This past April, I spend 25 days in Japan, where I took dozens of trips on railroads and subways. In each case, the train left exactly on time and arrived on time. This past weekend, my wife and I decided to take a ride on the “Gold Line” from Union Station in downtown LA to Arcadia, a trip scheduled to last 35 minutes. It was a disaster.
Los Angeles is unique in terms of its lack of rail service for a developed metro area with more than 10 million people (indeed nearly 20 million if you include suburban areas in neighboring counties.) They have tried to remedy the situation with some newly built rail lines, with more under construction. But it just doesn’t seem to be working.
Our trip took an hour and 20 minutes, more than twice the scheduled 35 minutes. Even though (we later found out) the delays were due to maintenance, and not unexpected, this information was not given to passengers who boarded the train. The ride itself was also unpleasant, with loud music being played by passengers. My wife vowed never again to take mass transit in LA, and I can’t blame her. I’ve had similar bad experiences in New York and Boston. Washington DC’s subway is better, but also has frequent delays.
Elsewhere in LA the situation is even worse. The even newer “Expo Line” is a surface line that must frequently stop at red lights, as cars are given preference over trains. Thus this hugely expensive new rail line (costing $2.5 billion) is little better than a bus route:
The Expo Line is infamous for being slow because in certain sections it stops at red lights. The poster nailed it with this: “If it is at-grade and doesn’t have signal priority you might as well write the word ‘train’ on the side of a bus and you’d save money ….”
More such lines are currently under construction in LA, funded by a recent tax increase. The planned high speed rail between LA and the Bay Area is an especially costly fiasco. It would make more sense to stop building new public mass transit infrastructure, and instead work on improving the existing lines. For reasons that are difficult to discern, LA’s mass transit has extremely high maintenance costs, despite being a relatively new system that is not heavily used (and not having to face ice storms, as in Boston.)
How should the current mass transit lines be improved? I see little evidence that spending more money would help—-the US already spends far more (per mile) on transit lines than other countries, with much worse service to show for it. Instead, we should look for alternative models of management. Here’s how it’s done in Japan:
In Japan, being in the railway business means being in the real estate business, explained Egon Terplan, SPUR’s regional planning director, at Thursday afternoon’s panel discussion about what the Bay Area can learn from Japanese transit station area development. “They are able to capture the value of the train stations they are building and beyond. One third of the revenue is from retail, services, hotels.”
That’s because rather than contracting out the business opportunities on the real estate around their stations, they own it all–everything from department stores to vending machines on the platforms. That has turned Japan’s six passenger railway companies–Hokkaido Railway Company, East Japan Railway Company, Central Japan Railway Company, West Japan Railway Company, Shikoku Railway Company, and Kyushu Railway Company–into hugely profitable corporations.
“These are companies listed on the stock exchange; they make money,” said Terplan. They also, together, carry nearly a third of the world’s railway passengers.
In Japan, the profit motive of real estate, retail, and office space–in addition to the trains–becomes a bit of a feedback loop. The Japanese railway companies want to maximize the value they derive from space around the stations. So transit oriented development isn’t just about housing. In Japan, it includes department stores, office buildings, shops, and hotels, and housing on different levels directly above and below the stations.
Obviously there are many geographical, cultural and political differences between the US and Japan, and I’m under no illusion that it would be easy to implement the Japanese system over here. But this is the sort of reform that policymakers need to start thinking about if they wish to improve rail service in the US. Get the right incentives in place. Simply throwing more of the public’s money at the problem won’t work.
READER COMMENTS
Brian Donohue
Nov 6 2018 at 3:22pm
Doesn’t it come down to population density? My hunch is that Canada isn’t very successful at public transit either. New York, Boston, Philly, Montreal, Chicago, Toronto, San Francisco, (maybe) Vancouver are probably the only real North American candidates for viable mass transit. Washington and Atlanta are more dubious despite having newish systems, and LA is hopeless.
LK Beland
Nov 7 2018 at 9:23am
Nearly a third of the Canadian population lives in the greater Toronto and Montreal areas.
They both have excellent mass transit (by N.A. standards).
LK Beland
Nov 7 2018 at 9:28am
Actually, as you can see, people in most Canadian cities use mass transit more than in major US cities.
https://en.m.wikipedia.org/wiki/Public_transportation_in_the_United_States#Usage
nobody.really
Nov 7 2018 at 2:37pm
Absolutely. Every time I try to work with people in LA, I remark on how dense these people are. Truly hopeless.
Vincent Passanisi
Nov 6 2018 at 5:10pm
Regarding California’s high speed rail fiasco, I think the best solution is to ditch the rail in favor of the first highway built strictly for driver-less vehicles. It could be funded by the private corporations that wish to produce those vehicles, and could be designed from the ground up to accommodate the technology necessary to make those vehicles safe, efficient, and cost-effective. I’m pretty certain the commuter estimates for such a highway would dwarf the numbers for the rail project, if only because of the immense flexibility for travel that a highway with individual vehicles presents.
I was unaware of the Japanese system for transit area development, but I had thought about something similar regarding a driver-less highway. If private property owners could benefit along the highway, imagine the opportunities to serve travelers that might emerge. Companies would have to figure out how best to accommodate these types of vehicles and travelers. And they would have the opportunity to solve problems like the best way to merge while entering or exiting the highway.
You could even have special lanes reserved for driver-less trucks and freight. My imagination is limited, but the possibilities are beyond comprehension no doubt.
Maniel
Nov 6 2018 at 8:45pm
Vincent,
I concur. I recently presented a paper on the subject. The abstract is available at the URL below.
https://ascelibrary.org/doi/10.1061/9780784481530.006
Maniel
Scott Sumner
Nov 6 2018 at 5:28pm
Brian, Yes, density is a problem here (except in NYC).
Vincent, If given the opportunity, I believe the private sector could find a better solution. The role of the public sector might be in putting together the right of way via eminent domain.
Thaomas
Nov 6 2018 at 6:14pm
It is hard to know the value of public transit without use and congestion of roads and streets being priced. That dedicated bus lanes is a better investment than most rail is a petty good bet, however.
Mark Z
Nov 7 2018 at 5:18pm
I’m almost inclined to generalize this statement and say buses are a better investment overall than trains when it comes urban public transit (though I’d probably prefer privatized buses). I don’t think trains are even significantly better in terms of carbon emissions (if at all; what little I’ve read on the matter seems ambiguous on that question).
Between the fact that vehicles have been getting consistently more fuel efficient since the mid 2000s, and vehicle miles per capita have been declining since around the same time, as people reurbanize, lower their commute times, and own their own cars less in general, I’m not sure even the environmental advantage of trains, most generously interpreted, would make them a worthwhile investment these days.
Benjamin Cole
Nov 6 2018 at 7:31pm
Scott, you will be happy to know I authored an op-ed essentially saying the same thing you say here, about a decade ago for the Los Angeles Business Journal.
Again, the problem comes back to property zoning. Residents around mass transit stops will not tolerate the density required to make those stops profitable from a development perspective.
As it is, the United States appears unable to build infrastructure for costs that equal benefits. The nation evidently can also not build more than 1.2 million housing units a year, partially due to property zoning.
So where to place all the immigrants?
Mark Bahner
Nov 7 2018 at 12:03am
No one should be doing infrastructure spending on public mass transit, because it’s yesterday’s technology. The future of transportation is transportation-as-a-service provided by autonomous vehicles.
The benefits of transportation-as-a-service provided by autonomous vehicles will include:
The service provider(s) will know where everyone is going and when. That will allow the providers to see where and when it makes sense for people to travel together. For example, the George Washington Bridge between NJ and Manhattan regularly experiences rush-hour slowdowns. These could be completely eliminated by simply packing more people into vehicles.
Speeds will be much higher. Interstate speeds, even within cities, will be 100+ mph.
Accidents will be much less frequent, and there will be zero “rubber-necking” that slows down traffic near accidents.
Autonomous vehicles will even allow traffic situations that would be insane with human drivers. For example, the GW Bridge is typically slow going into Manhattan in the morning, but coming out of Manhattan in the evening. With autonomous vehicles, it would even be possible to switch lanes that normal go out of Manhattan in the morning to going *into* Manhattan in the morning, and and vice versa. That would be insane with human drivers, because it would be impossible to make sure all the human drivers “got the memo” about the lane direction. It’s a piece of cake for autonomous vehicles.
In addition to traveling faster and with fewer accidents, the separation between vehicles will be significantly reduced while maintaining safety. In fact, traffic lights will probably almost completely be eliminated.
The majority of passenger miles traveled in the U.S. will be by autonomous vehicles providing transportation-as-a-service within 25 years. Contrast that to the NYC Second Avenue subway line, which was proposed in 1919, construction started in 1972, and the *first phase* of the line completed in 2017.
P.S. Here in Durham NC, we have a proposed multi-*billion* dollar light rail system. The only bright aspect of the proposal is that it almost certainly will never be built.
Don Geddis
Nov 7 2018 at 5:04pm
Lane switching doesn’t at all require autonomous vehicles. The Golden Gate Bridge in San Francisco already, today, has center lanes that change direction depending on the time of day (in order to match the commute, just as you suggest). It uses a movable barrier between directions, and a zipper.
Matthias Goergens
Nov 7 2018 at 8:23am
https://www.theatlantic.com/technology/archive/2015/11/why-dont-we-know-where-all-the-trains-are/415152/
There was a great article on New York transportation a while ago in The Atlantic.
Mark Bahner
Nov 7 2018 at 10:01am
Yes, that is an excellent article. Thanks for pointing it out! In my mind, this is the key point of the article, discussing how a new communications-based train control (CBTC) system will improve the subways:
One of the best ideas I’ve had in my engineering career is to change hurricane storm surge protection systems from fixed systems to portable systems (consisting of tubes filled with air and water, deployed offshore). One beauty of portable systems is that they improve as hurricane storm surge modeling improves (which is based on software and computer hardware improvements). Our abilities to model hurricane storm surge have been improving and will continue to improve very rapidly.
So what I take as a bottom line of that article and my portable storm surge protection system is that it’s good to transfer everything from mechanical problems to software and computer hardware problems. Both computer software and computer hardware can improve much faster than one-off systems like the mechanical systems now controlling NYC subways, or the levees and storm surge gates that protect parts of a very few cities in the world (e.g. New Orleans, Rotterdam, London, and Venice).
Matthias Goergens
Nov 7 2018 at 10:23am
Yes. Though if memory serves right about the article on the NYC subway the workforce is against any kind of productivity improvements.
Running public infrastructure efficiently isn’t trivial, but it’s not rocket science either. Lots of countries manage to do quite well, and the list expands when you take into account that the past is a foreign land as well. Ie even the US used to be good at it.
Matthias Goergens
Nov 7 2018 at 10:26am
On a similar topic, Wikipedia has an excellent Article on Privatisation of British Rail.
They show graph after graph of at least moderate success. And yet, opposition to that privatisation is almost universal amongst the public.
Alan Goldhammer
Nov 7 2018 at 11:43am
I rode the DC Metro to work every day for 24 years, commuting from my Bethesda home to a downtown business office. I would not have traded that for an automobile drive unless I was paid a boatload of money. We live within walking distance of the subway stop and in 99% of the times over this period the ride was about 18-20 minutes during which time I could read. The vast majority of the riders are respectful of the no eat or drink policy and listen to music only via headphones. It was cheaper for me to ride Metro than to drive when gas, maintenance and parking were factored in.
The Metro system is now recapturing real estate value as some of the parking lots are now being developed into housing and retail. I don’t know how much they will realize. The major problem is the system was not designed for the large ridership that materialized over the years and the dual track system is not well prepared to provide timely service when there are train malfunctions.
Mass transit is complicated and will not work in every area. Our DC Metro system is further complicated because it is overseen by three political jurisdictions.
JFA
Nov 7 2018 at 12:20pm
Alan, I don’t know when the last time you rode the metro was. The maintenance delays are pretty wide spread since DC Metro didn’t bother to 1) do necessary maintenance for years so now they have to do it all at once and (more importantly) 2) as you mention, they only have two tracks on each line, meaning that if one track is down, delays can become quite lengthy.
I would also say that a small but still large enough share of metro riders are completely oblivious of common courtesy: there are conversations over speaker phone and music being played via the phone’s speaker or some external speaker. Every time I take the metro, I encounter this behavior. I don’t use it every day, but I use it often enough to know that these aren’t aberrant behaviors. And it’s been this way at least since 2010, only now the service is worse. I don’t often encounter people eating food, so at least there’s that.
Mark Bahner
Nov 7 2018 at 10:03pm
Let me do a scenario with some hypothetical numbers for what your commute might look like in a future situation where autonomous vehicles are providing transportation-as-a-service. Let’s say, just for ease of analysis, that your door-to-door commuting distance was 10 miles and took you 30 minutes.
In the future scenario, you might check your smartphone a couple minutes before you leave for work. It might offer you a range of options:
Option 1 – You get a single-occupant vehicle that takes you right from your mailbox to the front door at work. Door-to-door time = 15 minutes. Cost = $1 per mile, or $10.
Option 2 – You’d walk a block or two to a single-occupancy car parked on a more main road, and you’d also be delivered a block or two from work. Door-to-door time = 25 minutes, cost = $8.
Option 3 – You’d get in a single occupancy car, and then meet with a minivan or small bus, and then get another single-occupant car in downtown DC, to bring you to your door at work Door-to-door time = 30 minutes, cost = $5.
Option 4 – You’d walk to a minivan, and then walk from the minivan to your door at work. Door-to-door time = 40 minutes, cost = $2.
There might be many more options. Your smartphone might hide the other options, based on knowing your history of choices for door-to-door time and cost. Your choices might vary based on weather, so the smartphone might even take into account weather in deciding which choices to put forward first. Your smartphone might also notice that your preferences might change depending on whether you were going to or from work (willing to pay more money for a shorter door-to-door time when going to work versus coming from work).
Jon
Nov 7 2018 at 11:51am
Even the portion of the gold line that doesn’t have stop lights along the 210 has a top speed of 35 miles per hour…
Taeyoung
Nov 8 2018 at 1:15pm
The fact that Japanese rail operators also generate revenue from department stores and shops located in and around rail lines helps them with their overall profitability. But the real difference between them and US operators is that their operating costs are so much lower. If they were paying as much per passenger-mile as US operators, they’d all have gone bankrupt long ago. Partly this is because they run trains more frequently — wait times (even on the intercity Tokaido Shinkansen!) are lower than I get on the Washington Metro. And partly this is just because each trains is, on average, carrying more passengers than a train in the US. So it’s not an entirely fair comparison. But when the operating costs for the Washington Metro are something like 70% of the operating costs of the Tokyo Metro, despite the Tokyo Metro servicing approximately 10x as many passengers per day, we’re obviously paying too much for what we get. And this is only exacerbated by the revelation, a year or two ago, that WMATA staff (specifically, a bunch of inspections staff) weren’t even doing the job we were paying them to do.
Patricia Chen
Nov 9 2018 at 3:19am
This is an extremely unscientific survey of two people taking one bad trip. I have taken the Blue Line to work most every day for the better part of the 28 years since it opened and I have found it reliable and convenient the vast majority of the time. I highly recommend it.
Larry Stevens
Nov 9 2018 at 7:27am
The combination of insane infrastructure costs with explosively advancing tech makes rail expansion an iq test that liberals uniformly flunk. Per mile construction in NYC is 7x higher than in Paris. The Ca and Hi projects show how demented the whole enterprise has become. Wake up America!
Floccina
Nov 12 2018 at 4:13pm
How about instead of the LA to San Fransisco rail, we work on try to speeding up the airports. It seems absurd to me that they tell you to be there 2 hours early for 1 hour flight.
Comments are closed.