61 Replies to “April 2011 Link Ridership”

  1. Wow, quite a jump in Sunday ridership. Probably because of that American Flag.

  2. How does April 2011 compare to April 2010? What’s the y-o-y increase?

  3. This is an interesting editorial comment from Martin:

    “This is actually up from March”

    Why the “actually”? Do you find the April numbers to be so low that it is surprising that the March numbers were “actually” lower?

    Or, what?

    And, I’m with you, Martin. I won’t feed the trolls.

    1. I love you Norman!

      I expect he’s only surprised because March was up 18% over last year, and gains that large in just one year were, at the time, interpreted as a fluke (perhaps driven by seasonal event). He was assuming ridership gains would drop back to a slow-and-steady pace in the next month.

      1. “Actually”, January ridership was up 31%, February up 23%; March up 18% and April up 9%.

        Notice any trend there?

        I love you too, Lack!

      2. I come up with a little less than 9% for weekday boardings:

        [ (21,870 – 20,129) / 20,129 ] X 100 = 8.65%

        Last year there was an 11% jump from March to April. This year the increase is only 2.6% which continues the trend indicating that ridership will be virtually the same year over year by this summer. I remember an article in the PI back around the time Link opened that said barring major changes, like adding a station or truncating bus routes, after about a year and a half ridership on all light rail tends to stabilize. I’ve been unable to turn up that article despite several search attempts.

      3. Seems about right. We’ll stabilize in the next year or so to a growth rate that roughly mirrors economic & population growth. It’ll probably hover in the 5% range for quite a while before settling down – my personal seat-of-pants prediction.

        But of course, what we currently have is a Rainier Valley line, with a tail to Sea-Tac. As the rest of the system opens, all these trends become irrelevant. The Capitol Hill station’s walkshed alone probably contains as many residents as all of the Rainer Valley (random guess).

      4. The Capitol Hill station’s walkshed alone probably contains as many residents as all of the Rainer Valley (random guess).

        Not too far off, here’s an interesting Neighborhood Map.

        Capitol Hill neighborhood
        Area: 5.863 square miles
        Population: 40,092
        Population density: 6,838 people per square mile

        Rainier Valley neighborhood
        Area: 10.734 square miles
        Population: 48,996
        Population density: 4,565 people per square mile

        I found it surprising that Capital Hill as a neighborhood on whole isn’t significantly more dense than the overall Seattle average of 6,717 people per square mile. Where is everybody else, Belltown, 1st Hill, U-District? Of course Capital Hill as defined by this map has a lot of park and open space.

        You’d hope that Link ridership is well ahead of general population trends. DT Bellevue, Bel-Red and Overlake for example are expected to absorb the vast majority of new housing and commercial development between now and 2030.

      5. Belltown, Lower Queen Anne, 1st Hill, U-District, Northgate. It was in the NYT tract-by-tract census map Zach posted ages ago.

      6. Here’s what ST predicts for Link ridership each year through 2016. These “forecasts” are as of January, 2011:


        ST is forecasting about a 19% increase in weekday average ridership for Link in all of 2011 over all of 2010.

        ST’s forecast for Link for 2011: 25,000/weekday.

        Actual Link ridership for all of 2010: 21,053/weekday.

        ST predicts a 6% increase in 2012 over 2011, then about 3% per year increase until a bigger increase in 2016, when U-Link opens in the middle of that year.

        I don’t think anyone on this blog has said whether or not they expect Link to reach ST’s most recent forecast of 25,000 boardings/weekday in 2011.

      7. In the latest station-by-station boarding figures for Link, the total average weekday boardings at Mt Baker, Columbia City, Othello, and Rainier Beach stations, both directions combined, was 4,418/weekday.

        I think ST is predicting a lot more than 4,418 boardings/weekday at the lone station on Capitol Hill.

        So, the Rainier Valley has 4 Link stations for a population of 49,000, and Capitol Hill will have one Link station for a population of 40,000, yet ST expects more boardings/weekday at the one Capitol Hill station than at all 4 Rainier Beach stations combined?

        You think that is a reasonable expectation, Bernie?

      8. Yeah, all are more dense than Capital Hill (except Northgate) but cumulatively it’s only 54,000 people out of a population of over 600,000 (about 9%).

        Belltown (Denny Triangle):
        Population: 9,493
        Population density: 12,610 people per square mile

        Lower Queen Anne:
        Population: 8,518
        Population density: 9,309 people per square mile

        First Hill:
        Population: 8,327
        Population density: 16,895 people per square mile

        University District (U District):
        Population: 24,547
        Population density: 8,527 people per square mile

        Capital Hill is average and Northgate is less dense than Crossroads in Bellevue.

        Population: 30,298
        Population density: 4,272 people per square mile

        I’m still mystified at where Seattle is putting all the people that bring the average up to 6,717 people per square mile.

        Population: 5,108
        Population density: 12,426 people per square mile

        Dense, not exactly huge in total population but surprising the Station, even combined with Stadium Station has such low ridership.

        Population: 2,477
        Population density: 7,582 people per square mile

        Also above average but very small in overall population. Queen Anne total has a population of 32,355 (7,359 people per square mile) which partially explains where the people are but it’s not where Link is going. Ditto for Fremont. Broadway is a big (and dense) and I’d would have to admit that it’s as well or better served by the Capital Hill Station as what’s delineated as the Capital Hill Neighborhood. So from the U District to Sodo Link is in relatively high density but outside of that it starts to drop off pretty steeply.

      9. Bernie, I remember that article, although it might have been in the Times. I think Link will prove to be an exception to that rule as there is so much development potential from Sodo to Tukwila.

        Anyhow, I think the quote came from the president of the American Public Transportation Association when he was in town for a conference or something. Hope that helps your search.

      10. ST expects more boardings/weekday at the one Capitol Hill station than at all 4 Rainier [Valley] stations combined?

        Broadway and Capital Hill neighborhoods combined are 64,500 people in 7.4 square miles (8,700/sq mi) which is double the density of the RV. So I’d expect at least twice the boardings. Capital Hill is smack dab between the two major destinations (UW and the CBD) so I’d expect a boost from that. And of course ST projections are for after a new portion of Link is open which will increase boardings at all stations so 4X the current combined RV ridership for Capital Hill seems reasonable. It’s stations north of the U District that I find a bit hard to believe. I also don’t see the steady rise ST predicts in 2012-2015 but I can see a 10,000 spike in weekday ridership when that segment opens (Montlake & Capital Hill)

      11. Thanks Jason. I think I was reading more of an op-ed piece that was probably written shortly after this but the money quote was:

        “There are several rules of thumb, but usually after about six months, you’ll have about half of what you’ll ever have,” Millar said. “And after about a year-and-a-half – unless you have big change in levels of service or land use – you’ll probably have what you’re ever going to have.”

        That seems to have been pretty much dead on the mark. Of course as the economy improves Link ridership and overall transit use will increase (it’s down ~10% from it’s peak). A healthy spike in gas prices that usually accompanies a strong economy won’t hurt either. That’s easily a 10-15% boost which with mild growth should have daily ridership around 30,000 prior to U Link. I’m skeptical that segment alone will double ridership but 40,000-50,000 per day seems pretty reasonable after it opens.

      12. Bernie: Regarding density, check out this insightful post: http://www.humantransit.org/2010/09/the-perils-of-average-density.html

        If you define Capitol Hill a bit more narrowly (perhaps around the station?), I think the numbers will look a lot better for the Capitol Hill station. Obviously, Capitol Hill is much more dense than Rainer Valley so if the numbers disagree with our reality then you’re right to question methodology :)

      13. @John Jensen, very good article. I’ll have to spend some time reading and thinking it through. My “epiphany” in looking up numbers for this comment stream is that there is a lot in how you define a neighborhood. For example in the neighborhood map I linked to the Arboretum is part of Capital Hill. So yes, “density is not destiny.” is true at least as far as it applies to transit planning.

      14. Don’t confuse ‘walkshed’ with the total area of Capital Hill.
        1/2 mile walkshed around a rail station is only .7 square miles. Granted it’s placed correctly around the greatest activity, but it’s still only one station.

      15. “ST expects more boardings/weekday at the one Capitol Hill station than at all 4 Rainier Beach stations combined?”

        Don’t forget Seattle Central Community College. Rainier Valley has nothing that size. While SCCC students aren’t necessarily concentrated around other Link stations (they may just as likely live in Queen Anne or Ballard), those who transfer downtown will find Link much faster than the 10/11/49, and may switch from those buses en masse.

      16. Hey Bernie… take a closer look at that page you linked. Capitol Hill Station isn’t in Capitol Hill as defined on that map… you want Broadway.

        Broadway neighborhood
        Area: 1.575 square miles
        Population: 20,447
        Population Density: 12,979 people per square mile

        That’s, as far as I know, the densest neighborhood in the state.

        That map you linked has Capitol Hill located roughly where the Stevens neighborhood is – an area almost entirely composed of single family houses, which explains the confusingly low density numbers you got. For clarity’s sake, the city uses the name Capitol Hill to refer to the district made up of all the neighborhoods from I-5 to the shore of Lake Washington.

      17. I noted Broadway and the definition of Capital Hill in a couple of places. It goes back to the definition of neighborhood and using that to define density. The article John Jenson linked to covers this very well. The map points to the Stevens neighborhood (still above the Seattle average) when you click on “locate” but it is the combined stats for Broadway, Stevens, Madison Park and Montlake (the latter two really bring down the density average and include the arboretum). It’s too bad the software doesn’t distinguish between broad areas and more detailed neighborhoods. Northgate is a similar compilation of what looks to be Maple Leaf Pinehurst, North College Park and Haller Lake. Lake City is another “super set”. In Bellevue the map doesn’t yet delineate Bel-Red but includes it as part of the Wilburton neighborhood which the statistics say is considerably less dense than Bridle Trails but will be getting two Link Stations (three if you count the Hospital/Whole Foods which I think the City officially considers part of Wilburton).

      18. For clarity’s sake, the city uses the name Capitol Hill to refer to the district made up of all the neighborhoods from I-5 to the shore of Lake Washington.

        Only as far south as Pike/Madison, though. Below that, you get into First Hill and the downtown street grid. Except for a couple of blocks on Pike near the convention center, Capitol Hill is entirely on the cardinal street grid.

        Anyway, for the purposes of land use and density, I think it’s more interesting to use the urban center/village definitions.

        The city defines the Capitol Hill/First Hill urban center as the neighborhoods of Capitol Hill (aka Broadway), Pike/Pine, First Hill, and 12th Ave (Seattle U). As you can see from the map, this corresponds very well to the actual urbanized area.

        For the purpose of census data, there’s also the Community Reporting Areas. These are defined as agglomerations of census tracts that correspond to particular areas.

        In this case, Capitol Hill is defined as tracts 74, 75, and 84.

        From the density map, it’s clear that these are some of the densest parts of the city. Tract 74, the western slope, had a density of 44,096 people per square mile *in 2001*. It’s undoubtedly higher now. 53.01 (the U-District) is actually a mistake; a huge number of students living in dorms were misreported in that district instead of 53.02. So really, the only part of town that even comes close to Capitol Hill’s density is lower Belltown (80.02).

    2. Whoops! “Actually” is left over from a earlier version of the post, when I had some bad numbers for March.


  4. In Norse legend, trolls are dumb, greedy, and have an affinity for accumulated gold. However, one limitation on their power is that if they lose track of time, like if distracted by counting the night’s “take”, the first beam of daylight that hits them turns them into a rock.

    Postings and comments presenting actual facts have the same effect. So more trolls the better. Lot of flooding this year, so Sounder needs the rip-rap.

    Mark Dublin

  5. Here is what I find most interesting about the latest ridership estimates: the difference between 2011 and 2010 weekday average ridership has been narrowing each month this year. For example, for the latest month, April, in 2010 April Link averaged 20,129/weekday; in 2011, Link averaged 21,870/weekday. That is an increase of +1,741 from April 2010 to April 2011, which is the smallest year-to-year increase of any month so far this year.

    Here are the year-to-year increases for each month this year so far in average weekday ridership on Link:

    January: +4,992
    February: +3,847
    March: +3,246
    April: +1,741

    So, each month of 2011, the difference from that month in 2010 has been smaller. This seems like a pretty obvious trend, to me. To round those numbers off, the differences are close to: +5,000; +4,000; +3,000; +2,000; for each successive month. So, average weekday ridership in 2011 keeps getting closer and closer to ridership in 2010.

    I had expected this to happen. The first month that year-to-year comparisons will be really interesting to me will be July, which was the peak of weekday ridership on Link last year. I was expecting that July 2011 might average around 2,000/weekday more than July 2010. But the difference is under 2,000/weekday already, in April.

    1. Based on your astute observations Norman, do you think Airport Link will reach it’s goal of 47,000 riders per weekday by 2020? This is the number used to justify Federal funding on the extension, and excluded any additional ridership attributable to future extensions, such as U-Link.

      1. It’s the bit from TIBS to the airport. That said, many of those projections, like the ~60k for North Link, are predicated on a full build-out of ST2.

  6. Early on, ridership grew due as bus connections changed, habit formed, etc. Today, such basics aren’t changing much anymore. So it’s a function of slower moving elements like jobs, housing growth, tourism, etc. All of which are recovering slowly.

    1. They need to terminate many more bus lines at Link Stations. And in the case of Mt. Baker, I’d find a way to put the transit center directly adjacent to the station. Not across the street but directly underneath.

      1. Who they? If you torture bus connections it’s possible to increase Link ridership by not too much less than you lose overall transit share but at what cost?

      2. Metro, obviously.

        One of the monstrous f*** ups of Central Link, even worse than the Mount Baker TC, but much less well known, is that Sound Transit did not decide what the fare structure for Link would be until far to late for Metro to factor it in to their restructuring of RV bus service in response to Central Link. They didn’t know if it would (say) cost $1 or $5 for a rider to transfer to Link at Columbia City. Very hard to plan a bus network around transfers in a low-income part of town when you don’t know what that’s going to cost.

        There is absolutely no need to torture bus routes in the RV to capitalize on Link and save money. A proposal like Martin’s, combined with some of the ballsier changes outlined in the recent 600k hour cut non-proposal scenario (running the 7X as a shuttle to Columbia City, terminate more of the Renton busses at RBS, split off the south part of the 34 and run it as a circulator between CCS and RBS) would provide comparable mobility at less cost to Metro.

      3. Bus routes shouldn’t terminate at a station, they should, in general, they should run past a station and continue on for a ways.

        That way, the bus can run perpendicular to the train line, both for folks going to the station, and folks transferring to the bus from the station. Add to that the folks who can usefully use the bus route without involving the train. Terminating at the station forces a transfer.

      4. Ideally, yes, and I would be in favor of extensions like that when they make sense. It’s even more important, however, not to run duplicative bus service in parallel with a rail service that’s nowhere near capacity and can take the load for essentially no marginal cost.

      5. Metro did plan to fill the crosstown holes on Genesee and Othello with an increase on the 39, but the County Council cannibalized for the 38 and 42. The 14 restructuring did succeed and covers McClellan. That leaves only Henderson where no solution was planned. Speculatively, there could be a route on Orcas or Graham but there was no money to consider it. The neighborhood said very loudly, “Don’t delete the 7. It works very well and is one of Metro’s most productive routes. Don’t break something that works for something that may or may not work.” Now it may be possible to delete the 7 and 7X, but it wasn’t then when Link was still theoretical.

        Martin’s plan would delete the 7. There’s no way to say that wouldn’t “torture bus routes” or have the biggest impact of any change in the valley. The plan has some advantages but it also has a significant risk, and those have to be weighed carefully and a real corridor analysis done. It may look good on paper but real people’s trips may fall through the cracks (i.e., be underestimated). How many people from south Rainier are going to downtown (and would be helped), vs how many are going to north Rainier (which they could do on the 9), vs how many are going to the International District (where they’d newly have to transfer). Rainier Valley does have a lot of Asians who go to/from the International District a lot.

        Bus routes should ideally have a destination at both ends, either a Link station or neighborhood commercial district. It can go slightly beyond those ends to bring in another residential neighborhood. Another model is a Link station in the middle of the route. But the worst model is a Link station at one end and going progressively less dense to a residential area at the other end, if there’s any possible way the line could be extended to a commercial/transit destination at the other end.

        Um, I suppose I just said it’s useful for the 28 to terminate at 145th/Aurora rather than somewhere short of that. I’m not so sure the northern part of the 28 is that productive. But that doesn’t apply to Rainier Valley, where there are many good destinations a route could be slightly extended to, to make new kinds of trips feasable.

      6. The modified RV proposal idea I mentioned above would improve non-downtown mobility, provide faster two-seat rides to downtown, and save Metro boatloads for money. If that’s torture, our entire bus network should be put to the rack.

  7. Week before last, I rode LINK from Westlake to Sea-Tac Airport talking with two Frontier Airlines stewardesses from the midwest. As with many new visitors I’ve spoken with, they were deeply impressed with both LINK and our whole concept of joint rail and bus operations.

    However, also per usual, they were harshly critical of the fact that they had to find out about the system by accident. It makes me wonder if ST needs a permanent line to Wikileaks.

    Example of necessary measure: ST marketing gets together with all airlines arriving at Sea-Tac to start including in Seattle tickets a packet including an ORCA card with a full-system day-pass pre-loaded and comprehensive transit route and schedule information. And detailed info showing routes to hotels and attractions.

    Successful business doesn’t wait for statistics. It takes action to create favorable ones.

    Mark Dublin

    1. How many millions of tax dollars has ST spent advertising Link, starting since before it even opened? “Ride the Wave.”

      How much adverstising was there ever for the #194 bus?

      1. The best part is it’s here, and it’s not going to go away. In fact, it’s expanding!

        I love link. And I love that it bugs people like Norman. He and the people who are still beating the Save Our Valley drum should go bowling or something.

      2. Link is a terrible waste of billions of tax dollars, as proven month after month with its dismal ridership figures.

        The same thing could have been achieved at a fraction of the cost with improved bus routes similar to SWIFT buses.

      3. Riding the Wave is more than just Link. Remember that people have been riding the wave since 1999.

    2. The airline stewardess who I spoke with, who actually works out of Seattle, was really angry that the 194 bus was eliminated, because it worked a lot better for her than Link.

      1. The only way SWIFT could perform on the level as Link would to build roadways grade seperated for the busses to run. Remember we have to talk about the same route that Link runs. (Not running up and down I5, we need to compare it equally) The moment you start to build that grade seperation…guess what you got that billions and billions, as you called it “wasted”. Unless you are just going to run a bus from Seatac up through the valley and to Westlake faster than the link can.

      2. The 194 north of the airport was not a Swift-like route. A Swift-like route would have stopped every mile or so along the way. The 194 was a point-to-point express from Spokane Street to the airport (with Spokane pretending to be “the beginning of downtown”). South of the airport, the 194 was more like Swift, and that was retained with the 574. (Although Swift stops at [future] TOD nodes and the 574 stops at freeway exits and P&Rs.)

        We might as well start from the fact that Link is slightly slower than the 194 for a Westlake-to-Airport trip in ideal conditions (no traffic). 37 minutes vs 28; that’s 9 minutes difference, or 25% of Link’s trip. Under normal daytime traffic, the 194 took anywhere from 28 minutes to 35 minutes to 40 minutes, so its advantage is partly or completely lost.

        10 of Link’s minutes is going from Westlake to Intl Dist, and that can be argued as just “downtown overhead” (i.e., the tradeoff for close stop spacing downtown). Starting from Intl Dist, Link takes a respectable 27 minutes to the airport (or 15 minutes to Othello).

        So any way you look at it, a trip from downtown to the airport takes 30-45 minutes, and Link is up to 9 minutes slower than the 194. Whoop-de-do about the 9 minutes. That’s one episode of buying a latte or putting on makeup or going to a phone booth you can’t do. In exchange for the 9-minute overhead, you get much more frequency. Link runs every 7-10 minutes until 10pm, and 15 minutes till midnight. The 194 ran every 15 minutes daytime, 30 minutes evenings, and not at all after 9pm. The 174 “replaced” the 194 after 9pm, taking 45 minutes. The effective trip time is the travel time plus the wait time. The “average” wait time is plausably half the maximum wait time, because only a minority of riders can choose when they arrive (i.e., only if they’re not transferring). So if you include the wait time, the 194’s advantage also vanishes.

        At longer distances, ST Express overtakes Link. If Link reaches Tacoma someday, it may be 10-15 slower than ST Express for Federal Way-Seattle, and 10-20 minutes longer for Tacoma-Seattle. Another issue compounds it: the 60-minute milestone. “It takes less than an hour” feels different from “it takes almost an hour”, and again from “It takes more than an hour”. So if Link’s travel time to Federal Way and Tacoma turns out to be worst-case, there’ll a bigger justification for parallel ST Express service. But that absolutely does not apply to the downtown-SeaTac situation. For north Link, I think Link will be able to stick to the “no worse than 10 minutes slower than ST Express” for downtown-Lynnwood and downtown-Everett, so those situations would be like SeaTac’s.

      3. Flight Attendants know to be kind to the elderly. They know that listening to them makes them quiet down.

        I’m happy you had a nice bus ride, Norman. Independence is everything, next to health, isn’t it?

    1. How does New Orleans’ comprehensiveness, frequency, and speed compare to ours? That shows what you’re getting for the $3. Our system may not be ideal but it’s within the top 10 of US cities, so we are getting something for the money.

    2. OK, after two days, and only one negative comment, I’ll assume there are no ideas out there to increase Link’s ridership.
      Thank you all for your participation.

      1. Spend the next decade or two building density around stations, add more lines, increase frequency, increase length of trains (rush hours and saturdays can get pretty tight), wait until the economy improves (tourists, commuters)…. In other words, keep doing what they’re doing. Short term, things are going fine.

Comments are closed.