Finance Salt Lake Social Justice

A Call for Minimum Service Standards

» All across the country, transit agencies are opening new rail lines with inadequate service.

At $37 million for two miles of track, Salt Lake City’s new S-Line, sometimes referred to as the Sugar House Streetcar, was one of the cheapest rail transit projects recently completed in the United States, with per-mile costs equivalent to the typical bus rapid transit project. From a capital cost perspective, it’s a great success.

Too bad the S-Line is such a dud when it comes to ridership. According to recent data from the local transit system, the project is serving fewer than 1,000 riders a day, far fewer than the 3,000 expected for the project. One explanation is that the short route doesn’t attract many people. Another is that the line’s frequency is simply too low to convince people to orient their lives around it.

The thing is, providing new rail lines isn’t enough — service standards really matter when it comes to attracting people to use transit. And on that front, too many transit agencies around the country are failing to offer the services people can rely on. The problem extends far beyond New Orleans and encompasses a large share of the cities that are investing in new rail lines today, ultimately limiting their effectiveness and cutting down on ridership.

We must commit our transit agencies to providing a minimum level of transit service on their lines, particularly those in which it has been deemed necessary to invest millions of dollars in capital upgrades.

Certainly part of the answer should be speeding transit up. In an urban environment where automobiles dominate, making sure that buses and trains can move as quickly as possible reduces commute times and, ultimately, reduces the appeal of driving by providing a time-competitive alternative. At an average of 10 mph, the S-Line is certainly no stunner.

But the streetcar’s bigger problem is that trains only make the 2-mile, 12-minute trip every 20 minutes, or 3 times an hour.* If you miss a trip, you might as well walk, because you’ll save virtually no time waiting for the train. As Jarrett Walker has noted many times, frequency of service can be just as important as speed, since the frequency at which a vehicle on a line arrives determines how long most people have to wait — especially when they’re transferring between services, an essential element of any big-city transit network and one that cannot be significantly improved with real-time data.

An examination of the operations of 49 new or extended light rail or streetcar lines built in the U.S. after 2000, summarized in the table at the end of this article, suggests that the situation experienced in Salt Lake is hardly unique, particularly at off-peak hours. While the large majority of these services offer at least four trains per hour (one vehicle every 15 minutes in each direction) at peak hours, 35% offer fewer than 4 trains per hour at midday and 73% offer fewer than 4 trains per hour in the evening (indeed, 33% offer 2 or fewer trains per hour, or a train every half hour, in the evening).

The difference for a passenger using a transit service offering robust frequencies — 6 trains per hour, or one train every ten minutes — versus mediocre ones — 3 trains per hour, like the S-Line — can be dramatic. A hypothetical rider in the robust city who has to take two 15-minute train trips that involve one transfer between them will spend an average of 40 minutes commuting in each direction (30 minutes on both trains and 5 minutes waiting for each individual train). In the mediocre city, on the other hand, average waiting times of 10 minutes for each train would increase commute times to 50 minutes, or a 25 percent increase. In the worst circumstances, where a rider just misses each train, the rider in the robust city would require 50 minutes to commute while her peer in the mediocre one would need 1h10, a full 20 minutes more.

To create a transit system that is attractive enough to pull people out of their cars, high frequencies of service at all times of the day are essential.

Los Angeles, Minneapolis, and Seattle, as the table demonstrates, have chosen to outfit their new rail lines with a robust level of service that befits the major investment that has been put into them and recognizes the time constraints of their riders. Others, from San Jose to Sacramento to San Diego and even to Portland, have simply chosen to give up on their passengers at night. What they’ve effectively decided is that only people who truly have no other choice should rely on transit outside of peak hours.

The poor service offered on these lines produces infrastructure that is massively underused. One of the frequent arguments made by proponents of investment in rail is that “people know” that the trains will come because of the fixed track and supposedly high quality of service. But inadequate operations make this benefit disappear.

The federal government, which has funded the majority of these projects, has failed to enforce any sort of minimum level of service that these lines must provide. Rather than mandate that new services funded through grants offer service at least every 15 minutes, for example, the Federal Transit Administration simply requires agencies to “develop quantitative standards for all fixed route modes of operation” for issues like vehicle headway. In other words, if a transit agency provides service every three hours on a just-built rail line, that’s fine — as long as that information has been submitted in triplicate to Washington in advance.

The federal government is throwing money at these projects with little supervision over how they are operated. The results are underperformance and relatively low ridership.

Certainly many transit agencies will suggest that the reason they do not offer better service is that they cannot afford to do so; as I wrote last week, the way that transit funding is allocated results in perverse incentives that encourage transit expansion over transit service. But local governments that commit to new projects should be required to identify adequate funding to cover operations if they are awarded federal money for construction.

Other agencies might argue that the service they provide simply matches the demand; there is no need to offer more than three trains an hour on the S-Line because only 1,000 people a day will even ride the thing. This fact raises questions about whether it made sense to build the project in the first place — if it’s not serving many people, is it needed? Or it represents a self-fulfilling prophecy: Of course the line serves few people because the service it provides is so poor.

Choosing to invest in better services comes at a cost. But it’s one that our political leaders and local transit activists should be fighting for. Now that we have some rail lines constructed, let’s start running more trains on them before we rush out to build more track!

Service levels for new or expanded light rail or streetcar lines in the U.S. since 2000
CityExtension Project or New LineYearPeak trains per hour (8-9 AM)Midday trains per hour (12-1 PM)Evening trains per hour (9-10 PM)
CharlotteBlue Line2007643
DallasStreetcar Extension2013442
DallasBlue to Rowlett2012433
DallasRed to Plano2002433
HoustonRed Line200410103
Jersey CityTonelle Ave20061074
Jersey City8th St2011532
Little RockStreetcar2004223
Los AngelesGold to Pasadena2003956
Los AngelesExpo 12012556
MemphisMadison Line2004440
MinneapolisBlue Line2004664
MinneapolisGreen Line2014666
New OrleansCanal2004334
New OrleansLoyola-UPT2013332
NewarkLRT Penn to Broad2006422
NorfolkLight Rail2011644
PhoenixLight Rail2008553
PittsburghBlue Line2004222
PittsburghNorth Shore20121586
PortlandRed to Airport2001443
PortlandInitial Streetcar2001353
SacramentoSouth to Meadow View2003442
SacramentoGreen to Richards2012220
Salt LakeRed Line2003444
Salt LakeBlue to Draper2013444
Salt LakeGreen/Airport2013444
Salt LakeStreetcar2013331
San DiegoGreen: Mission to Santee2005442
San FranciscoT Line2007663
San Joseto Alum Rock2004442
San JoseSacramento to Winchester2005422
SeattleCentral Link2009766
St LouisCross County to I-442006533
St Louisto Scott AFB2003433

* This is the level of service provided all day. In its submission to the federal government in 2010, Salt Lake claimed it would provide service every 15 minutes at peak and every 30 minutes in off-peak periods.

Image at top: S-Line Streetcar in Salt Lake City, from Flickr user Paul Kimo McGregor (cc)

60 replies on “A Call for Minimum Service Standards”

This is really good. Minimum standards of service should absolutely be included in federal funding requirements. That being said, I see one potential drawback: in many places, new rail systems have been accused not just of having inadequate operations funding for themselves, but of drawing away operations funding from the rest of the system (primarily buses) as well. Obviously in some cases those concerns have been overblown (cough cough LA Bus Riders’ Union), but it does seem to be a serious issue in some cities (Portland for sure, I’ve seen reference in SLC). Should we enforce an even higher standard–that cities/agencies opening new transit have to identify and guarantee funding not just for high-quality operations for that line, but to keep non-redundant existing services at their current levels as well?

It’s worth noting that the S-Line headways are constrained by its construction methods, not just the UTA. Phase I was largely built with single track sections, so trains can only pass one another at the few stations located along the route. While headways could be slightly shorter, it’s unlikely that they could be improved by much without incurring new capital costs. The project was also built along an abandoned freight right of way, which had the advantage of cutting costs and avoiding mixed traffic operation, but also meant that the initial segment runs along a right of way that has little, if any, commercial or residential frontage. The S-Line, along with the bike/ped greenway also built in the right of way, was intended to bring development to the area. And Phase II will take the line into the main business district in Sugar House, where ridership demand will be much greater.

I absolutely agree that the UTA should do everything it can to reduce headways, at least to the 15 minutes that were originally promised. But we shouldn’t judge a new line’s ridership when it is blazing a new trail, especially when much of the promised development along the streetcar has taken longer than expected.

This is a great post. It points up importance of locating new rail lines in existing activity corridors rather than as stimulus to greenfield development. This is especially true in cities with skimpy rail networks, as the rail line will have limited use in attracting development if it won’t take the occupants of that development where they want to go.

A useful addition to the table would be the length of the line (where applicable, the whole line, not just the extension). Longer headways aren’t so bad if you’re talking about a long trip. For example, the route and land use of the Camden River Line are more like a commuter railroad, so this route suffers much less from long headways than the Sugar House Streetcar or Neward Broad Street LRT.

Tampa’s streetcar is a pure example of Tourist Transit… the idea that politicians went to Disney World, rode the monorail and then came home and said “Gee, we should build something like that here.” Other examples of Tourist Transit include Kenosha’s streetcar, the Seattle Monorail and the Las Vegas Monorail.

So therefore, Tampa went ahead and built a streetcar line that is only about 2 miles long and serves a bunch of tourist-only places like Ybor City that are only busy in the afternoon and on weekends. Lo and behold, there is no reason to run it in the AM peak… who gets up to go to the Florida Aquarium or to the bars at 8am? Check out the website here… you can barely find the map/schedule/fare info buried in it

In summary: Only 13 out of 49 offer service every 15 minutes of better all-day… and only six (!) offer service every 12 minutes or better.

The US always seems happy to spend capital money, but never operational…

Maybe this is a reason to outsource operations, so there’s a capitalist able to get his finger in the pie and thus willing to lobby for more operational dollars.

Right now, the US Government is terrified that transit workers might make a decent wage. :)

In Dallas, the McKinney Ave trolley operations are severely limited by a long stretch of dead-end track. That should pick up significantly this year when they finish a loop extension. Still, it is primarily a historic/sightseer attraction, not a means of regular transportation. The DART light rail lines are constrained , at least at peak times, by having to filter through downtown without grade separation. From what I understand, that was an existential decision. Building the system never would have moved forward if it had included the cost to elevate or bury the downtown service. I think minimum standards are great, but I wouldn’t want them to kill off a nascent project that needs time to grow up, so to speak.

More significant than the grade crossings, DART light rail is constrained by having all 4 lines run through the same center city tracks, so the maximum frequency on each line is only 1/4 of what it could be. If a bypass route is built for 2 of 4 lines downtown, the frequency can be doubled on each line.

Signal priority for light rail downtown is also important – it’s not clear to me if this exists yet.

I think it’s pretty simple why this is happening.

1. Most Americans don’t take transit. Maybe they took it on vacation a couple times and thought it was pretty cool, but on a day-to-day basis they drive to work and drive to run errands.

2. Most Americans are intrigued by the idea of building transit, but not necessarily ’cause they imagine themselves suddenly taking it to work or to run errands. Maybe they could take the trolley the next time they’re on Beale Street or to get to the Mexican border when they visit San Diego or the next time they’re in NYC. But they can’t really imagine not just driving everywhere in their regular daily lives.

3. The idea that service frequency is so important is not intuitive to someone who doesn’t take transit regularly in their daily lives. Instead, people who don’t take the bus are very concerned that their tax dollars are paying for “empty buses” they see lumbering along the road at 9pm on a Sunday as they pull out of the parking lot. They aren’t ever going to take that bus. But they might take that new little trolley that will run in a loop downtown every 40 minutes from 10 to 4. Or at least they think they might.

We’re paying for things that the non-transit riding majority of the public imagine themselves to one day ride. We don’t care about the needs of riders. That’s the perverse version of democracy we have.

The Chicago Transit Authority is planning an extension of the Red Line – their system’s busiest route – from 95th Street to 130th Street, adding four stations. There is a yard at 98th Street (and two other yards much further north).

The CTA could easily opt to run regular service on the existing route (between Howard and 95th Street) and limited service on the extension, from 95th Street to 130th Street. The yard and crossover tracks make short-turning trains a possibility.

Likely because of our fractured regional transportation planning environment in Chicagoland, using existing infrastructure to serve the residents south and southeast of the busy 95th Street terminal has not been reviewed. Instead, the CTA, Metra, and their regional oversight authority, the RTA, acted in silos. The CTA reviewed the corridor by itself, looking at different modes that could serve the area, and the possible alignments. To my knowledge they didn’t review a non-CTA solution. But that’s expected in our setup; CTA is in charge of CTA and Metra is in charge of Metra. The RTA is, on paper, in charge of both, but hasn’t taken a leadership role in this project that will cost well over $1 billion.

The existing infrastructure is Metra Electric, the most on-time route in the commuter rail system. The line already has stations on each of the streets (except 119th) where CTA proposes new metro stations. The line runs within blocks of Altgeld Gardens, the detached neighborhood at the terminus of the extension.

Will CTA commit to the right levels of service for the extension?

Presumably, CTA and especially Pace would relocate a lot of the buses that terminate at 95th St now, especially the ones that duplicate other routes just to get there from places a lot further south. Pace has a really big incentive to do so, since they waste a lot of time/miles in the city limits just getting to a terminal when all their suburban customer base wants is to get to a train. But with a longer Red Line, even CTA would be smart to straighten out routes like #111, #112, #115 and #119 to operate purely east-west and hit the new Red Line stations instead of all crowding onto the same streets (Michigan, 95th) to access the current Red Line terminus.

If all bus routes except #29, #95E, #95W, #100 and maybe #34 left the 95th terminal for newer stations fewer south… it would drive higher ridership and thus justify lower headways at the new 130th St terminus. Given all of that, I think CTA would be able to easily justify at least 4 trains an hour during midday and at night south of 95th St.

So why is this type of policy/approach only a good idea for fixed route transit? Walker’s notion of the appeal of the frequent network applies to all transit, not just fixed route services.

The notion of the frequent network does apply to all transit, however, non-fixed-route transit (= demand responsive transit) usually doesn’t have any significant FIXED capital costs. You simply buy a bunch of demand-response vehicles and then start operating… and if you or the Feds later determine that the vehicles are being underutilized, you can sell them off at anytime and recapture some or almost all of the initial grant money that the Feds gave you.

The whole point of this particular article though is that fixed route transit is different. For rail, it requires fixed capital costs – you have to lay down track and build stations that are essentially forever locked into their current placement. Even for bus, it requires bus shelters and stops to be put into the ground. (You can repurpose or liquidate parts of them maybe, but only after spending even more money to modify or rip them out of the ground… and there are a lot fewer buyers around for used fixed transit equipment.) So if it’s determined that these facilities are being underutilized, there’s very little their owners or the Feds can do about it… it turns into a pure waste of those very limited capital grant dollars from Washington.

MAP-21 did include some allusion to service standards in the definition(s) of Bus Rapid Transit. It doesn’t give exact numbers, but it does say it should have: “short headway bidirectional services for a substantial part of weekdays and weekend days;”

Sounds vague. Could a transit agency be brought to task through some sort of lawsuit for violating that requirement when using MAP-21 federal money?

Since that requirement applies to FTA New Starts/Small Starts funded projects, what exact requirements you are held to gets defined in your agreement with the FTA for funding the project.

As a committed rail person, several years ago I had the opportunity to take the Baltimore Light Rail from Penn Station down to the Inner Harbor. Not even counting the wait time at Penn (the trains operate on 30 minute headway) the slow creep down Howard Street (where the train even enjoys its own lanes) took much longer than I could have made it on foot.

Baltimore Light Rail problems run deeper than one too many stops on Howard Street. Alignment over the northern segment of the Blue Line Light Rail is so far away from major traffic generators that patronage is too low to run more frequent service in its North-South Corridor. Baltimore Metro Area has also underinvested in Light Rail & Metro expansion in other corridors, while overbuilding suburban freeways.

The best thing Baltimore Metro Transportation officials can do now is unify and lobby the FTA to fund:

1. Building Red Line in the East West corridor asap.
2. Building the Yellow Line as 22mph (grade separations, tunnel) high frequency Light Rail in the highest traffic corridor between Lutherville-Towson-Towson State/Med Ctr-Loyola U-John Hopkins U-Penn Station-Charles Center-Inner Harbor-Camden Yards
3. Extend Green Line from Johns Hopkins Med Ctr-33rd Street-Morgan State-Overlea.

What about automated guideways? Why aren’t we building those? Will the Federal Transit Administration pay for those? Those seem like they could provide more frequent service at less than the cost of any of these LRT and streetcar systems.

At least as of now it requires grade separation everywhere, and that costs money, and for older systems (like the CTA) it requires a lot of retrofitting of existing infrastructure, which also costs a lot of money, enough that automation doesn’t often present any significant advantage over one-person train operation. Plus a lot of American systems aren’t purely urban systems but somewhere in between urban and regional rail, so they don’t necessarily need high frequencies less than ten or fifteen minutes on their most distant branches.

I think the S-Tog in Copenhagen is set up to accommodate future automation and it’s a not-grade-separated-everywhere regional rail system—unfortunately rail technology, at least in the states, moves slowly, but I wouldn’t be surprised if we start to see automation on systems with grade crossings (but their own ROW, i. e. not-street-running) outside the US in the next ten or fifteen years.

The Canadian made SkyTrain is automated. It is computer controlled. There are no drivers who regularly driver the train, but there are on-train personnel who can operate it in case of an emergency. It is, obviously, totally grade separated even at ground level where it must be protected from intruders, animals and the like. However, it hasn’t prevented suicides.

Trains generally do not have personnel on board. They are often at stations for fare checks and security reasons. The benefit of the automated trains is the cheap and quick ability to increase headways to respond to demand. In Vancouver, for instance, during am/pm peaks, trains run on 90 second headways. The trains are very short but are offset by the very high frequencies. Even in off-peak hours (9pm-1am), headways are generally no more than 5 minutes.

What you say is true. My complaint is the extremely bad bus-feeder headways — bad enough to keep me from using SkyTrain. I am also waiting to see what the headways will be like when the new Evergreen line opens in two years. Another common complaint is that the SkyTrain shuts long before some bars do in Vancouver. If New York City can operate 24 hours on its subway system, there is no logical reason that the SkyTrain can’t run until at least 2 a.m. This is not a personal issue but a safety issue.

Wow, these abysmal service frequencies reveal that many streetcar projects are more about real estate development than about public transit per se. I suppose one could argue this is a revival of 19th century real estate practices, in which streetcars were built to access the new “TODs” of that era.

(The differences being: [1] the “TODs” of that era were actually transit-oriented due to necessity, whereas many of today’s infrequent toy trolleys do little, if anything, to reduce one’s need for a car, and [2] the real estate companies of the 19th century actually had to finance, build, and run their streetcars themselves. Of course, many were vertically integrated (real estate development, transportation to said development, and electric generation for both all under one roof) and eventually turned their routes over to public agencies.)

But if we’re gonna revive 19th century real estate development, why saddle public transit agencies with limited resources – the feds that do the reviews and grants, the states that sometimes chip in, and the local/regional transit agencies that ultimately have to operate all these showpieces – with the duty? Why not give developers the regulatory freedom* to finance, build, and operate their own streetcars again if they’re so convinced they have to have them to spur development? If the lines prove useful, then they can turn them over to public operation if they want, or continue running them themselves. If the lines prove duds, then developers can just discontinue them with ease after the development has already taken place.

*This might be a tough sell in the built-up east, but it might be more palatable in southern and western Sunbelt metropolises with ample, empty, developable land both within and around cities. Most of the streetcar projects are already concentrated in the Sunbelt anyway.

The service frequencies on some of these lines are so poor that I wonder if adjacent developers are more interested in the *idea* or *image* of transit than in actual useful transit. Some lines offer such spartan service – yet still attract development – that we might get an absurd scenario in which simply adding decorative tracks to the street, but offering no service over them whatsover (except perhaps the occasional tourist shuttle), might be enough to spur development.

In this case steel tracks will be just another development-spurring streetscape improvement, analogous to the five B’s from the 1980s (bollards, brick sidewalks, benches, banners, berms). And eventually too, their development-spurring ability will fade once they’re overapplied. But public agencies shouldn’t have to pay for, plan, build, and operate streetscape improvements that are essentially irrelevant to public transit.

“Some lines offer such spartan service – yet still attract development – that we might get an absurd scenario in which simply adding decorative tracks to the street, but offering no service over them whatsover (except perhaps the occasional tourist shuttle), might be enough to spur development.”

This is essentially what is happening here in DC along the H Street corridor. The tracks have been there for years… but no actual service has ever operated and City Council just cut funding for it… and yet, redevelopment is spreading like wildfire throughout the area. Read the linked article below:

The Streetcar-Minus-Streetcar Plan Worked for D.C.

Is Utah Transit Authority Counting one Person like they 4 or 5 People in Order to Justify getting more Federal Taxpayers Dollars to Pay for Operating Cost for the Streetcar.

Here in Vancouver BC on our light rail system we have trains every 2 minutes or less in peak hours. Seldom more than every 5 minutes in off peak hours. Otherwise what is the point.

A train that only goes 2 miles and runs at 10 MPH is a total waste of money. One city bus running the same 2 miles can do way better than that and for way less money.

What were they thinking?

It appears DMF and I agree that Vancouver has a successful system.

I understand the Skytrain has >100% farebox recovery of op costs – something on the order of 2 times the best US / LRT rate. And I believe our capital cost compares well. This supports Freemark’s point about frequency and effective speed.

Obviously Americans are too stubborn and proud to learn from Vancouver. Believe me, even living in the U.S. I’m dead sure that Canada does EVERYTHING better than the U.S.

Translink doesn’t publish separate cost recovery figures for SkyTrain, since the system’s fare-integrated with the buses. SkyTrain does publish figures for cost per boarding on each bus route, and (in harder-to-retrieve places) on SkyTrain, and for both SkyTrain and the busiest bus routes, it’s a bit more $1 per unlinked trip. So if you’re transferring from a trunk route to SkyTrain the system probably breaks even on you, after accounting for both monthly discounts and the zone system. But then if you’re using less frequent buses to get to the train, you’re being subsidized heavily.

Anyway, Vancouver is doing a lot of things most American cities don’t: upzoning in downtown and around SkyTrain is the big one, but also timed transfers between the buses and SeaBus, a relatively coherent grid in Vancouver proper, and routing buses to serve secondary centers like Metrotown and not just downtown.

I can confirm DMF’s observations of Vancouver’s SkyTrain. Standing on the eastbound platform of the Burrard Station at 3:30 p.m. on an October work day last year purposely I timed the frequency at four 6-car trains every five minutes, which of course works out to one train every 75 seconds. I went to the westbound platform below and timed it almost exactly the same. That’s a dozen trains entering and leaving the station complex every five minutes at rush hour.

I don’t know the average speed on the Expo Line (Alon, do you have that figure at your fingertips?), but do know that the top speed between stations is 80 kph.

To me all that’s bloody remarkable.

The Canada Line is a little different. While it uses the same driverless technology and is therein grade-separated (both elevated and underground), it was built using a private partner who specified a different rail company (Hyundai’s Rotem instead of Bombardier, which makes SkyTrain) and downgraded the capital cost and in the view of many shortchanged the project design. Thus, we have 42 m platforms on that line instead of 80+ m, and the private operator never puts all 20 trains on at rush hour and thus saves money. The Rotem is a wider car that carries more people than the narrower SkyTrain car, but the frequencies are a lot slower and therefore the crowds pile up on key hub station platforms at crush hour and the trains are packed like sardines.

To some that’s a measure of success, especially compared to the anemic frequencies and ridership Yonah iterates above. To other like me, we can do a lot better in future.

Thanks for this article.

The absolute minimum frequency for urban transit should be 4 tph (trains per hour), which is 15 minute service, with the exception of night-owl hours (10PM – 8AM at most).

15 minute service is the bare minimum level for not consulting a timetable.

Minor quibble: AM rush hour starts at 7am in most cities, 6am or even 5am in some. Ending your night owl service headways at 8am wouldn’t be early enough to get anyone to a normal-business-hours job.

I understood that one problem is that Federal Funding is available for initial construction, but not for subsequent operations and maintenance. This is an incentive to cut back on frequencies.
And I recently rode from Chicago to Milwaukee Airport to have a look at the connection there. The train, of course, continued on to downtown Milwaukee, waited a bit and came back again. I caught the same train back – and was surprised to see how many of the people I’d seen waiting on the platform when I got off the northbound train were waiting for that southbound service. They didn’t seem too bothered by frequencies! But yes, I accept that this was a bit longer than the average transit trip.

That’s a 90 minute one-way trip there. Waiting more than 15 minutes to go on a 90 minute trip is nothing… witness all the well-to-do people voluntarily showing up at airports 60-120 minutes in advance of a 90-minute trip. What this article is speaking about is that nobody wants to wait more than 15 minutes to make a 10-30 min trip across town.

Low frequency is what keeps me from using transit where I live — 60 min (1 per hour). I can drive to my destination and back in less than 60 m. They are extending a SkyTrain out through suburb I live in. But they aren’t providing any extra parking. The current lots are used up by commuter train passengers. Waiting 55 minutes for a bus is totally unacceptable. I will continue driving. Oh yes, I am a fan of public transit (preferably electric), but not if it is bloody inconvenient.

K Macgowan: Most people living the area served by Translink do not live within walking distance of a SkyTrain Station. Most buses outside the City of Vancouver and even some within have terrible service. If they had decent service, one would not be able to get on a SkyTrain esp during the rush hours and it would be packed during the day. One thing that the area is promoted is developments within walking distance of the SkyTrain. They account for a sizable percentage of the riders during the day. Rush hour succeeds because of the very expensive parking rates downtown. Of course, people without cars have no choice.

This is the difference between U.S. and Canada (and other developed countries). While other countries are building efficient transit lines that bring in really high ridership and worthwhile taking instead of driving, the U.S. continues to build useless s**t to nowhere that runs really slow.

So why are American transit planners so FUCKING STUPID??? Forget about Europe or Asia, just look north to Canada for successful rail transit!!! If I were to grade the efforts of new light rail lines since 2000, most of them would get a clear F. The best ones only deserve a C-.


@george lee
Transit service is crap because the majority of people don’t demand high standards or hold politicians responsible for the failures. And the biggest thing is that most people still drive, so they don’t give a f**k anyway, as far as they’re concerned public transport is for the poor and hipsters that read blogs like this one. Every time I go home to the SF bay area for the holidays (coming from Tokyo) I feel like I have arrived in the third world wrt to public transport.

What say is mostly true, but ridership on rail lines is slowly increasing. Reward them with better service.

A useful addition to the table would be end-to-end average speed by schedule. That way more of us could point out the flaws of Light Rail lines that still need grade separations and how to judge proposed Light Rail lines.

Shouldn’t the FTA require that future Light Rail lines to receive substantial funding meet these minimum standards:

22 mph average speed
5 trains per hour + 3 cabin trains (morning and evening commute time)
4 trains per hour + 2 cabin trains (off-hours)

Otherwise, why fund them?

Some of the comments note Vancouver’s success, but let’s not paint all of Canada with the same positive brush! Coming from the Greater Toronto Area, I can say its clearly not the same in some of these parts.

For example, the provincial and federal governments are contributing significant capital dollars (billions) to the York Region VIVA Bus Rapid Transit system (learn more – While the system is currently hybrid BRT (until full build out), the service frequencies are based on current travel demand which range from a bus every 3 to 19 minutes during the AM peak. Service standards note a bus every 15 minutes all day every day on all routes. Even at full build out, service plans note a bus every 7 minutes in the peak period. While this is more than some of the rail lines noted above, I would call current and future service pathetic in light of the project goal – transforming modal share in a suburban car oriented region.

why you may ask? some might note that the provincial government does not contribute to municipal operating costs. (as it once did). They are there to cut the ribbon, but there is no commitment to support money losing lines that have regional (and provincial) benefits. And municipal leaders in York and other nearby regions are often hesitant to support ’empty buses’ regardless of ‘if you build it’ arguments and evidence.

From a policy perspective perhaps state and provincial governments should be building funding programs that are longer term and support capital and operating. For example, an initial capital subsidy with service benchmarks at launch and then an ongoing (small) operating commitment to sustain these levels. The operating commitments would slide higher if service is increased more than the initial benchmark. This would last a certain period of time. In general it would encourage a sustained commitment until ridership reaches a point of no return, or at least when the cost recovery ratio hits a point that is affordable for cash strapped local governments.

Interesting article! But why does the table have an entry for San Jose: “Sacramento to Winchester”? I think that’s supposed to be Mountain View to Winchester. San Jose (really Santa Clara County) light rail does *not* go to the city of Sacramento.

Sorry I didn’t get to this discussion earlier. I was in Berlin, where dozens of lines were running on 20-minute headways till midnight, with owls running every 30 minutes. Oh, and I was in Amsterdam, where the streetcar line that best served me was on a 15-minute headway. The transit system that I just retired from tries to run a 15-minute headway on each rail segment for as many hours of the day as the budget permits.

The article focuses on only the aspect of headways. Experience long ago in Edmonton, and my experience in Denver since, and in talking with service planners in other cities, is that there are several factors going on at the same time. Were the federal government in the U.S. to add regulations regarding headways, there would be some perverse results.

Some factors to consider:

1. The above-mentioned cities, including Denver, have overlapping lines in the denser areas, so a table of headways simply listed by line is misleading.

2. 15 to 20 minutes has proven over and over to be the range in which casual riders are inclined to show up at the stops without having checked a timetable.

3. The quality of connections is almost as important as the headways. Berlin will put a bigger bus or tram on a growing line (or longer train on the S-Bahn) in order to keep the headway wide enough to make consistent connections.

4. THE CONVENIENCE OF A HEADWAY IS DETERMINED BY THE LENGTH IN TIME OF A PASSENGER’S TRIP. A 30 to 45 minute headway on a peak commuter trip with few stops and a long ride will attract almost as many riders as a 15 to 20 minute headway. The same headway in a dense area with many possible destinations within shorter distances is unsatisfactory to most customers.

5. Comments assuming that all of the rail aspects are driven by real estate developers ignore the role of customers. This includes agitating for service cutbacks on routes other than their own. That is especially true during economic downturns, when customers on low-productivity routes will demand cutbacks on routes that have more service — even when those other routes need the extra service for capacity reason. In the recent Great Recession, one Colorado municipality had staff going through light rail numbers trip by trip. Their report fingered the pull-out trips from the yard as being unwarranted.

6. The role of cost-savings in infrastructure was touched on in the discussion, but it is a much bigger problem. In order to overcome some “value engineering” results, it is necessary to either buy more rolling stock or widen headways. At InnoTrans this past week, I had an off-the-record conversation with a German working on a new light rail line who was experiencing the same issue, so there’s no need to beat up the U.S. alone for doing this. Much of the problem is driven by modern media, which tends to get fixated on a single cost figure. If the mature European systems were being built in today’s media spotlight, they’d be deleting the short-turn track and crossovers, too.

If you want people to ride the rails without a timetable the minimum tph should be 6, not 4. 15 minutes is too long of a wait for the next train. for peak hours, they should run 10 or more trains per hour. you should only have lower frequencies if several lines are bundled for a significant portion of the system. Seattle’s Link, for example should have no excuse for not running 10 tph.

In Cleveland, we have two rail lines–the heavy-rail Red Line, half of which was built as an express corridor for interurbans, and the Blue and Green Lines, collectively known as the Shaker Lines, which are light rail and which were also intended to carry express interurban traffic. As it happens, Cleveland RTA operates maybe 70% of the service it did ten years ago. The base service on our rail lines is every 15 minutes. Maybe 20 years ago, there was a major push to improve service systemwide, and the Red Line at least had 10-minute midday headways, 12 minutes on weekends.

RTA has the highest transit ridership in Ohio. It also has vastly reduced service in primary corridors. Granted, it’s not at all easy to maintain service in a county where the population dropped by almost 40% between 1970 and 2010, but this also led to a failure to expand rail service in more corridors. We have rights of way which could be used, for expansion to the west, southwest, northeast and southeast. There isn’t the political will to do it, but there’s also the question of who the hell would pay for it.

In all likelihood, the Shaker Lines wouldn’t qualify for New Start funding today. There is a proposal to extend the Red Line northeast, to the city of Euclid, via one of two paths. Both corridors are hemorrhaging population, but either one gets you to I-90 and a very convenient connection to two suburban corridors. In the west, service to Lakewood and (maybe) Rocky River would almost certainly generate high enough ridership to justify construction. As it stands now, the “priority bus” project in the Clifton corridor doesn’t do a damn thing to return bus service to a 15-minute midday headway…so there are millions of dollars being spent on a median, and on silly brick bus shelters which look like something out of a misbegotten Safety Town. But that’s what we get, including continued crappy service, because there is no funding and no political will to build the rail line which would–even at 15-minute headways–be an immense improvement.

What is the threshold for service? In Cleveland, we just might be able to regenerate enough parts of town to justify train service back at a 12-minute, even 10-minute headway, and an extension of the Red Line might mean 7 1/2 minute headways from University Circle to Downtown (and possibly west). But what has been pushed on us is the “Health Line” BRT project, also under consideration for eastward expansion. Sure, the service is better, but it seems that in all the self-congratulation about its ridership increase, nobody at RTA ever stopped to consider whether they could upgrade to double-articulated buses when headways hit 5 minutes or less. Meanwhile, the signal activation for those buses has been switched off, with no public explanation.

Yes, I do think it’s irresponsible to plan to build any infrastructure where there is no provision for funding intensive operation. In Cleveland, with ample fracking to the east, one of the saddest reasons for our service collapse is that our transit authority scrapped ALL natural-gas-powered operations, so they’re stuck paying for diesel when natural gas is far cheaper. We at least have the excuse of a monumental infrastructure mistake, compounded by declining population. What’s the excuse in San Jose, or San Diego, where rail headways are no better than ours? What’s the excuse in Miami, where a functional plan for Metrorail expansion was consumed by corruption? In DC, they’re spending billions of dollars on the line to Dulles, knowing full well that the trunk tunnel across the core of the District doesn’t have the capacity to carry every scheduled train. I guess we in Cleveland should be grateful that we can maintain a 15-minute headway, and that we can very probably maintain it.

I’m from the Raleigh NC area. We don’t have any rail service, but we have some similar issues. Though clearly better than some areas such as Detroit’s struggling SMART service, in comparison to even nearby cities like smaller sister city Durham NC Raleigh’s bus service is lacking. From what I understand, one of the busiest routes in Raleigh (Capital Blvd) runs every 30 minutes- at peak… Keep in mind this is a route on Raleigh’s busiest corridor in a city with a population of over 400,000. Other routes barely reach the areas that need service the most (low income areas in southeast Raleigh). Of course, the complete lack of sidewalks and immense number of country-style city streets contribute to the issue (not walkable at all!), but as we begin to discuss the issues and propose expanded service (and even a proposed light rail!), it seems to me we really need to improve the service and infrastructure we already have. At least 5 or 6 buses an hour during peak times on a corridor that busy makes sense to me. However, at least they seem to be making an effort to improve so they deserve some credit.

I may be wrong about the 30 minute headway. It may be 15 minutes at peak. I live about 30 minutes outside of Raleigh and I do not get to use the transit.

Leave a Reply to ThomasD Cancel reply