit debates are largely unseen by the rest of the world (and vice-versa) because of low language literacy here, so the news are often limited to just optimistic statements by politicians.
@Jarrett. Point well taken. But then again, I don’t think they should run much transit in mixed traffic at all–that is, if there’s any traffic to speak of.
But, I could see some situations where giving the streetcar dedicated right of way could happen in the future, just like creating a bus lane. At least streetcar is cheaper than heavier rail to alter and expand, and even a short, bad stretch only popular among yuppies and tourists (read: Portland) can be the beginning of a network that gets better over time. But, I really have no hope for the U.S. until I see Rasengleise ;D
Bustitution of lines in the US “transit holocaust” period is evidence enough for the rider preference for trams. Mixed-traffic streetcars (decades old) were replaced with *brand new air conditioned* mixed traffic buses.
Ridership plummetted like a stone overnight, in Philadelphia, Boston, and elsewhere. This is easy enough to look up in historical records.
“Pedestrianization of street has been virtually an universal success in Europe. You could draw the radically opposite conclusion in US.”
Actually, it’s been quite successful in many places in the US. Times Square?
You have to pedestrianize the *right streets* however, which has been an ‘epic fail’ in many places. In a city without gridlock, targeting a street just *off* of the main car thoroughfare is going to work a lot better than blocking the main car thoroughfare.
Nathanael, late answer but:
Beside time Square, do you have other example to cite: Time Square looks to be the exception.
You have to pedestrianize the *right streets* however, which has been an ‘epic fail’ in many places. In a city without gridlock, targeting a street just *off* of the main car thoroughfare is going to work a lot better than blocking the main car thoroughfare.”
hum…Is not time square at the confluence of 7th and Broadway?
so the only successful example you give seems to be in plain and full contradiction with your statement.