The inanity of airport connectors

Despite my issues with how new transit projects are implemented in America today, I’m generally happy to see them built. Even though they’re flawed, heavily-subsidized government creations, they make upzoning more palatable and can later be sold off and privately managed. There’s a lot I’d do differently, but on net I think most new transit projects are a step, however imperfect, in the direction of market urbanism. But there’s at least one form of transit that I can almost never get behind: the airport connector.

The airport connector is a special beast of a rail-based transit system that’s a relatively recent phenomenon outside of transit-dense regions like Western Europe and Japan. So manifestly wasteful that it generates more animosity towards mass transit than it does riders, it’s a project that only politicians and unions could love. Unlike more integrated networks where the airport is just one station on an otherwise viable route (like Philadelphia’s Airport Line or DC’s proposed Silver Line), airport connectors generally serve only the airport and one local hub. With no purpose other than to get people in and out of the airport, they provide neither ancillary transit benefits nor TOD opportunities.  Oftentimes they don’t even reach downtown, acting instead like glorified park-and-rides.

The most egregious example in the US would have to be BART’s proposed Oakland Airport Connector. The rail line will extend for a little more than three miles, replacing what is now a bus routes.  The $3 fare will double, along with the half billion dollars that it will cost the government. Like the current bus route, it will only connect Oakland’s airport to the nearest BART station with no intermediate stops. It’s opposed by transit activists, who would rather convert the bus into a dedicated BRT lane and spend the rest of the half billion elsewhere. Its support seems to lie entirely with unions eager for the work, and one commissioner actually called it “too big to fail”…as an endorsement of the project! Even the feds knew better and took back their $70 million in funding, citing its impact (or lack thereof) on Oakland’s low-income population. Still, the city is pressing froward with the guaranteed boondoggle and is continuing to allocate funds for it, though transit activists vow to keep fighting it.

"Rhode Island can't build its way out of traffic congestion," said one parking garage-cum-rail station backer

Oakland is hardly alone, with the Rhode Island DOT recently reaching a deal on its $267 million “Interlink” project, which entails building a station at the airport on an existing line, along with a commuter parking garage and a rental car facility. The station is only expected to see six trains a day initially, which is probably for the best since Providence’s T. F. Green Airport isn’t exactly O’Hare. No word on whether any additional density is being allowed around the new station, but something tells me the answer is no.  New York’s five-destination Stewart Airport could also snag a rail line, as might a couple of other airports if Obama’s high-speed rail plans ever see the light of day – and I doubt they will, but that won’t stop consultants from being paid to consider them.

America, however, has no monopoly on ineffective government, and like all forms of excess, the Chinese have perfected the genre with their maglev airport connector in Shanghai. It literally levitates above a track through a system of magnets, and with only air friction it can achieve speeds of up to 268 mph. The Chinese government can acquire land at zero cost and labor is cheap, so it only cost $1.3 billion to build, but it makes up for its low cost in utter uselessness: in addition to being too expensive for ordinary Chinese to ride, it doesn’t even take you downtown – “it virtually goes nowhere,” as the Asia Times puts it. The project can be seen in the broader context of Shanghai’s misguided rivalry with Hong Kong, where it seeks to copy the trappings of the wealthy city-state while avoiding the attendant economic liberalization.

There may be a limited place for short airport connectors in large, transit-rich cities like New York City, but many of the projects turn out to be far too expensive for the limited service that they provide. They are often a sort of cargo cult urbanism that seeks to emulate the frills of good transit systems isn’t willing to make the hard decisions necessary to actually build a robust network and allow the density to fill it. In the case of the the Providence airport, lawmakers said they hoped the station would attract international service to the currently domestic-only airport – as if Providence can acquire the amenities of a big city without allowing itself to become one. Airport connectors instead are often little more than highly inefficient subsidies to the airline industry, wealthy frequent fliers, and construction unions – which, now that I think about it, might explain why legislators love them so much.

  • mogreen

    so i take it you’re not a fan of airtrain in nyc?

  • Brandon

    To tell you the truth, I wish SFO had an airport connector rather than the current wye-based airport stop. The number of people going there is pretty decent (I dont know if this is the story in Oakland), but its not worth running full size BART trains there. If BART ever goes down the Peninsula further it will be an even bigger problem since having to stop at the airport will affect more than service to Millbrae. (they sit at the airport for awhile, going to/from Millbrae after 7 when the dedicated trains stop is a pain)

  • Brian

    The current Oakland airport BART bus connector is just fine; I’ve ridden it many times and it runs quickly, affordably, and frequently along uncrowded streets between the station and the airport. There is no need for half a billion dollars that will literally not improve current service at all.

    I wish the SFO connector to BART had never been built. It drains CalTrain (the actual local transit) of funding, cost nearly $2 billion, and runs slower than both the old CalTrain shuttle bus and the MUNI express bus from the airport to downtown. That’s $2 billion that could have built new lines with new equipment throughout the region and made San Francisco a much nicer place to live car-free. The old express bus ran just as often as BART, cost about the same, and connected to BART, MUNI underground, cable cars, and light rail with a one block walk. I’ve used all those connections (cable cars are a little expensive but they make sense for some of my trips).

    In Chicago, the L goes nicely to O’Hare, but the many stations along the way are park-and-ride abominations and the train runs in the center of a freeway. The MDW connection in Chicago is better. At least both airports have transit. I found that in New York City, of all places, I can’t get to EWK or LGA without insane delays and connections through awful local busses. JFK doesn’t seem much better. That seems like a bigger hole in regional transit than an eighth and ninth NJ tunnel (they should just make one Lincoln bore a dedicated 24-hour BRT route and save $15 billion).

    Salt Lake City is building a better airport light rail connector. Local studies indicate that the airport is the destination of 20% of all car traffic on the most crowded local highways. The intermediate stops are good candidates for better development and the mayor is working hard on legalizing it but local property owners are pushing back every step of the way. Mostly it’s institutional users committed to a car-oriented strip and a minority of the small businesses holding things up. When the route is done, it’ll make my life easier since I travel to the airport.

    Mexico City’s airport has long had a station on the subway. I use it all the time but most tourists and locals prefer to take a taxi, even though traffic can make the taxi ride take longer than the train. The subway does prohibit connecting with more than a carry-on but it’s Mexico so there’s no real enforcement. Recently a new terminal opened up with a boondoggle expensive connector train between terminals. The new train requires you to have a boarding pass (so you have to print one out at home) but charges no alternative fare, so if you’re hoping to check in at the new terminal or pay to ride, you’re just out of luck — hire a taxi between terminals for US$5. I’ve looked up the nearest subway station to the new terminal but I haven’t tried walking there yet; it looks like a freeway-only entrance. It’s a one-km walk.

    Still, I love the access and find Mexico City’s connector near ideal. The lines that run around the airport are fully integrated into regular lines through the city with no artificial “connector” portion.

  • JoelT

    I don’t know if it’s fair to include the new RI airport link here. It’s part of one of the “integrated networks where the airport is just one station on an otherwise viable route” – Boston’s commuter rail.

  • Stephen Smith

    It might be well-integrated, but it’s not a very large airport, and it looks like most of the money is actually being spent on the parking garage and a rental car facility (?!). And it will only see 6 trains a day initially, with 8 or 10 (I’ve seen both figures) max later on. And spending a quarter of a million dollars just to build a station and parking garage? It doesn’t even look like they’re laying any track!

  • Stephen Smith

    NYC and its airports are pretty big, so I guess it’s not the end of the world, but if I could’ve allocated the money, I definitely would have just given them BRT lanes and spent the rest of the billions elsewhere.

  • Pingback: Weekly Roundup « The Gondola Project()

  • Aaron Brown


    Agree with your points about NYC – I never understood why LGA and JFK weren’t connected to the subway, especially since they run fairly close by (although the “SkyTrain” makes the JFK trip a bit easier). Given the cost of building a connection today, I think the best option is to convert the M60 bus into a dedicated BRT line. It not only connects Manhattan with LGA, but also goes through a number of densely populated neighborhoods in Queens and Harlem, so I’d guess that it’d be well-utilized by non-travelers as well.

    On Chicago, you’re right that the Blue Line to O’Hare is pretty ugly as it gets closer to the airport (once it start running in the highway median), but most of the stops along the way are in fairly densely-populated neighborhoods where residents regularly use mass transit (think Jefferson Park, Logan Square, Wicker Park). Given that, the extension to O’Hare made a lot of sense, and I see a lot of people using it for travel.

    That said, Mayor Daley’s push to add an expensive express service to O’Hare should be put down immediately. The cost to build would be astronomical (because of the dense neighborhoods mentioned above), the service would likely delay regular riders on the current Blue Line, and there’s already a fairly quick way to get to the airport via public transit (~40min from downtown).

  • Benjamin Hemric

    While I don’t have time to comment on this interesting topic, I do want to ask a question about what appears to a techical problem with the website — most recently with two comments, by the webmaster Stephen Smith, that appear to be missing from this thread (if I’m using the website correctly, that is). (I believe I’ve noticed the same problem with some recent comments by other commenters too.)

    If I understand the Market Urbanism website correctly, on the right side of the home page there is an “unlabled” list of commenters that shows the most recent comments that have been made on the website and the threads that they belong to. (This list is located just below another list that is of links to other websites that have linked to Market Urbanism articles.) In the past, when one clicked the name of the commenter, one was lead (at least at one time) directly to the person’s comment in the thread. (Somewhere along the way, though, this seems to be changed so that one was just linked to the thread itself — but not to any particular comment in the thread.)

    If this is a correct description of the way the site works (and if I’m using the website correctly), there appear to be two comments in this thread by Stephen Smith (from a few days ago) that are “missing” from the thread. To put it slightly differently, when I put the cursor over Stephen Smith’s name in the list on the right side of the website’s home page, I get to see the first line of what appears to be a comment by Stephen Smith in the thread. But when I click on the link (or click directly into the article), I don’t see any comments by Stephen Smith at all.

    Is this a problem with the website, or perhaps I’m doing something wrong? (By the way, this seems to happen from various PCs at various locations.)

    Benjamin Hemric
    Sun., October 17, 2010, 4:35 P.M.

  • Stephen Smith

    Yeah, I’ve noticed that too – my comments disappear occasionally. I’m actually not the webmaster – it’s Adam’s site, although I’ve been posting the vast majority of the articles within the last few months. Sometimes he comments though under the name MarketUrbanism, whereas I usually use my own name. I’ll send him an email though in case he doesn’t see this…neither of us are professional webmasters so I’m not sure what he can do about it, but I’ll let him know.

  • West Coaster

    Seattle just replaced its service between downtown and SeaTac with a light rail connection. Travel times are now twice as long. The light rail line meanders over kingdom come, and is mainly a revitalization effort for the Rainier Valley area of low-income people. For a connector about five to ten miles south of downtown, the line first goes a mile east and then a couple of miles west. Save us from more “connectors” like this!

  • David Jackson

    “n the feds knew better and took back their $70 billion in funding, citing its impact ”

    I think you mean $70 million.

  • Stephen Smith

    Haha, thanks. That would have been one hell of a rail line!

  • Anonymous

    The original plan for BART to SFO was to have a regular station under the main parking structure. The foundations for the parking structure have gaps for the BART tunnel to enter, and this modification cost quite a bit of money in the 1960s. The current compromise solution means that BART riders have to change from one train to another, which limits ridership.

    The Oakland connector idea is awful, though if it could have been done for $50million instead of $500million, it might be worthwhile. There’s nothing worth stopping at between OAK and the Coliseum BART station, so that part of the plan is ok, but since it will require a transfer from one train to another, it’s not much of an improvement over transferring from a train to a bus. Not having to step outside the station at Coliseum BART has its attractions, though – that’s one of the worse neighborhoods to have a BART station.

  • Worker201

    The Seattle system might seem odd on its own, but I think it works in the context of the larger Sound Transit network, including both current routes and planned developments. I took the train from Tukwila to downtown a number of times (before the airport link was finished) and I found it a pleasant and comfortable experience. And there is already tons of new development along the MLK part of the route, so I guess it is working as a revitalization project. I’d also like to point out that the voters in King County are overwhelmingly for public transportation, and voted in favor of a tax levy to fund new rail development.

  • Joseph Alacchi

    We in Montreal are having a similar debate about our airport connector. The frequency required for a good service (Montreal is a transit city) require that dedicated tracks be built. Furthermore, there are two stations downtown to connect to. Gare Lucien-L’Allier is the closest to the airport but is far from downtown hotels. Gare Centrale would entail a less direct route but is closer to downtown hotels and is also where all the intercity trains run to. The airport (a private company) wants a direct train line from the airport to Gare Centrale. The AMT (regional rail operator) wants to merge the airport project with the existing Vaudreuil-Hudson train line to create an all-day frequent commuter train line serving both the airport and Montreal’s western suburbs and linking them to Gare Lucien-L’Allier. The city of Montreal would rather see a route running between the airport and Gare Centrale with many intermediate stops serving the southwestern, poorest part of the city.

  • Andrew

    I ride the Blue Line daily: Damen-Rosemont 6x a week and Damen-Jackson 4x a week, all for work purposes. It’s certainly true that in terms of aesthetics, the highway median stations in Chicago leave something to be desired, compared to the elevated or subway lines in the rest of the city. Clearly these stations present barriers to walk-in pedestrian traffic and they don’t do very much to knit rapid transit close to the neighborhoods they serve, and cars rushing by in the dead of winter is not the most appealing image.

    Without regards to aesthetics, though, I’d have to say the Blue Line does a pretty good job with what it has. The key here is that much employment in metropolitan Chicago has been shifting in a northwestern direction towards the affluent suburbs out by and past O’Hare (think Schaumburg, Hoffman Estates, etc.) so while the highway-median stations may look ugly, they serve a decent amount of customers traveling to the northwest areas. In terms of raw ridership, the Jefferson Park, Rosemont and Cumberland stations — all highway-median stations — all have monthly ridership, as of Aug ’10, higher than all Orange Line stations exclusive of Midway; and Irving Park (another median station) sees more riders than all stations aside from Midway and Pulaski. Raw ridership obviously only tells you numbers, but in terms of being “better” than the Orange Line, the one to Midway, as mentioned above, I don’t know if it’s a really a fair conclusion to draw.

    Clearly, there are a number of external factors. The Orange Line opened in 1993 and the O’Hare extension (Jefferson Park to the airport) opened in 1984, so the southwest rapid transit line is still relatively new to the city. Both are situated in Chicago’s tracts of less-dense single-family housing, as opposed to the apartment buildings and high-rises we see closer to the downtown core and along the lakefront. I suspect that if the preexisting rail trackage did not exist for the Orange Line, that line would have been run down the median of the Stevenson Expressway (I-55) as well.

  • Alex2000

    Your information on the Shanghai Maglev is misleading.

    Sure, it only runs 19 miles, from the airport to not-downtown *now*, but that was phase 1! That was never the intended full length. Phase 2, which was greenlit in march, will extend the track to 124 miles, further into Shanghai to the Shanghai-South Railway Station, and southwest to neighbouring Hangzhou city (pop 3.5 million!).

    The train will also stop at Hangzhou airport along the way, meaning that travelers will be able to go the 34 miles from Shanghai airport to Hangzhou airport in 15 minutes.

  • Pingback: News & Notes — Greater City: Providence()

  • Cities21

    Oakland Airport Connector: Personal Rapid Transit beats expensive Automated People Mover:
    More stations, faster, cheaper, more economic development, more jobs, not “blingfrastructure for the rich” – BART Director Radulovich, redevelops parking, higher BART ridership, helps OAK market position, enhances AC Transit bus service, not an eyesore, less construction hassle, and more extendible.


  • Anonymous

    Totally untrue. First of all, the line is 15.6 miles long, while the most direct route from downtown to the airport would be about 14 miles. That’s 1.6 more miles to serve tens of thousands of extra commuters. And it’s not even a dedicated airport connector; it has to go past the airport to get to population centers to the south anyways! It does not take close to twice as long as the bus that went from downtown to the airport. The bus took 32 minutes and didn’t stop between those two places, while the light rail takes 36 minutes while stopping a dozen times along the way. 4 extra minutes to stop in 8 or so extra neighborhoods is a great deal!

  • Stephen Smith

    Okay, but is it worth the $267 million cost of this one station? Even if 1,000 people got on at this station every day for the next 20 years and there were no fares or operating costs, the subsidy would still be $36 per ride!

  • thunda

    Just a correction about Philadelphia’s Airport Line, formerly the R1, before the recent rebranding of rail services:

    The Airport route isn’t actually a “station on an otherwise viable line,” but a ~4.5 mile spur off of the Northeast Corridor to serve PHL. Much of the trackage already existed, so the investment in the line in the 1980s was relatively low, thought ridership isn’t spectacular (

  • Stephen Smith

    I appreciate the comment, but what the hell was that YouTube link??

  • Jaded

    I like th SFO connector. Lucky for me I live on the direct BART line. But it meets an important criteria. You land in the airport and you don’t need to go outside. So it is pretty seamless. At least as much so as driving, parking off-site or in long term parking.

    The PAC on the other hand, makes you go outside. And walk pretty far. I voted for the 2000 proposal because it improved the neighborhood with intermediate stops. There is a huge development with Walmart and a lot of other stuff near the airport, the OAC should be serving this area as well. I am really disappointed they are not holding up their end of the deal by delivering the project that was pitched on the ballot.

    The worst thing? The station is stopping at Terminal 1, and now where near Terminal 2 or the planned #3. To go from BART to the security line is going to take 30 minutes if you need to go to Terminal 2. How is this an improvement. And it’ll cost me $6 each way. I could split a cab from BART to the airport at that price. This isn’t a good use of my $500M.

  • Sam Russell

    Airport connectors are a lifesaver. There is nothing worse than arriving tired in a city and having to wait at a curb for some crowded bus, wondering if it’s the right bus, and being stuck standing up in traffic. Or taking a taxi and paying 50 to 100 dollars. God forbid you could take a comfortable, hassle-free train which would take you ideally downtown, or at worst to a connector station. My biggest complaint is when the airport service doesn’t run frequently enough, or is badly coordinated. Newark Airport’s monorail connection to the Northeast Corridor is a revolution that finally makes the airport a contender to NY’s others, but the schedule on the NE Corridor leaves regular gaps of 40 minutes between trains, which is too long to wait and makes the cab look tempting. And at 15 dollars, the trip is far more expensive than it should be. These systems should be subsidized as much as possible, to help ordinary travellers who can’t afford to spend more on a cab than they did on their flight, and generally to get those damn taxis off the road.

  • Stephen Smith

    Of course there are worse things that arriving tired in a city and having to pay for a cab. Like, say, living 24/7 in an auto-centric world without a car or access to transit, which is bound to happen more often when scare public funds are wasted on projects like airport connectors.

  • Anon256

    The NJT #62 bus gets from Newark Airport to Newark Penn Station for $1.50, running every 15 minutes all day long, and PATH gets from there to NYC for $1.75. This is much more useful than the costly and infrequent monorail-to-commuter rail option. As for subsidising the latter, when transit systems can’t even afford to keep running the basic services that people of all incomes need to get to work and school, there’s no way they can justify the huge cost of providing a gold-plated transit experience for a trip made only infrequently and only by people who can afford to spend hundreds of dollars on a flight. A potential compromise would be to fund the airport connector through a surcharge on plane tickets to and from the airport in question, but as an “ordinary traveller” I’m quite happy to keep my cheaper airfare and ride the #62 bus.

  • Mike Orr

    “Travel times are now twice as long.”

    The trip increased from 28 minutes to 37 minutes. It’s also 33-50% more frequent until 9:30pm, and 100% more frequent after that. So the wait+travel time is a wash, and people hate waiting even more than they hate slow trips.

    That “28 minutes” assumes no traffic or accidents on the freeway. The train also created new express service to the airport from southeast Seattle neighborhoods, and the next phase will provide that to northeast Seattle and the suburbs. The time loss travelling east-west is around three minutes, not the end of the world, and it allowed one neighborhood (Beacon Hill) to get a station that would otherwise have no station.

  • Anonymous

    I landed in Newark once returning from vacation and it took me forever to wait for the *&#! bus 62 or the other one that goes to the Upper East Side. So I like the airbus. My question on airport monorails is why don’t they make it part of a network of public transit trains? Wouldn’t that encourage public transit usage in the sprawling areas where the airports reside? Or for that matter, why not extend a regular train route to the airports? Of course, one disadvantage is the crowded trains you have to take, with all your luggage taking up space. That’s what you have to do if you take the E train to JFK in NYC.

  • Montreal Limousine

    Airport connectors are poorly integrated with existing transit systems and usually have only two stops—a local hub and the airport—they offer little or no secondary benefits or opportunities for transit-oriented development.

  • Limo Montreal

    Montreal Limousine services have figured out  on how to serve you better and decided to offer you the lowest possible price than any other company. 

  • Montreal Limo

    Montreal Limo Services carries a fine fleet of limousines that are sure to meet your Limousine Service needs wether they are sedans or stretch Limos.

  • Montreal Limo

    If you live in Montreal and you want to ride in style and be treated like loyalty then it will be best if you hire the airport limo service.