Boosting Transit Ridership With New Stations, Not New Track

Boston's new Orange Line station in Somerville is a great example of how older cities can boost transit ridership inexpensively with new stations in strategic locations. Image: MBTA
Assembly Station in Somerville, outside Boston, is a great example of how older transit systems can draw more riders with new stations in strategic locations. Image: MBTA

Yonah Freemark at the Transport Politic calls them infill stations: new transit stops built in gaps along existing rail lines. Current examples include Assembly Station just outside Boston in Somerville, DC’s NoMa Station, and the West Dublin/Pleasanton BART station.

Infill stations are a pretty brilliant method to get the most out of older rail systems without spending very much, Freemark says. He’d like to see more cities adopt the strategy:

The advantages of infill stations result from the fact that people are simply more likely to use transit when they’re closer to it — and from the fact that the older transit systems in many cities have widely spaced stations that are underserving potentially significant markets. Erick Guerra and Robert Cervero, affiliated with the University of California-Berkeley, have demonstrated that people living or working within a quarter mile of a transit station produce about twice as many transit rides as people living or working more than half a mile away. In other words, with fewer stations on a line, the number of people willing to use public transportation as a whole is likely reduced.

Assembly Station, which has been in the works for several years, promises significant benefits — 5,000 future daily riders taking advantage of a 10-minute ride to the region’s central business district, at a construction cost of about $30 million. The station fits in the 1.3-mile gap between two existing stations and is the first new stop built along Boston’s T rapid transit network in 26 years. When combined with the $1.7 billion Green Line light rail extension planned for opening later this decade, 85 percent of Somerville’s residents will live within walking distance of rapid transit, up from just 15 percent today.

The cost-per-rider comparison between the two Somerville projects is indicative of the value offered by infill stations: While Assembly Station cost about $6,000 per rider served, the Green Line Extension will cost $38,000 per rider served — six times more. Both projects will provide benefits, but the cost-effectiveness of infill stations in terms of attracting riders is clear. While infill stations will reduce transit speeds to some extent, within reason the number of new riders they attract will more than make up for the change.

Elsewhere on the Network today: Strong Towns comments on perverse transportation engineering standards that create dangerous streets in the name of “safety.” Systemic Failure says Caltrain will have to choose between bikes and bathrooms in its new electrified trains, and it should go with the former. And Beyond DC shares a quote that gets to the heart of the reason protected bike infrastructure is so important.

0 thoughts on Boosting Transit Ridership With New Stations, Not New Track

  1. Except the MBTA is dragging their feet on the orange line expansion south – which has been on the books since the 1940s. Adding just one station south of Forest Hills (there’s space for both tracks and a station about a mile south in Roslindale Village) would allow them to reconfigure the 9 bus lines that serve a single 1 mile stretch of road and potentially save them about $1million in operating costs a year. I believe they haven’t bothered to do a single station study because it’s always been tied into a bigger expansion that involves two different subway lines, there is already some weekday commuter rail service (although expensive) – and there hasn’t been any pressure from the city – which would consider this expansion crucial to economic development in the southern end of the city.

  2. Also in the Chicago area, preliminary plans have been made for an Asbury stop on the Yellow Llne, which for the first time in 66 years would put parts of West Rogers Park within reasonable distance of rapid transit and, according to Evanston’s feasibility study, might increase ridership on the line by as much as 45%. Not that anyone now living will actually see that station, what with the city and state broke and the Party of No Transit holding the purse strings in Washington for the foreseeable future.

  3. The challenge, I think, to adding more infill stations on Metra’s UP-North line is that the trains currently run on diesel power, and that means they are slow to accelerate. Now, if this line were electrified, like the Metra Electric line to the southeast, it would have great potential for more infill stations. Adding a third track, which the new bridges will allow space for, would allow local & express trains to co-exist on the UP-N line.

    All of these same suggestions and opportunity exist for other Metra ROWs–which, like in Boston, exist in areas of the city where light-rail doesn’t serve.

  4. While West Dublin/Pleasanton is accomplishing the goal of more riders, most of the points this article mentions do not apply to it. It was built and is mostly functioning as an expanded parking lot for the Dublin/Pleasanton station at the end of the line. I’m not sure there is actually any housing within a quarter mile.

    For BART the rock star station of the entire system is actually an infill station–albeit a very old one–Embarcadero station in San Francisco, the first station after crossing the Transbay Tunnel, was not part of the original plan and opened a year or two after all the other stations. It consistently has the highest ridership of all stations.

    Of course, maybe we would still have a worthwhile ferry system if that BART station wasn’t on the Ferry Building’s doorstep.

    Personally, I’d like to see the mythic 30th Street BART station in Sam Francisco. $500,000,000 seems cheap compared with the Central Subway boondoggle.

  5. Infill stations in the right places are definitely a good idea. The key thing is to make sure new stations don’t unduly slow down the service.

  6. Although a bad idea. That is going to happen in the not too distant future.

    A station instead at Devon would have made more sense with perhaps trading out the Lunt stop for a new one at Howard.

    The station at Peterson will literally have a 1/4 semi circle for 1/2 mile in the southwest direction populated by dead people.

    A station at Devon would have had potential to still serve S&C still serve RP / Edgewater and WRP. Also the surrounding area houses far more people, a major employer in S&C and a commercial strip along Devon that could have benefitted from a transit link.

    This along with under utilized lots just south of Devon and also east of track that could have provided an opportunity for TOD.

    No Instead in this town we place a transit link on an auto-sewer (Peterson) surrounded by dead people (Rosehill) and low density autocentric pedestrian hostile Peterson.

    A terrible terrible waste….par for the course in the town of Chicago.

    A town that wastes more transit assets than most cities even have.

  7. On the subject of Chicago, what about a station at Ashland on the Southwest Service? It’s over a mile from the nearest L stop and could connect to the Ashland BRT when/if that happens.

  8. They should have built Addison with the brown line connection when the land was available 10 years ago.

  9. I used to think infill stations were critical also, but my thinking has changed.
    As mentioned already, this solution has a natural and important flaw: it sacrifices speed for ridership. While adding infill stations sounds great, the reality is while each new stop may boost ridership 5% (let’s say), it degrades the service for the 95% who were already using it. Play this solution out: if it’s such a great idea, we should do it everywhere – degrading current service with each new addition. (In the East Bay, you could make a strong case that every inter-station neighborhood between Lake Merritt and Union City could support an infill BART station. That’s 7 new stations! How would service hold up?)
    I believe the ideal transit system has to include circulators that collect riders from these between-station neighborhoods and deliver them to stations. (And I don’t mean separate systems like AC transit that have schedules and routes independent from BART, as well as separate governing boards, financial considerations, etc.; I mean BART-owned and -operated circulators that serve to widen the effective radius of their stations.)

    Say we have a line with stations called A, C & E, with neighborhoods B & D between. Instead of adding a station in neighborhood B, you have a circulator from station A serving that half of neighborhood B and another from station C serving the other half. Depending on local geography, the same station C collector also collects riders in neighborhood D, or there may be a second station C collector working that side.

    This would achieve increased access without bogging down the main lines. In my opinion, the best mode for this service would be minibuses, as used throughout Asia; fast & nimble. Picture 1-2 Sprinter vans based at each station. [Alternatively, in areas where they make sense on their ownb merits, this function could be served by streetcars. Though streetcars would obviate the main point – low cost – of the post, they do have a whole other range of thoroughly-discussed benefits.]

  10. You could think about it that way, or you could look at the dense apartment complexes in the NW quadrant, populated by a diverse mix of immigrants with low car-ownership rates.

    If Chicago’s BRT plans work out, Peterson is a natural candidate for a fast east-west bus line. Even the current configuration could support decent express bus service with signal priority.

  11. It’s really sad that it’s been 26 years since the MBTA opened a new rail station. I wonder how many were eliminated in that same time period… oh, the entire Arborway streetcar line!

  12. Not every point between existing stations would need a new stop. The San Antonio District in East Oakland would be a prime spot for an infill station. The neighborhood is dense and walkable.

  13. Except in this town even the Ashland BRT looks like it will be watered down. The city has already backed away from the original plan to include Western ave BRT.

    And Peterson was eliminated as a route in at least one study I am aware of for BRT. Do not have the link off hand.

    Better to place assets where demand already is in place versus relying on at best temporally distant speculative proposals like the Western BRt and the eliminated Peterson BRT.

  14. Yes and Addison or an IP station would seem to make sense especially if Metra could ever move from the 19th century to the 21ts century and electrify the UPN

  15. What demand are you referring to? Metra service patterns are only useful for people working downtown, and there aren’t a whole lot of those people living along the UP-North line. Ravenswood ridership stagnated for years until the neighborhood gentrified and filled up with downtown workers.

    I don’t see why factory workers at S&C would take hourly Metra service instead of driving. They’re probably not living on the North Shore, either, so where would they go on the train?

  16. The best candidate for infill station in New York is on the new #7 extension: 41st and 10th Ave. I believe this was in the plans originally, but was eliminated due to costs.

    As one can easily see from the subway map, this leaves a large area just west of Times Square without subway service.

  17. Sure that is why the UPN Rogers Park stop sees no I mean no boardings and alightings….OK.

    I mean I know when I get on / off at Lunt I am the only one.

    And there seem to be plenty of S&C workers as well especially near morning and evening rush when service at lunt runs quite more frequently than once per hour and does so for a few hours….

    But what do I know….you seem to know it all.

    Oh and of course the area particularly at least this point south of Devon east to clark well I not to sure you are up to date on your demographics; and that is beginning to spread north of Devon.

    That being said….I’m pretty sure the dead in Rosehill will not be demanding much service at all; perhaps you have evidence to the contrary

  18. Sure, plenty of downtown workers live in Andersonville – south of Peterson. Why would you want to move it further north?

    My take is that no matter where you put the station, it will take a few years before the area fills in with additional development and with downtown commuters. I think Peterson is a stronger location BECAUSE Peterson itself is a good feeder route for the entire Far North Side, so you don’t have to wait for that development to fill in before seeing ridership. Lunt has terrible accessibility.

  19. People are most likely to use transit when their DESTINATIONS are near high-capacity, rapid transit lines. If you live right on top of a station but you shop at malls built in the middle of nowhere and work in an industrial park at a freeway interchange, you will never take transit.

    It is not housing that must be massed right next to stations, it is businesses, shops and offices (which also generate 2 to 5 times as many trips per square foot as housing). If you really want to increase ridership, simply convert your transit station into a mall with a small amount of underground parking. You don’t even need to spend anything, just set up a public-private partnership and have a promoter pay for all the changes to the station.

    High-density housing can be a bit further out, a 5- or 10-minute walk or bus ride away.

    Adding transit stations can be a good idea sometimes, but you have to be wary. The more stops transit has, the slower it gets. If it gets too slow, the farther reaches of the line become too far from other destinations and people may go back to cars.

  20. Stations on BART are much closer together in North Oakland/Berkeley and in San Francisco than on the Fremont line south of Lake Merritt. So there’s room within BART’s operating model to add selected infill stations.

  21. My point was that the demographics between Devon and Peterson esp east of the tracts is more favorable than you seem to believe and this is beginning to change even more rapidly.

    The census tracts surrounding either Devon or Peterson share two of the same tracts; the unshared tracts show more density in those surrounding a Devon site.

    I agree Lunt has terrible accessibility. I would give up Lunt…an inconvenience for me if a new stop was placed at Devon over Peterson and the Lunt stop moved to Howard where there would be better intergration with the red and yellow line.

    Additionally I really wish the UPN would consider running increased frequencies during the day at least 2 ( preferably 3) per hour and continue such service on the weekends perhaps cutting it to 1-2 per hour on sat, sun at least in the inner portions of Metra (thinking like wilmette south to OTC)

    That combined with some infill stations perhaps 2 – more for instance (foster vs peterson )or (IP or addison ) and likely Belmont could do wonders for transit use and further development of the northside.

    It could even eliminate the need for BRT north of Clyblourn.

    Toronto has experimented with increasing frequency on commuter rail and seen significant postive effect

  22. I don’t think the bias between origination and destination is that large.

    Mass-scale transit networks will always have last-mile issues. When these issues are greatly reduced on either end of the trip, it increases overall attractiveness of the system as long as the other end has some solution (like a local light-rail or shuttle buses) if it is not within walking distance.

    So if there one lives in front of a station and works in an office park, but the office park is connected with shuttle vans on high frequency patterns to other train station, the system is convenient.

    I agree, though, that since people almost always keep a car on their home garages or on street in front of their residences, having a totally disconnected origination point of a commute is less damaging to transit potential usage than having a totally disconnected destination.

  23. Historically, successful TODs like streetcar suburbs and present-day train station in Tokyo have always grouped businesses near the transit lines with housing a bit further out. Take an office, a regular office has perhaps one employee per 150-200 square feet of floor area, whereas in the US, there is one person per 800 square feet of space in housing, even in urban areas, that average may be down to 400, but it’s still 2-3 times less than the space each office worker gets. For shops, which have workers and clients, it’s far more than that.

    At work, I have access to the Trip Generation handbook, based on empirical data, and I can assure you that businesses do generate quite a lot more trips than housing for the same area.

    If most of a region’s main attractions (offices, stores, etc…) are within walking distance to a rapid transit system, you’ve just eliminated half of the last mile problem. Bus feeder lines also have quite low capacity. A subway can carry 20 000 people per hour per direction and more, a bus line with regular buses every 5 minutes can carry about 600-700. So connecting a major shopping area to a station by a feeder bus line can result in congestion if transit were the dominant mode to access the shopping area.

    There is also another reason why concentrating commercial developments at stations make sense…

    Think of malls at freeway interchanges that happen all over the place in North America, why are they there? Of course, the freeway allows them to have a larger pool of clients because of high speed, but it’s more than that, since commuters come mostly from the freeway and get out at a few exits, there is a lot of traffic passing through these areas. So placing malls there make a lot of sense because they have the potential for a lot of “pass-by” clients who are passing through the area anyway and for whom it’s quite easy to stop and shop on the way to work or the way back home. For these customers, this is very efficient, they don’t need to go home then go shopping, they can go shopping on the way home, two birds, one stone. For the store owners, that is a strong customer base that is easy to tap.

    Same thing for transit. The last mile problem can in fact be an opportunity more than a problem. Since a lot of people will get out at the station, if you put stores at the station, then you have made it a lot more convenient for transit users, for whom shops are now “on the way”. If they need to buy a pint of milk, they can easily do it at the station, or maybe get a few groceries while they’re waiting for the bus to take them home (if they don’t live within walking distance of the station). It makes a waiting time for a connection useful instead of wasted time. Also, feeder bus lines will play a double role, sure they will bring people too far from the station to it to take rapid transit, but as stores are concentrated at the station, they will also double in use as lines to take people to the local shopping area. The result will be higher frequency than shuttle services would have alone and higher capacity as you have many lines converging on the shopping area, not just one shuttle line.

  24. Your calculation assumes infill station at a place where 95% of ridership travels through.

    Even on heavily directional-terminal patterned commuter lines, I think it is a stretch to think of any rail line that, at any given point, has 95% of its regular ridership going through a single sector (maybe special event trains are the exception). Usually, you have people boarding and alighting on different stations, even if the train ends its course at a major terminal like GCT or PATH WTC.

    Obviously, the effect exists and should be evaluated. Sometimes rail systems can cope with a handful of extra infill stations with a concurrent revamp that brings better rolling stock (more focus on acceleration, less focus on top speed) and other scheduling optimization strategies.

Leave a Reply

Your email address will not be published. Required fields are marked *


Boosting Transit Ridership With New Stations, Not New Track

Assembly Station in Somerville, outside Boston, is a great example of how older transit systems can draw more riders with new stations in strategic locations. Image: MBTA Yonah Freemark at the Transport Politic calls them infill stations: new transit stops built in gaps along existing rail lines. Current examples include Assembly Station just outside Boston […]

Parking Madness 2019 Round 1: Oakland vs. Atlanta

We’re on to our second round one match in this year’s Parking Madness competition — comparing different cities that have repaired their parking craters with useful infill development: more city! This competition highlights a really hopeful trend underway in cities; Parking craters all over the country are disappearing. In our first matchup, Minneapolis knocked off a […]
At Bourg-la-Reine, outside Paris, the rail station is surrounded by dense, mixed-use development and walkable streets. Image: Google Maps

What American Commuter Rail Can Learn From Paris, Part 2

In Europe it's common for regional rail systems to get ridership comparable to that of the subway in the central city. But in America, this is unheard of. One reason for the discrepancy is land use: American commuter rail stations are typically surrounded by parking, while in the Paris region you see a different pattern with ample development next to suburban train stations.

Despite Problems, Boston’s MBTA Should Continue to Expand

2nd inbound orange line track to Boston isn’t even cleared of snow between Oak Grove & Malden! #MBTApocalypse #MBTA — Cory Thomas (@coryt3) February 11, 2015 Cross posted from the Frontier Group. As you may have heard, we’ve been experiencing a few public transportation problems here in Boston of late. Record-smashing snowfall, coupled with extreme cold […]