# what’s wrong with the “transit score”?

Carter Rubin, on Los Angeles Metro's blog The Source, says exactly what needs to be said about WalkScore.com's Transit Score product.  The product is in the news today because of their new ranking of US cities by Transit Score.

Everyone’s favorite mapping tool, Walk Score, has launched a new service called Transit Score and used it to determine what it believes are the best American cities for public transit. The Switchboard blog has a recap and analysis. Los Angeles clocks in at 11, just behind Portland, Ore., and ahead of Denver, Colo. — respectable enough company, I suppose. That said, I do have some constructive criticisms of Transit Score’s methodology.

Excerpt:

To calculate a raw Transit Score, we sum the value of all of the nearby routes. The value of a route is defined as the service level (frequency per week) multiplied by the mode weight (heavy/light rail is weighted 2X, ferry/cable car/other are 1.5X, and bus is 1X) multiplied by a distance penalty. The distance penalty calculates the distance to the nearest stop on a route and then uses the same distance decay function as the Walk Score algorithm.

The first issue that jumps out at me is how Transit Score weights different modes to give a bonus to “better” modes: “heavy/light rail is weighted 2X, ferry/cable car/other are 1.5X, and bus is 1X.” There’s such an incredible diversity of service quality within each mode that this strikes me as an oversimplification. Is the Metro Los Angeles Orange Line really half as “good” as the Boston Green Line — a nice enough line, but it plods along 100-year-old tracks — just because the former runs on rubber tires?

There are arguments either way, and reasonable people can disagree. Those differences, however, are simply glazed over with a simple multiplier. And, I don’t even pretend to understand the 1.5X weighting for ferries and cable cars.

Early last year, when Transit Score first rolled out, I discussed it with WalkScore's Matt Lerner, expanded on this very critique, and suggested a better (though computationally intense) approach.  It involves aggregating the content of their travel time maps – effectively "maps of your freedom" — over all likely destinations from any residence, so the two digit score is actually a percentage, a composite of answers to the question "What percentage of jobs, retail, etc can you get to in __ minutes, on transit, from here?"   It needs refinement, but that's the only truly factual measure of access that could be reduced to a two digit number.

I understand why Walk Score folks settled on this methodology; it's more about the limits of compultational power than any modal bias on their part.  Still, Transit Score is the only WalkScore.com product that I can't recommend.  I love everything else that WalkScore does, but Transit Score assumes that a slow streetcar is better for you than an express bus, and people need to make that decision for themselves.

### 49 Responses to what’s wrong with the “transit score”?

1. Matthew April 27, 2012 at 4:37 pm #

I agree that is silly. The grade-crossing Green Line (“B”, “C” and “E”) is slow because it stops as often as a bus and has no signal priority. The bustituted “A” line is now the 57 bus and parallels the “B” over the worst stretch of Comm Ave. The bus regularly beats the train by a minute or two except in the worst traffic. However, the frequencies on the Green Line are much better off-peak than the bus. But that’s not a fundamental characteristic of buses, just Boston’s transit backwardness.
I could see some multiplier distinction between services, but it should be more about grade separation and reliability than rubber tires-vs-steel wheels. And ferry/cable car? Ridiculous.

2. d.p. April 27, 2012 at 6:13 pm #

Methodology problems abound, and yet the results more or less conform to what every well-traveled transit user knows about American transit cities.
It’s actually good that they’re focusing on mobility within urbanized areas rather than including and averaging entire Metropolitan areas, leading to skewed results. Suburban Seattle actually has marginally better bus coverage than, say, Westchester county, but at that point your just comparing the severity of the gaps and the degree of inferiority to driving. Better to compare the transit utility where transit utility (beyond P&Rs or commuter trains) can actually mean something.
On the other hand, using only proper municipal boundaries (rather than a more esoteric measure of urban extent) — based on the population figures in their master spreadsheet, this seems to be what they’ve done for the rankings — leads to some arbitrariness. Perfectly square, perfectly dense San Francisco likely gets a rankings boost when compared to sprawling Chicago proper or gerrymandered Boston proper (with its inherently transit-disadvantaged outposts of Hyde Park and West Roxbury), especially since urbanized Evanston, Oak Park, Cambridge, Somerville, Brookline, and the like get left out of each city’s averaged and ranked “score.” Minneapolis and St. Louis, both halfway acceptable transit cities (by American standards), get left off the list entirely because the proper cities’ boundaries are just too small.
Nevertheless, Transit Score’s top-ten list is infinitely better than this asinine internet story that made the rounds a few months ago, causing every Boston refugee I knew in Seattle or Los Angeles to wretch: http://finance.yahoo.com/news/best-cities-live-car-free-070000558.html
I lost hours of my life explaining the gaping holes in their methodology to people. The clearly non-transit-using author seems to have measured the number of buses passing in any direction within a certain distance of a starting point, calculated the mean over a certain period of time, labeled it “service frequency,” and weighted it heavily in his calculations. Thus Seattle, with hundreds of disparate, poorly coordinated routes that mostly each run half-hourly, magically achieved a “service frequency” of 8.8 minutes — better than Boston’s actually-frequent trunk routes.
The author also seems to have placed undue emphasis on the percentage of “jobs reachable in 90 minutes” throughout entire Metropolitan Statistical Areas. Not only does this lead to the above-describe problem of inherently uncompetitive intra-suburban service diluting any valuable information about the ability to get to and around the actual cities, but it’s also an insanely high baseline. Who cares how man jobs you can reach in 90 minutes on the bus? Nobody wants to do that every day, especially if the distances covered are equivalent to only 30 minutes in a car! I want to know how far I can get in 40 minutes. In 25 minutes. In 15 minutes!
In Seattle, many 90-minute bus commutes are less-than-20-minute drives. And yet transit’s “success” in getting you there in “only” an hour-and-a-freakin’-half helps push it up to “5th best” in the author’s estimation. But if he’d used a 30-minute baseline, he’d have gotten a very different picture.
Transit Score isn’t perfect. But if it goes viral in a way that replaces the prior cancerous meme, it will at least provide an improved picture of American transit to those encountering such comparisons for the first time.

3. Miles Bader April 27, 2012 at 7:21 pm #

Of course it’s simplistic, but that doesn’t mean it doesn’t work well in practice.
Maybe a better thing to do would be to explicitly encode the various factors which generally are the reason rail is more popular than buses, e.g. smoother ride, more consistent service, more comfort, “more permanent” feel to the routes, etc.
But in the end, it’s very hard not to not make some simplifications, and it ends up being a judgement call whether or not those they made are “reasonable, on average” or not. You apparently think not, but it’s certainly an arguable point; there’s nothing wrong in principle with what they did, nor does do the particular factors they chose seem unreasonable. Using their method, a great bus service will still get a higher score than a crappy rail line, but “all things equal,” the rail line will score higher—and that reflects how people think of transit in real life.
I guess the question is: what would you have done in instead? The factors that generally cause people to prefer rail over buses shouldn’t be ignored—they can have a significant impact on ridership—but may be hard to express quantitatively.

4. d.p. April 27, 2012 at 7:22 pm #

…With apologies for all the typos. I was running to catch one of Seattle’s patented half-hourlies, so it was either “post it” or “lose it.”
Next time, I promise to take advantage of your lovely proofreading feature.

5. Miles Bader April 27, 2012 at 7:28 pm #

… btw, I should add that among the hard-to-quantify factors, are those which don’t reflect “service” so much as “impact on the development of the urban environment.”
Rail, especially heavy rail, can have a very different impact on the way a city develops compared to a bus system of equivalent capacity. That doesn’t get you to work faster next month, but if you want the sort of place that develops around rail lines, you’d best invest in rail lines, even when that’s more expensive in the short term…

6. Jonathan April 28, 2012 at 5:38 am #

Hey,
I think assigning trains and trams a higher multiplier over buses makes a lot of sense. Walkscore is mostly about assessing apartments to rent and homes to buy, so it matters not only what the situation is now, but how it will look in that area in two or five or ten years.
Jared can probably make a strong statement about how this actually works, but when a transit agency looks at cutting a route, if it cuts a bus route it saves one driver per seventy passengers, if it cuts a tram line it saves one driver per two hundred passengers, which affords a degree of protection to the tramline. Further, if demand patterns change the bus line can be moved – good for the transit agency and all the new passengers, but from the point of view of someone living along the old route, it’s actually a downside.
Yours,
JMH

7. Joseph E April 28, 2012 at 7:34 am #

The suggestion of using an access-based score is good. But then you suggest this: “”What percentage of jobs, retail, etc can you get to in __ minutes, on transit, from here?”
I don’t think a percentage of access within the metro area or city is helpful. When I lived in Long Beach, CA, I didn’t care what percentage of metro area jobs I could reach in an hour by driving, or what percent of restaurants in the city were accessible by transit. I cared about the absolute number (and quality) of destinations available in a certain time.
A small town of 10,000 people might have 95% of destinations available in a 20 minute walk, while Los Angeles only might have 25% of destinations within reach by transit within an hour. But in Los Angeles, that 25% is a huge absolute number of jobs, shops and people.
Perhaps both score could be listed. The percentage would be useful for transit planners to see if how changes to the transit system could improve access in the city. But the absolute number would be more helpful for people looking for places to live or work, which is how I use Walkscore.

8. TransitPlannerMunich April 28, 2012 at 7:50 am #

Well, I can only repeat what our expiriences are when we open a new tram connection. Most recently we opened a 4 km tram line replacing a bus line running the same way. The high quality metro-bus line was running before every 5 minutes (even on sundays every 10 minutes), the tram mostly only every 10 minutes. Still already in January, 4 weeks after opening the line, the passenger numbers increased over 50% compared to the bus line before with over 13,000 in the strongest segment. Currently it is growing more,so that the tram service will be probably changed to every 5 minutes.
So however you look at it and how strange it looks to you, people love rail and use it. And they use it more than they use bus lines, no matter how good the bus line is.

9. Jarrett at HumanTransit.org April 28, 2012 at 7:50 am #

Joseph E.  How about "Of all jobs in a 7 mile radius, what % can you get to in 30 min?"  It will still be arbitrary, and anyone wanting to dig under the score could get to a place where they could set these parameters themselves.

10. Jarrett at HumanTransit.org April 28, 2012 at 7:56 am #

TPMunich.  No disagreement with your data, though as usual I note that modal preferences are facts of psychology and culture while travel time outcomes are mostly physics and math, which is why I feel more confident in the durability of the latter.
Is there a European consensus on how much additional travel time people will accept in order to use rail rather than bus?  Interested in what you know about that.  It would need to be calibrated on cities that really do strive to make the bus and rail experiences as equal as possible in terms of comfor, legibility, and delight.

11. Jarrett at HumanTransit.org April 28, 2012 at 10:06 am #

Jonathan.  Again, rails in the street are a popular symbol of permanence but a poor indicator of permanence.  If they were, we'd still have our streetcar networks from a century ago.  The real indication of permanence is the permanence of the market.  That's why it's important to provide indicators of relative permanence within the bus network — such as the Frequent Network designation.  Jarrett

12. Eric O April 28, 2012 at 2:49 pm #

Jarrett, what would you think about replacing the mode weight with the stop profile / route type? Factor 2x for express services (heavy rail, commuter express bus), 1.5x for Rapid transit, 1.0 for local. That would help offset for the distance penalty that would otherwise be charged to Express and Rapid stop profiles.

13. Jarrett at HumanTransit.org April 28, 2012 at 6:03 pm #

Eric.  Yes, that could be a reasonable simplification, except for the difficulty of weighting frequency vs speed.  Jarrett

14. Ben Smith April 28, 2012 at 7:18 pm #

I think they are coming from the psychological perspective of rail vs. bus when determining rail’s advantage over buses. In most transit maps, rail is shown more prominently over bus routes, regardless of the speed and frequencies of the lines themselves.
It would be like comparing which cities are best to drive in. One city with a lot of highways might score higher than one with a lot of artery roads, even if in the latter traffic moves smoother because it is less congested and better planned.
That said, anyone can look at a transit map and see which cities are better covered by rapid transit. Looking at the frequency, speed, and priority measures of the routes as well would yield better results for everyday life.

15. TransitPlannerMunich April 29, 2012 at 11:18 am #

On our premium bus product, the MetroBus lines with a bus every 5-10 minutes includiung sundays, we have around 3,500 passengers per km of line.
They offer bus priotiry at the traffic lights and proof of payment (like in all Munich bus lines) with ticket machines also in the bus.
On our busiest bus line we have 40,000 passengers per day (the line is 13 km long). So
As a comparison: the famous orange line in Los Angeles has 25,000 passengers on a line with a length of 22 km. 1,100 passengers per km opposed to the 3,500 passengers per km in Munich. The orange line has a travel speed of 14 km/h – while our lines in Munich have 16-20 km/h travel speed.
Still we noticed that tram/light rail is even more successful – so when we convert a line we get 50% -100% more riders.
Certainly it is psychology, cause our trams are not so much faster than our buses. Also we have an excellent subway system. So you should normally expect rather less people riding a bus. Still we have in average (all buses in operation, not only MetroBus) around 450,000 passengers per bus and year.
But what can we do – when we notice a line converted to rail gets more passengers, then we take them. And continue to build more rail.
As I mentioned once, when we know our customers want purple buses with lights outside and a cocktail bar inside we would offer it. But strange enough they go for rail, no matter how Jarrett or I or my boss (who would be happy to spend less money) wants to view it.

16. TransitPlannerMunich April 29, 2012 at 11:32 am #

I must make a small correction to my post, travel speed on the orange line is in fact between 25 an over 30 km/h due to the relatively few stops (18 on 22 km).

17. Al Dimond April 29, 2012 at 2:04 pm #

I’ll take Transit Score over Walk Score any day. Walk Score, which tries to measure walkability while ignoring what little information there is about where and how you can walk and what the experience is like (since you rarely walk unimpeded and as-the-crow-flies, this matters). It basically just measures retail density and diversity, which is nice to know but only loosely related to walking. Transit Score has a hard task in measuring transit accessibility, but at least it knows where the buses and trains run. Furthermore those bus and train lines were designed by agencies that understand the city, so typically (though with exceptions) frequent lines will go to the most important places people need to go. So there’s that additional level of intelligence built in.
I think you dislike Transit Score simply because it prefers trains over buses. It’s not true that trains are inherently better, but when you look at the systems that actually exist it’s probably closer to accurate to assume a train is better than assuming a train is the same — even if you limit yourself to uncontroversial transportation metrics like speed and reliability. If you have speed and reliability data go ahead and use it. If all you have is information on mode, and the modes don’t exactly line up from city to city, well, assuming a rail line with a particular schedule is better than a bus line with that schedule is probably pretty reasonable.

18. Alon Levy April 30, 2012 at 12:13 am #

I don’t know whether the weights they assigned are sufficiently granular and close to actual experience, but the example you quote of a bus being as good as light rail really isn’t. The Boston Green Line gets over 200,000 boardings per weekday. It nearly ties the Red Line on ridership. The on-street segments are a crawl, and the D line serves moderately dense suburbs, but then they feed into a fast central subway. I visited Boston this Saturday, when the central segment of the Red Line was out; there was a shuttle bus making all the Red Line stops, and it took 10 minutes to get from Park Street to Charles-MGH, vs. about 2 on the subway when it runs normally.
The reason subways are so popular is that in large, dense cities, with transit-friendly geography, the streets are jammed with cars. You go on-street, and even with dedicated lanes and state of the art signal priority you get less than 20 km/h average speed. You go below grade, and you can do 30 on ancient legacy systems and 40 on new systems.

19. Sasha April 30, 2012 at 12:31 am #

I’m not even at a point where I can take a critical look at TransitScore because I still haven’t gotten over WalkScore’s flawed way of rating walkabililty.
WalkScore doesn’t incorporate the built environment into its ratings. All you have to do to see how flawed it is enter a cross-street right next to a large shopping mall in a suburban hell hole near you and it’ll tell you it’s a highly walkable area.
None of these on-the-fly indicators should be taken seriously.

20. Nathanael April 30, 2012 at 1:27 am #

“Is the Metro Los Angeles Orange Line really half as “good” as the Boston Green Line — a nice enough line, but it plods along 100-year-old tracks — just because the former runs on rubber tires?”
Yes. Have you actually tried both of them?
The “Orange Lie” needs to be light rail. It used to be light rail. The buses are a poor substitute.
Not that I think this sort of score is more than a rough indicator (you could theoretically have a really nice busway, and I know there are some truly crappy railways), but bad example.

21. Jonathan April 30, 2012 at 1:30 am #

Jarret: I disagree (sort of). Let’s say that there’s some frequent network supported by high population density, which exists on its farebox returns; supplemented by some subsidised extension to that network for social mobility goals. The transit agency hires some good, smart guy like you, and the network is as well-optimised and clear as can be. That’s all fine (and probably that best-case doesn’t actually exist anywhere, but I trust that this part can be left to you…?).
In many places, demand for housing is so high that you could simply turn off restrictions on building more densely, and the increase in population density would let you expand your frequent network. Also fine – or rather, the problem should be argued about in an urban planning forum rather than a transit forum (in many places, I think the ‘streetcar’ addition and the zoning restrictions often happen together, which is nice, but does tend to confuse the issue).
I mostly care about the marginal case – where you would need to add value (for people living there, and therefore to the land itself), in order to make it worthwhile to build more densely. Greater access – as can be provided by access to (the rest of the city’s) frequent network can add that value. But that value needs to be captured by the developer (in order to make it profitable), which means that the guy buying the flat (either to live in or to rent out) will have to pay the developer for that value – and he needs to be sure that it will remain until he sells the flat (and similarly, the guy buying from him needs to believe that value will remain, and so on and so on). This is sometimes called the ‘law of iterated expectations’, and can also be written as ‘if I can predict how I’m going to think tomorrow, I might as well save myself some time and think that way today.
If you’re saying there’s no way for a transit agency to signal permanance, then we have a problem, because there’s no way to access that ‘good equilibrium’ where everyone pays for added value – because no one will pay to add the value in the first place. It doesn’t matter what the signal is, so long as it is signalling the agency’s real intentions.
To put it another way, I don’t just have to believe that the route is permanent – I have to believe that the transit agency believes that the route is permanent.
Yours,
JMH

22. Nathanael April 30, 2012 at 1:31 am #

TransitPlannerMunich wrote:
“As I mentioned once, when we know our customers want purple buses with lights outside and a cocktail bar inside we would offer it. But strange enough they go for rail, no matter how Jarrett or I or my boss (who would be happy to spend less money) wants to view it.”
🙂 Thanks for that humorous way of describing the situation.
The citizen preference for rail is simply a fact, one Jarrett really wants to deny. We can discuss why it happens — and I do think that motion sickness (trains give a better ride) is a large part of it — but regardless of the cause, it’s just a fact.

23. Nathanael April 30, 2012 at 1:34 am #

Jarrett: “TPMunich. No disagreement with your data, though as usual I note that modal preferences are facts of psychology and culture while travel time outcomes are mostly physics and math, which is why I feel more confident in the durability of the latter.”
But the modal preference for rail over bus is:
(1) cross-cultural
and
(2) has been true since the invention of railways.
You can feel more confident in that than in travel time outcomes, and here’s why.
Travel time only matters to the extent that people care how fast they travel, which is a *cultural matter* which *actually* varies between cultures and across time periods.

24. Nathanael April 30, 2012 at 1:39 am #

Right now, the US is fairly travel-time obsessed, because most people in the US work too much under punishing conditions for insufficient wages, in a culture which demands that people be working all the time. Lack of free time == demand for fast travel.
It’s not at all clear that that will remain true.

25. Eric April 30, 2012 at 5:14 am #

Nathanael, if people value their leisure and free time, then they won’t want to waste it on travel. Which implies that there will still be a demand for fast travel.
You might think that Europe has a better work/leisure culture than the US, and thus is less obsessed with travel time – but in fact Europe has been spending lots of money on projects like HSR in recent decades, while the US has allowed its infrastructure to stagnate.

26. EN57 April 30, 2012 at 7:36 am #

Most US cities would need to lift their transit mode shares by 400-500% to approach the level of transit usage in Ottawa, which relies largely on buses. Lack of gold-plated,frequent and “complete” rail networks continues to be a great excuse for Americans who don’t want anything to do with transit.
TransitPlannerMunich – is your recommended transport strategy (for any city) to eventually replace all bus lines with trams? Would you start with the poorest performing bus routes to really give them a patronage boost? The success of tram network extensions in your city has more to do with proper land use and transport planning and putting the trams in the right places, under the right conditions.

27. Jonathan April 30, 2012 at 8:08 am #

Hey,
@TPMunich – I guess that a lot of why people like trams in Europe is that in Europe tram is equivalent to light rail, so the bus-equivalent is a dedicated busway, which will automatically get you better service than any vehicle in mixed traffic.
Another feature of tramways is that because they are fixed, the transit agency has to put in the work to make sure they get it right the first time. The flexibility of buses means you can get away with sloppy implementation (or, even if you do it right, other stakeholders may not contribute as fully, thinking ‘well, we can make some changes later’).
Yours,
JMH

28. Carter Rubin April 30, 2012 at 8:43 am #

Nathaniel of this comment: http://www.humantransit.org/2012/04/whats-wrong-with-the-transit-score.html?cid=6a00d83454714d69e2016304f6f202970d#comment-6a00d83454714d69e2016304f6f202970d
Yes, I have taken rides on both. My point is not that one is unequivocally better than the other: each has relative strengths and weaknesses. Rather, it’s an apples-to-oranges comparison that is not adequately captured by the simple weighting system used by Transit Score.
Carter

29. Jonathan Hammond April 30, 2012 at 10:05 am #

Given that WalkScore weights a tiny corner bodega with erratic hours as equivalent to a 24-hour supermarket, I am usually fairly skeptical of their metrics. While it does tell me whether or not any kind of amenities exist in a neighborhood within walking distance, it also tells me remarkably little about the quality of amenities in a neighborhood I haven’t actually visited.

30. francis April 30, 2012 at 11:47 am #

Re: bus/rail weightings: I feel it’s being used as a poor proxy for stop quality. Typically heavy rail or light rail have indoor stations or at least covered platforms, while a bus stop may be just a sign (or here in SF, some yellow paint on a street pole). Have there been any studies on bus ridership increases when stops are upgraded with shelters?

31. JJJJ April 30, 2012 at 4:07 pm #

The oddest part to me is weighting cable cars and ferries higher than buses. Theyre certainly more fun to ride, but when are they ever faster?
And yeah, an express bus can and will beat a train every day. I was recently looking at transit options between New Brunswick NJ and NYC. The megabus is cheaper….AND faster than the train!
And in Boston, as mentioned above, the 57 will beat the B train every day, and twice and sundays. The only time it doesnt is during a snowstorm when the roads arent plowed.

32. Andre Lot May 1, 2012 at 5:43 am #

First, I think frequency should be taken at least on a squared basis, as reduced frequency exponentially increases transfer expected waiting. Or at least that is the case if the transfers are not timed, which is usually not the case for systems operating high frequencies.
Second, but more importantly, having transit service available says not much about how efficient is to use it as means to supply one’s mobility needs. San Francisco is a place where frequent services doesn’t translate necessarily into efficient mobility to/from the Bay Area given its awful low average speed and bad connections with Caltrain.
The latter also calls into questions the old discussion of mobility vs. access.

33. TransitPlannerMunich May 1, 2012 at 8:39 am #

@Jonathan:
Certainly we try to build in Germany trams with own right of way. But still, I live at one of Munichs most popular and heaviest used tram lines, Tram 27, and his has in the most important section no own right of way.
You can get an own impression here:
http://www.youtube.com/watch?v=LdlwgiJ-Wu8
During construction works on that line on an important intersection the tram was replaced 1:1 by an even more frequent bus service with low floor articulated buses – the ridership dropped to less than half.
And the new tram line that replaced a bus line recently has now own right of way – but the bus was almost as fast and cause of almost no traffic jams in that part of the street network had no delays. Still the tram got now this strong passenger increase.
@EN57:
Despite 100 km of subway network and 80 km of tram/light rail (generating 350 million passengers for the subway and 100 million passengers for the tram) in a city with 1.4 million inhabitants we have an extensive and frequent bus network including bus priority and now mostly articulated buses, traditionally also proof of payment and on board ticket machines.
By bus in operation we generate in average 450,000 riders, per MetroBus on our most frequent lines 1 million per year. Still we increase our passenger numbers by expansion of the tram network compared to the bus (we have also over 180 million bus riders in addition to the above mentioned passengers on subway/tram and also around 100 million alone within the city limits on the S-Bahn commuter rail (heavy rail) system plus 150 million heavy rail commuters from outside the city.
So I think we have a nice bus system, please show me a city of around 1 to 1,5 million inhabitants in America where you reach 180 million bus passengers. Not to mention the other figures. (And yes, we have cars, not only the BMW headquarters but over 700,000 cars on 1.4 million inhabitants).
Still as a rule of the thumb a in Germany famous transit planner pope, Dieter Ludwig from Karlsruhe, usually said: everywhere where you can run an articulated bus every 10 minutes druing peak hour you can replace it by tram/light rail and generate much more passengers and farebox revenue.
Our agency in Munich covers 100% of operational costs and a big chunk of investments from the farebox. So I guess we know what we are doing. We run under private company law and would not like to risk losses (even if our owner is the city).

34. Matthew May 1, 2012 at 6:33 pm #

Boston has 111.5 million bus riders annually (2010). The city has 600 thousand people but the metro area is much larger, not sure how that compares to Munich.
The 57 bus normally beats the “B” line except when it has to stop a lot to pickup or dropoff passengers, in which case it is forced into a subservient position by cars that will not let it merge back into traffic. Then the “B” beats it. Both the bus and the train have pretty terrible operating procedures (especially by German standards), and extremely small inter-station gaps, which slows them down immensely.
If the “B” had fewer stations, improved boarding, and signal priority, then it would beat the bus.

35. Chris Standen May 1, 2012 at 9:02 pm #

On a related note, Transportation Nation has published this BRT Rating Standard http://transportationnation.org/2012/05/01/brt-systems-getting-an-international-rating-standard/

36. M1EK May 2, 2012 at 8:20 am #

Yes, this “permanence” rebuttal of Jarrett’s is highly disingenuous. Streetcars were torn out, it’s true, but it took a looong time – and none of them were effectively ‘moved’.
And all that happened basically one time in the last 100 years.
In the meantime, most cities have seen buses shift far more times. Even trunk routes make at least minor shifts even if the general corridor stays mostly the same.

37. TransitPlannerMunich May 2, 2012 at 10:22 am #

Today I just read an article in a German publication that in Zurich they expirienced that changing a normal bus line to a trolley bus will bring 10% more passengers to the line…
(Tram even more).
And that just by a change from diesel to wire.

38. Wad May 2, 2012 at 7:17 pm #

@Nathanael wrote:
The “Orange Lie” needs to be light rail. It used to be light rail.
The Orange Line won’t be light rail. It should not be converted.
The carpenter’s adage of “measure twice, cut once” also applies for major capital investments. There are no do-overs.
If we peg Orange Line at \$400 million for the current busway, which has yielded 27,000 boardings (current ridership), we must remember that a light rail equivalent would have cost \$1.2 billion. Rail might be worth it if you could plausibly get 81,000 riders — and the Blue Line is the only modern light rail system in the U.S. to attain that figure.
The problem now, though, is the marginal cost. A do-over would cost \$1.6 billion — remember, the \$400 million busway is a sunk cost. You would now need to get the ridership that the L.A. subway gets in order to justify rail. There’s no way rail would draw 81,000 along the right of way; and no corridor in L.A. but the the subway extension will add 108,000 boardings.

39. Nathanael May 3, 2012 at 12:11 am #

@EN57: Ottawa’s building a subway.
@Wad: The Orange Line’s roadway is already deteriorating badly. It already needs to be rebuilt. Consider the \$400 million *truly* sunk, in that the ‘investment’ has already crumbled into broken asphalt, and decide what ought to be built there next….
I suspect you’re also wrong about the cost of light rail on the line. There was some really strange stuff going on at the time, because there were local laws *prohibiting* surface light rail on the eastern end of the line, which inflated the costs for any potential light rail system *massively*. I think those laws are still on the books; if they were repealed, one might be able to take a reasonable look at the costs.

40. Eric O May 3, 2012 at 6:01 am #

If you are looking only at certain isolating metrics of the routes with stops in proximity, then Andre Lot’s suggestion that frequency be exponentially factored is a great one. However, I would apply a kind of “mode weight” still to reflect capacity and speed of the service. Typically, people are willing to walk further to reach higher capacity modes or express services, for example. Modes that “compete with walking” should be weighed less.
Andre, to your second point, I think it is important to note that Jarrett’s suggested alternative gets exactly to the root of measuring the efficiency of the system. Basically, what percentage of the destinations in the service area that you can reach in X minutes is a specific measure of efficiency. Basically, Jarrett’s important datum is the 30 minute isochrone, and comparing what’s in it to what is outside of it. How well your connections and routes integrated geographically to get “where I want to go, when I want to go” is a measure of how hard your transit system is working for you.
The WalkScore folks already know how to create isochrone maps, so Jarrett’s suggestion is not an unreasonable analysis for them. Unfortunately, they are getting carried away by their own mode fetishism to clearly see the value of the tool they created.

41. TransitPlannerMunich May 3, 2012 at 1:11 pm #

About the Orange Line:
It is about 23 km long. So according to German standard calculations it wold cost around 300 million dollars (10 million euros per kilometer) to build a light rail including stops, landscaping, electrical equipment, power supply stations, wires, curbstone to curbstone design, signal priority etc. – plus around 3.5 million dollars per light rail vehicle. Those costs are valid for Germany, even if Germany is a high wage country.
But the rails will last for 30 years, the vehicles for 35 years.
So we build trams/light rail for reasons of economy to save on maintenance, save on vehicles and get more riders. Busways would be too expensive in the long run.
The result is, you get something like this (there is high density housing and offices close by):
http://www.flickr.com/photos/frederikb/7136497405/in/photostream/
or this:
http://www.flickr.com/photos/frederikb/6990407092/in/photostream/
at least this:
http://www.flickr.com/photos/frederikb/6970198080/in/photostream
but not – for the same price or more strange enough, something like this:
http://en.wikipedia.org/wiki/File:LACMTA_Orange_Line_Balboa_Blvd.JPG
And yes, we also have BRT,
http://www.flickr.com/photos/frederikb/6919731894/in/photostream/
but seriously, the tram examples above looks nicer, or?

42. Wad May 3, 2012 at 4:20 pm #

@TransitPlannerMunich,
The U.S. is incapable of producing any kind of project at that cost. The Orange Line, as is now, cost \$400 million.
The U.S. is incapable of controlling costs on public projects, and L.A. is acutely bad in the U.S. at controlling said costs.
We just completed a light rail line at \$950 million that should have been \$700 million. Our subway extension will cost \$6 billion for about 10 miles — this extension will cost more than the existing subway (17 miles), the very same one that was banned with a minimum of 2/3 support countywide.

43. John Bailo May 4, 2012 at 12:44 pm #

Here’s what I don’t understand…why is it “walkable” to go half a block and get on a bus?
Seems like that’s not much walking at all.
People in rural communities walk all the time…but we walk half miles and miles like drinking a glass of water.

44. d.p. May 4, 2012 at 3:59 pm #

You know that’s not true, John. Most of Kent, WA doesn’t even have sidewalks, and nearly all of its car-owning residents would use them just to cross the street.
Maybe you’re the exception, but I suspect you’re just as dehydrated as your neighbors.

45. Nathanael May 4, 2012 at 6:36 pm #

“Basically, Jarrett’s important datum is the 30 minute isochrone, and comparing what’s in it to what is outside of it. ”
OK. Let’s do that. But let’s consider *reliability*.
I remember an interesting instance in Boston where a group of us tried to go across town from the Hynes area to Porter. We walked to the bus stop for the supposedly fast, frequent direct bus.
After a couple of minutes I said “I’m taking the train”. Several of us walked five minutes to the train station, rode the Green Line to Park Street, changed trains to the Red Line, rode from there to Charles/MGH, — where the Red Line bridge was CLOSED FOR CONSTRUCTION — got on a shuttle bus (they had something like 30 of them, running continuously) to Kendall/MIT, and continued on the Red Line to Porter; then walked two blocks to where the bus stop was. We then went to the cafe we were heading towards.
We got there a FULL HOUR EARLIER than the people who waited for the direct bus. For purposes of “30-minute walkshed”, the #1 bus *does not exist* in Boston due to unpredictable unreliability.
This shows not only the issues of mixed traffic running, but also *attitude* on the part of American transit agencies. They simply don’t *care* if the bus routes are disrupted, but if the rail lines are disrupted, they will make serious efforts to deal with it or at least to notify people of it.
You have to figure this sort of stuff in to any attempt to calculate an actual transit score. The rough-and-ready “is it rail?” analysis is a good approximation in most cases.
Wad: you have a point about costs. But in counterpoint, we know the specific reasons for most of the price increases on the Expo Line (several of which involve community lawsuits and some of which involve contractor incompetence). As for the Subway to the Sea, it has some of the ugliest geology to tunnel through ever (tar!).
You’ll also note that the Orange Line, as is, was ALSO cost-inflated; “all costs are inflated” is not an argument to make the less effective project, it’s an argument to figure out where the money is going.

46. Wad May 5, 2012 at 11:36 pm #

@Nathanael, there are some factors that go above and beyond local ability to control costs.
For one thing, transit projects in particular are dependent on a cartel of construction firms to design and build these systems. Each addition of an MIS-level transit project has grown progressively more expensive with each addition. This is counter to most productive economic activity, where each increment is supposed to reduce unit costs rather than increase them. In other words, it’s Moore’s Law in reverse.
It doesn’t have to be this way. You and I and most people reading this can come up with simple ways to subvert this system. It will never happen short of a power vacuum.
Why? Our present decision-making system (in the U.S.) perpetuates a consensus of failure. We have two political parties: One positions itself as the honest broker juggling hostile constituencies. The other will tolerate failure because it serves as a racquet (an organizing force that constituencies can fixate their hostilies toward but never really undo).
There’s a self-interested consensus among both camps to tolerate high costs and inefficiencies, because to actually eliminate them eliminates their need to exist. The first camp needs voters to think they need the honest broker; the second camp needs voters to think fraud, abuse and waste are inevitable and they are the only ones who will stop it and hang the miscreants to the nail of their deeds.
Reforming this, though, is the Gordian knot because it involves rewiring hard-wired and incompatible values systems. See George Lakoff’s theories on cognitive politics for one explanation of this phenomenon.

47. Ben Pease May 6, 2012 at 7:26 am #

I agree that rating cable cars higher than busses is odd. Here in San Francisco the \$5 fare, and the long line at the terminals, sorta divorce them from the regular transit network – in high season they are so full of tourists there is no room for more passengers at intermediate stops. Early in the morning, with a Fast Pass, they are much preferable.
Re JJJ’s comment… Back in the day I rode cable cars to school for \$0.50. If I missed the California car I could try my luck with the 55 Sacramento bus. If it wasn’t too full, the old GM diesel busses were a little faster than the cable car (and much faster than no cable car). But the last block of Nob Hill is so steep, if the bus was full, the driver would order the passengers to get out at Powell Street and walk up the steep hill to reboard the bus at Taylor. Less of a problem now with the 1 California trolley bus.

48. cph May 6, 2012 at 10:34 am #

Sure, I could go for giving a subway or light rail, with its own right-of-way, more “weight” than a local bus. Don’t know about mixed-traffic streetcars–they’re about the same as a mixed traffic bus in terms of performance. (The systems in San Francisco, Boston, and perhaps Philly get a pass because they feed into a subway–sort of a “pre-metro” system.)
Commuter rail is tricky though. If they have frequencies like LIRR/Metro North/NJT, certain Metra lines, or Caltrain, then they should get some weight. But most of them just run a few times per day. Same with ferries or “Commuter Express” buses….
And the difference between a true BRT (LA’s Orange Line), a limited-stop bus “Rapid Bus” and a regular local bus should be noted….

49. Andrew May 9, 2012 at 7:36 pm #

I do not care whether I use bus or rail on my trip, so that should not affect the transit score. The main thing I care about is the travel times, including waiting time, to a wide variety of destinations in the Greater Toronto Area from the area where I live. Furthermore, is transit competitive with, or faster than driving in heavy rush hour Toronto traffic? Obviously I would tend to prefer commuter rail/subway over light rail or bus because the former tends to be faster, but I do not have any preference between a commuter rail line and an equally fast express bus (assuming no crowding issues). The obvious way of calculating a transit score is to use Google Transit to generate travel times from a given location to various random destinations including wait time in rush hour and compare to drive time “with traffic”, and do the same for off peak times and off peak traffic. By this metric Yonge & Eglinton is very high since I can get to a large chunk of the GTA with one transfer, Don Mills & Eglinton is fair, and 16th Avenue & Highway 404 is poor.