Friday, December 28, 2012

Valley Transportation Myopia

The San Jose Mercury News has an article about the 25th anniversary of the Santa Clara Valley Transportation Authority's light rail system, quoting a few supporters and one critic. The supporters are VTA transportation planning manager Kevin Connolly, "the train's godfather" Rod Diridon, and a few riders, all of whom have Spanish last names and say that they ride the trolley because they don't own cars. The single critic quoted is Tom Rubin, an accountant who gets paid as a "transportation consultant" who mostly gets paid to attack transit. Sometimes Rubin will make arguments that seem "pro bus" in order to attack rail, but when there's no rail involved he's happy to attack buses.

On Twitter, some transit advocates did criticize the VTA. They're probably right. I've never been to Silicon Valley, but it sounds like it was relatively badly planned. It also sounds like there's a zoning issue behind the fact that many of the farms bordering on the lines haven't been built up with dense, walkable neighborhoods. That said, there's a bigger factor at play, one that I've touched on many times in the past: transportation myopia.

In the twenty-five years since the first VTA trolley ran, the federal, state and county governments and the VTA have widened four competing highways and built numerous interchanges and other "improvements." Here are a few that I could find details on:

CA 237Convert to "freeway standards"1997
I 880Widen from US 101 to Montague Expressway2004$76.3m
CA 17Add auxiliary lanes2007$28.2m
CA 87HOV lanes south and north2007$121.9m
I-280Ramp Metering and Widening2010$5.5m

As you can see, just about every branch of the VTA light rail system has seen millions of dollars invested in competing roads. Add to that the cost of constructing the 101, 280, 680 and 880 to begin with, which only happened within the previous twenty-five years. Those wouldn't have affected the design of the system, and thus the "Cost to run a light rail vehicle for an hour," but they have definitely sapped ridership, which affects all the other indicators mentioned.

Rubin is actually half right when he says "I think the original concept was very seriously flawed." Whatever the flaws of the original trolley concept and zoning, they pale in comparison to the flaw in the concept of building a trolley system at the same time as you expand the competing road network.


murphstahoe said...

The highways have not sapped light rail ridership. The light rail is at a standard below which anyone not taking it would bother. If the traffic congestion were so bad that light rail was preferred, then the result would be that people would leave the Valley and move to somewhere they can get around.

Traffic congestion is bad as it is. The sections denoted in that graphic are notoriously gridlocked. They opened a HOT lane on 237 and supposedly it saves 25 minutes on the drive from 880 to 101 coming from Milpitas. We ride on a bike path that parallels 237 and we easily outpace the cars next to us on 237.

That is why you see housing prices skyrocketing near Caltrain stations. That is why you see the private shuttles and skyrocketing housing prices near the shuttle stops. People are more than willing to switch to transit but the transit has to be nominally useful, by my definition that has to at least be "faster than a bike". This one isn't.

Following the LR line from Mountain View to Downtown SJ on a bike I could easily beat the Light rail. Not to mention that on a bike I could take a more direct route.

They put about 50% too many stops on MV->Tasman/First. So you can't even solve the problem by eliminating half the stops - then you would have too few. No signal priority. etc...

I was very excited when this line opened and it took maybe 2 rides to just decide it was "for emergencies only".

Good transit systems are seeing massive ridership increases. Bad ones are not.

Stephen Smith said...

Better zoning would certainly help VTA light rail, but zoning is not the primary issue. The primary issue is a lack of connections throughout the Bay Area. It basically connects downtown San Jose (which, despite styling itself as the "Capital of Silicon Valley," has very little housing density and only contains one name-brand employer: Adobe) to a bunch of parking lots dotted by the occasional office park. You can transfer to Caltrain, an hourly off-peak (and not in any clockface manner!) and expensive commuter rail line, but it won't even take you directly to San Francisco's financial district, where all the jobs and regional transit connections are (getting from a VTA light rail stop to, say, Berkeley's campus would involve a hilarious number of connections and different tickets). Some day, if Carl Guardino gets his way, you'll be able to transfer directly to BART from VTA light rail which will make it somewhat more useful, but you'll be connected to SF through a circuitous East Bay route.

Around BART stations, there is generally demand, and the problem is zoning and the towns not allowing enough development near them. But BART actually takes you somewhere you want to go – VTA light rail doesn't. If there's demand around VTA stations for development (and there probably is), it's not likely to be significantly more transit-oriented than development in the rest of San Jose that's not light rail-accessible.

Stephen Smith said...

I should amend that by saying that the downtown San Jose segments of VTA light rail actually are harmed by zoning, because there's actually demand there, given the direct Caltrain, ACE and Capitol Corridor access, as well as the very very very very very faint markings of urbanism in downtown San Jose.

If downtown San Jose were more developed and were more of an anchor, I could see zoning being an issue on the rest of the light rail corridors.

calwatch said...

A decade ago (almost to the week, wow!) I compared and contrasted VTA with another suburban transit agency, OCTA in Orange County, CA:!search/vta$20octa/ba.transportation/zBc5O3WDdcU/4B58tvPNEa4J

VTA has had problems for the last two decades. They are an inefficiently run system with a lot of slough. The OCTA ran a comprehensive bus system on grids to complement the commuter rail network. Bus ridership has stayed stable despite massive freeway expansion in the OC. Although there was a setback in 2002 due to an overly aggressive restructuring, they hit their zenith of service in 2008, with the central county grid covered with 15-20 minute service during peak hours. OCTA also used contracting judiciously to run smaller vehicles when necessary, and used commuter rail as a backbone to provide "extender" service. VTA meanwhile has had high labor costs, even before the 2000 boom, and have never been able to get that under control.

Even today one major complaint about OCTA is crowded buses, and to use fare increase money to alleviate overcrowding. I never hear about overcrowding in VTA land.

Peter said...

I've lived in most parts of the US and have lived in downtown San Jose for almost 3 years now -- my general take is that the VTA train system/management is not much worse than most other light rail systems in the country.

We built the trains to service all the high tech companies, forgetting that people with money don't take transit unless they have to (or if they live in some density anomaly like NYC or SF).

That SF and NYC have crazy-high population density (thus, light rail/subway ridership) is not because they're smarter or less psychopathic than SJ -- it's because they're bounded by water.

I live in downtown SJ, right next to a light rail station. I've ridden it more than a few times, but generally do not - it's too inconvenient/seldom, slow, often late, too expensive, and lacks safe/convenient/dignified access to the stations, among other things. Same as most other US light rail systems.

Still, I like having the train there. I like hearing the bells chime starting around 5:30 am weekdays. I'd gladly pay a $100/mo surcharge just to have access to the train when I want it. Transit advocates in America need to learn from Japan.

One simple solution to making ridership skyrocket would be to make it free.

But free transit fares won't punish poor people, so it's not something transit advocates will get behind here.

The trains could be automated, but Silicon Valley is concentrating on getting cars to be self-driving first, to make sure we protect the auto industry -- I guess there's more money in that.

And we're concentrating on BRT -- to make sure people stay in their cars.

I've lived in most parts of the US, and San Jose has always seemed the most sprawly to me. That includes Austin, Kansas City, Atlanta, and other horrors of sprawl.

Here, we have highways _everywhere_.

My 'locate your business here' pitch:

"San Jose -- Anywhere to anywhere in 10 minutes!"

This place reminds me of Buffalo -- such an easy commute to everywhere.

There are more than a few light rail-oriented developments popping up, tho, like this one, which is now almost done -- along with the one in the distance:

But we'll be exactly the same as the idiots and capitalists in San Francisco -- we'll have plenty of parking under all these living spaces, only we won't be constrained by water on three sides -- so we'll stay sprawled, have less ridership, etc.

The people who most need to use the trains will be priced out of the walk-shed to the train stations.

Biking here is still near-impossible and bringing a bike on the trains is accommodated, but in a very awkward/uncomfortable fashion.

Bike sharing is coming to the entire Bay Area...soon?? -- that'll help train ridership, in theory.

There was some overcrowding of some VTA buses up until recently, when the class warriors raised the fares again.

It'll probably take another 30 or so years before we start to use the trains. The Bay (meaning, the body of water known as 'The Bay') has to start flooding some rich people out of their homes and offices before we make the train system useful, and that may happen sooner than most expect.

Anonymous said...

All SJ shows is that you can't get lots of riders by focusing on low-density office parks. Doesn't say anything about tech workers not wanting to take transit (they do, disporportionately, as shown by the fact that in the same region, so many take the employer shuttle buses).

Matt said...

^What everyone else has said.

Unfortunately we seem to invest so much in transit lines to serve suburban job centers, but then we build freeways and encourage companies to locate in large campuses behind parking lots. Total myopia.

murphstahoe said...

Caltrain isn't expensive. The anti-Caltrain argument is always based on the non-clipper one way fare. Clipper provides a discount, but if you have a Monthly Pass the train is ridiculously cheap. Monthly SF to MV is 179. Factor in $125 pre-tax benefit and that drops closer to $140. I rode daily - $7 RT, which means I was paying $3.50 for a 40 mile train ride.

Those ragging on Caltrain's off peak schedule neglect to mention than on-peak the frequency is higher than that on BART's trunk lines. The response? Not all trains stop at all stops. Go rag on the NYC subway then. Or BART that is now giving power points about "express service" 15 years in the future. And Caltrain actually can be improved from the decent baseline. BART is maxed out.

It's less than 10 minutes tops by MUNI to any market street destination, 5 minutes by bike. The majority of the stops are central to peninsula downtowns. The San Jose stop is a disaster and that is on San Jose, not Caltrain.

Caltrain is the poor stepchild but the ROI of that system is ridiculously high. And if you look at their major capital expenditures, they have added value - baby bullet, grade separations, removal of holdout stations. Compared to various VTA LRV extensions or West Dublin BART/OAC? No comparison.

Bureaucromancer said...

Re: murphstahoe

Looking at the situation from across the continent it's seemed to me for a long time that Caltrain really should be given to BART. Admittedly that probably means a fight with the BART Around the Bay advocates (though BART as locals that can deviate from the rail corridor and quasi HSR limited stop Caltrain expresses might be a good option if full quad tracking and grade separation isn't happening on the peninsula), but you get a more obviously connected system, fare integration and probably better funding with what is at the end of the day mostly a branding and administrative change.

As far as SCVTA goes my impression is that it's problems could largely be solved with a small number of low cost (as such things go) rail projects and a reorganized bus system. As was said above, there really is a need for a (quasi? the urban structure doesn't necessarily seem right for a pure grid) grid based system well integrated into feeding the rail lines. The east valley project is pretty vital, and should be rail (even if more streetcar than LRT) but beyond that SCVTA need to get it's bus system in order if it want's to attract riders.

Anonymous said...

I've been living in Santa Clara Country for the past 4.5 years, and have had to take VTA almost every day. Here's my take on the VTA and its mediocre light rail. First, the operational issues that are relatively minor:

1. The MV-Winchester line needs to run at 15 minute intervals throughout the day. Stop terminating certain trains at Gish.

2. Run trains from Alum Rock to MV. Transfers are transit killers. Plus its confusing for first timers.

3. Have more late night trains from MV that coordinate with Caltrain. As it stands, the last SB lrt from MV leaves at 10:50 7 days a week.

VTA problem in general:
They dont' give a shit about transfers. Often you'll be a minute or two late, and have to wait half and hour, or even an hour, to wait. People dont' want to wait 30-60 minutes at some godforsaken suburban light rail stop/bus station.

In general, the light rail system is very poorly laid out. Ideally, most of it would be torn up and laid down anew. Ironically, the much maligned freeway alignments provided pretty fast service when I lived off of Snell Road in South San Jose. Its all about having a dedicated right of way.
There's no way to make north 1st and Tasman feasible streets for light rail other than to elevate it (which would in one sense be easy, as there's few people out there to complain about blight).

On Caltrain:

Off peak service is too infrequent. Who cares about weekday commute hour frequency, try waiting for 45 minutes at Millbrae BART on a weekend for it. It sucks.

Also, Caltrain...well, I think its expensive, and if you've ever taken it during the weekday commute, you'll notice most of the people are white collar professionals (regardless of how they actually dress). Its not a poor person's way of getting to work regularly.

Alon said...

Santa Clara County widened 101 in the late 1990s; Caltrain ridership on the Gilroy-SJ segment took a nosedive and hasn't recovered, even as SF-SJ ridership keeps going up.

Anonymous said...

Two years ago, the VTA put signs up at the Mountain View light rail station saying there'd be no extra New Years Eve service (Caltrain runs extra late nite southbound trains that night). They are running late night trains thus year, but it's that extra special attention paid to fucking over passengers and persuading them not to ride that sets VTA apart.

J said...

murphstahoe hits a major point; as someone who grew up in and still lives in SV, I do NOT see highway widening and road expansions have been the biggest single problem promoting low VTA ridership. Traffic is horrible on 101, 237 and 880 during commute hours as is (and will continue to be even with wider highways), but SV commuters still drive because driving is ultimately faster and more convenient than what VTA offers.

Promoting strong ridership on VTA won't be easy simply because of SV's geographical layout--SV is a massive, sprawling region without any concentrated focal point. Stephen pointed out that SJ often calls itself the capital of SV, but that's really only because it has the highest population count. SV companies (and their workers) are sprawled out across the entire region: A handful in Palo Alto (HP, VMWare, Tesla) a handful in Mtn View (Google, Intuit, MS), a handful in Milpitas (KLA Tencor, Flextronics) a handful in Sunnyvale (AMD, Amat), another handful in Santa Clara (Intel, Marvell)--you get the idea. Caltrain and BART can depend on a heavy segment of its ridership converging to SF (or in BART's case, also Oakland and Berkeley). But for VTA, there isn't one single point where SV workers all converge.

Sprawl-inducing policies are likely the culprit, though I don't think it'd be possible to pinpoint just one particular policy that should be changed. SV today is the result of decades of suburbia-loving workers meshed with the lack of a unified planning vision across the tens of cities

calwatch said...

Why does Orange County Transit work so well then? Because they have concentrated on the transit dependent rider, with some "low hanging fruit" services for choice riders primarily as direct shuttles from Metrolink. If you beefed up services in the central core, and started thinning buses in the office park areas, you might get more productivity.

neroden@gmail said...

I'm going to make a deeply obvious comment. If you looked at an aerial photo of the San Jose area.... *would you put the light rail line there?* And if you did *would you build it this way?*

No, you wouldn't.

The Winchester to downtown route is a plausible route, I guess. But it gets worse from there.

The Alum Rock route is... huh, seriously? Who thought of this? It connects Alum Rock to where exactly?

The Mountain View route has the right endpoints, but is very excessively circuitous.

The First Street route is OK as a route... but built to be slow -- which is not the right choice on the trunk segment. You could get away with this with a densely developed downtown like Portland, and even there it's not a good idea, but in San Jose? For that many miles?

The Almaden and Santa Teresa routes -- highway and low-density sprawl. Not where I'd put a rail line.

Oh, there are more mistakes made in VTA. It's pretty much a poster child for how not to do light rail. Contrast San Diego.

Marc said...

Most of Caltrain's daily ridership is, in fact, commuting to jobs on the peninsula. The ridership stats show that during morning peak 5800 people get off the train in San Francisco. Even if one assumes most people who get off at Millbrae are taking BART to SF (which is a poor assumption) the total goes up to 6900. By comparison, over 7700 people get off the train at one of the stations between Hillsdale and Mountain View. Of those, 5100 get off in Santa Clara county between Palo Alto and Mountain View.

There are concentrated employment pockets in Santa Clara county, VTA light rail simply does not serve them very effectively...

Anonymous said...

There are also massive swathes of Santa Clara County that are simply not served by light rail-Cupertino, Mountain View, Sunnyvale, Santa Clara, and Palo Alto (yes, technically the MV-Winchester line goes through parts of Santa Clara, MV and Sunnyvale, but not where the majority of people live). I also think the VTA has to be eyeing the last of the Vasona line, when the quarry in Cupertino goes under/stops shipping by rail. How effective of a light rail line it'd be I don't know.

arcady said...

It's not enough for a light rail line to serve residential areas, and it's not enough to serve employment centers either. The light rail line has to actually get people from where they live to where they work, and do so reasonably quickly. The problem is that significant parts of the VTA light rail network fail at that, in particular the whole Mountain View line, which has a long and slow run through the office parks, and seems to mostly be effectively a very expensive shuttle to get people to the Caltrain. The fast parts of the lines to Santa Teresa and Winchester are actually reasonably successful, and the line North 1st is not too slow either and gets decent ridership, but the slow, circuitous sidewalk-running alignment through Downtown probably hurts ridership considerably, given that Downtown isn't that huge of a destination.

Anonymous said...

I think that North 1st, Japantown to Tasman, is worse than downtown SJ proper for slowing trains down.

J said...

This is a fantastic conversation to be having. Arcady hit it by saying that it's not enough to serve just residential areas or just employment areas.

I think what really differentiates the success of Caltrain and VTA LRT is the development around their respective stations. Yes--Caltrain has a lot of people getting off at SF, as well as Peninsula stops too like PA and MV. But look at the PA and MV stops (or any of the Caltrain stops, for that matter), they all have a vibrant downtown surrounding the station with restaurants, shops, residential patches and jobs. In short, Caltrain stations are surrounded by dense development (in terms of density, jobs, and other activities that attract people). Caltrain (or, the rail system that existed a century ago) spurred TOD.

Now take a look at VTA LRT. Have you seen Evelyn Station? That place has nothing around it. And every single stop on First St (the backbone for the LRT system) serves some SV company--but basically nothing else. VTA LRT actually does an excellent job serving SV workers in terms of reaching their offices--just off the top of my head, LRT serves the doorsteps of Lockheed Martin, Cisco, NetApp, Sanmina, Yahoo, etc. But these companies are all in office parks with nothing else around them. The stations become useless except during rush hour. You get ultra low ridership as a result (and even lower given that most SV workers don't even bother with the slow LRT speeds).

Perhaps the biggest takeaway from VTA LRT's failure is that density around its stations (and not just residential density, but also density of other development such as employment and activities) is the key to transit's success. Of course, VTA LRT also does a lousy job of transporting people efficiently to begin with...

arcady said...

The thing with density, though, is that it can slowly increase, and is in fact doing so. They're building taller and taller office buildings in the office parks, going up as high as 8 stories at Moffett Park. And there are huge new developments going up at Race and Fruitdale stations. But neither of thoe will change the fact that train just takes too long to get from the one to the other, except slightly and indirectly, through increased traffic congestion.
The thing that does make the existing light rail more effective is more mixing of uses along the line, which is also slowly happening. There's a new development at the Fair Oaks station with condos and a supermarket, there are some huge new condo buildings at Mountain View, and there's also something pretty big going up by the Baypointe/Tasman stations. This will put more residents within a reasonable travel distance of jobs on the light rail, and should increase ridership.

And @amandainsjc, we can argue about what's slower, but it takes about as long to get from Japantown to Tasman (5 miles, 17 minutes) as it does to the Diridon Station (about 2 miles, also 17 minutes). A 17 mph average speed is pretty good for a median-running ROW, though only half as fast as a fully separated line (with proper grade crossings) would be. But a 7 mph average speed is just pathetic. And the indirect route through Downtown makes the effective speed even slower, to the point where it's almost faster to walk.

neroden@gmail said...

The indirect routings are the real killer here. And the Mountain View one is the worst by far, which is particularly bad given that that's the one heading towards the dense employment of the Peninsula...

Squirrelly, slow detours are not a good thing. Sure, the Winchester line has to do some tricks to get under the mainline rails, but the other detours are just gratuitous.