Ridership Patterns on King County Metro Route 7
Ridership Patterns on King County Metro Route 7

Connecting Downtown, the International District, and a swath of neighborhoods in the Rainier Valley, Route 7 is one of Metro’s most important routes. Along with many of Metro’s trolleybus routes, it was originally a streetcar, and also like many of those streetcar routes, it was put in to help spur development in what was previously untamed land: transit- and pedestrian-oriented urban sprawl, circa 1890. In a previous post, I discussed some of the problems suffered by the current incarnation of Route 7, and discussed one possible improvement. In this post, I’ll focus mostly on current ridership patterns.

First, some notes on the chart. As I discussed at length in the aforementioned post, Route 7 has a split service pattern, with most trips turning back at Rainier Beach High School, and others continuing to Pentice St. This manifests itself as a discontinuity in the data in the outbound (right) pane. More importantly, Metro’s APC system assigns both the outbound and inbound legs of the Prentice St loop to the outbound trips. So, reading upwards in the right pane, from Rainier/Hend to Prentice/64th the bus is heading south; subsequently it is heading north.

Analysis after the jump.

  • Blockbuster demand on Jackson St. At all times of day and night, riders are traveling both directions on this segment; we saw the same thing with the 36 and 14. Other than 3rd Ave, no other route segment can touch this one.
  • Relatively weak between Mount Baker Station and Dearborn. Even though that area is densely built up, it doesn’t have the pedestrian-oriented streets found elsewhere  on the 7.
  • Service workers going from the RV to the Eastside? One interesting quirk: lots of inbound pre-6:30 riders get off at Rainier/I-90; presumably they’re heading east. We don’t see this in any other time period.
  • High turnover inbound at Mount Baker Station. The two stops closest to MBS, Rainier/Forest and Rainer/McClellan, exhibit a strange asymmetry, with far more rider churn inbound than outbound. I can’t explain this.
  • Mixed main-street/neighborhood pattern out to Henderson. The route exhibits steady turnover at each stop in this segment, somewhat like the 358, but overall, the bus is gradually unloading.
  • Complicated after Henderson. Looking at the right pane, in all time periods, the bus still has five to ten riders heading out to Prentice St. Those riders deboard heading south, but virtually no-one boards heading north. This is presumably due to the inbound trip being interrupted by a layover, as discussed in my prior post. My take on this is that there’s latent demand on the Prentice St loop that is served reasonably well by the outbound 7, but not by the inbound. This is part of the motivation to fix the south end of the 7.
  • Surprisingly low average loads. Bredas nominally seat 56, and my limited anecdotal experience riding the 7 on Rainier suggests those buses are usually not standing room only, but certainly feel more full than the average of 25 shown in the chart. My guess is that buses are getting bunched, and full leading buses are offset by nearly-empty trailing buses.

36 Replies to “Ridership Patterns on Route 7”

  1. to the last point: i live in columbia city, and about 1/3 of the time i see a seven (between hudson and i90) it’s bunched with either a 9 or another 7. when i first moved here 8 years ago it was 4/5 of the time, so its actually not so bad now. but still ridiculous.

    1. I’m not surprised. When I used to live downtown, most of the time when I saw an inbound 7, there was another one right behind it.

    2. I see 7’s bunched up all the time on Rainier. It’s not uncommon to see 3 7’s within 3 blocks going southbound on Rainier.

  2. I wonder if most people on the Prentice St portion of Route 7 just walk to Renton Ave S to catch the Route 106 bus inbound. It would probably be easier than trying to figure out the complicated transfer at Rainier & Henderson.

  3. The low ridership between MBS and Dearborn makes me worry less about eventually fixing the transfer at MBS and stopping the 7 there.

  4. On inbound turnover at Mt Baker: Franklin High is probably the main factor.

    But another factor could be that transferring to the train has a shorter average wait than transferring to the bus because of shorter headways and less bunching on the train. So there may be people who take the bus to Link in the morning who just walk home from Link in the evening (or only take the bus if it happens to be there).

    1. It could be happening because when people are going to work they care more about being on time, so they switch to Link, whereas when they are coming home it is not such a big deal.

  5. A note about Jackson. Part of my convoluted commute starts at about 16th and Jackson. I walk toward downtown until OBA tells me a bus will come soon. Frequency can be quite good by the time I get to 12th, but it’s terribly inconsistant. On many mornings I’ve walked all the way to King Street Station to catch a bus. On these days I pass large groups of people waiting at each bus stop.

    I’d complain, but my hope is the streetcar will make everything better. Of course if it’s anything like the frequency of the SLUT I might find myself walking anyway.

    1. The streetcar can’t go faster than a bus, so it’s only advantage along Jackson is having a clear ROW with fewer stops. Park a Fedex truck even close to the tracks and the streetcar is stuck. Signal pre-emption will help, but when signals are queued up for a block in every directions (12th), then all the fancy timers in the world won’t make it go away or clear out your section to do much good. And the buses can take advantage of that too, so really no advantage to the streetcar. The buses may bunch up, but at least the drivers are smart enough to go around all the idiots that park along Jackson for veggie days.
      Gondolas rule!

      1. I’m so tired of this idea that streetcars can’t go faster than buses. With level boarding and multiple doors, streetcars can load and unload much faster than buses. And FedEx drivers don’t park in front of something that can’t go around them.

        They may not be a good investment, but they can be faster than buses.

      2. Fedex and UPS and other random clowns can and do, happens all the time in SLU. I don’t think it’ll be a problem on Jackson, though, due to the center running, and it being more of a car street; all of the businesses have front parking for drivers to pull into.

        Trolleybuses with low floors, passive restraints, and off-board payment, and streetcar-like stop spacing can get you 85% there for 15% of the money. Unfortunately, building a decent rapid urban bus/trolleybus system seems largely beyond the imaginations of most transit agencies in the US.

      3. Parking in front of the streetcar: Huh. I haven’t seen that happen, but I haven’t ridden the SLUT in a long time. I wonder what it would cost to put wifi cameras in the streetcar so as to be able to automatically ticket people parking in the streetcar lane.

        Better trolleybuses: agreed, especially if you can get wider aisles and a faster wheelchair lift. I’m not sure it’s only 15% of the money (trolley buses aren’t cheap and streetcars aren’t all that expensive), but it’s definitely a good deal if you can get it to happen.

      4. “automatically ticket people parking in the streetcar lane”

        Or just add a cow catcher and some extra steel reinforcing. They’d only park illegally once.

      5. One advantage streetcars have over buses is passenger capacity. Both in terms of passengers per hour and passengers per vehicle. The FHSC segment on Jackson is perfectly placed to capture all of the demand there.

      6. No, it’s not though. It doesn’t go all the way into downtown, and the buses that compete with it will run more frequently and not require crossing to a center island. Had the streetcar been curb running, your statement would have some truth to it.

    2. if the streetcar is implemented, it may South Jackson Street flow worse; note that intending riders will wait at the curb for routes 7, 14, and 36 and on islands for the streetcar, so they will not be able to take advantage of the combined service frequency; one can see this today on Fairview Avenue North near Fred Hutch with the SLU stop inside and the Route 70 stops outside; with both the ETB routes and the streetcar stopping in-lane, but in different lanes, and more general purpose traffic due to the south portal of the deep bore, South Jackson Street may get crowded; the island stod in Little Saigon may slow traffic between 12th and 14th avenues South as well. the streetcar is supposed to have 10-minute headway in the peak periods; routes 7 and 36 both have 10-minute headway; Route 14 is atop of them.

  6. The question of unloads on I90. I commute from Genesee/Rainier to and from downtown.
    My commute frequently involves transferring at the I90 stop and going to and from downtown by the endless buses serving the east side. Especially out of downtown at the end of the day. Hop in the bus tunnel, wait a minute or two for an east side bus, then hop off at Rainier. No traffic, 2-3 minutes from the ID to I90. Then I can hop on just about any bus going south (7, 7x, 9x, etc.). It’s the fastest trip between the ID and the valley. You try getting from the ID to Rainier/I90 in that time at 5pm.

    So, I will somewhat frequently hop off a 7 or 9x at I90 in the morning if there’s any traffic or crowdedness and transfer to a bus coming from the east side. Especially if it’s food bank day in the ID!

    1. I agree, and I know people who do that. But pre-6:30AM, the time period where I noted all those deboardings, Eastside buses won’t be that frequent, and Rainier traffic won’t have reached gridlock. It doesn’t seem like a transfer at that time would offer much advantage.

      But again, I’ve never ridden the 7 at rush hour, so I have no anecdotal experience of where those people might be going.

    2. I can see how your trick is much faster than taking the 7 all the way, but compared to taking Link all the way, I’d bet it’s closer to a wash depending on how long you have to wait for a bus. If I were doing it, I’d probably go into the tunnel and hop on either Link or a 550, whichever comes first.

  7. Bruce: Great graphs as always. It looks like there are 10-15 people still on the bus at the bottom of each graph, which doesn’t seem right. The bus should be empty at the last stop. Am I reading too much into the graphs?

    1. I should have discussed this in the post. There are three issues here, one manifested on both ends of the chart, one on the bottom (downtown).

      Nights and Sundays, the 7 is through-routed with the 49, and thus has different terminal stops and different ridership patterns. Rather than present data that would, by necessity, either be confusing or oversimplified, I chose to cut off the chart at Seneca and Union, to which point the routing is straightforward.

      On the top of the chart, in the AM peak, midday, and PM peak, you can see evidence of (what I presume is) cumulative APC error. As hundreds of people churn on and off, the APC system slightly overcounts boardings over the course of the day, so there are are “ghost riders” left at the end of the route. This was also a big problem with the data for the 14S.

      Finally, I don’t know what the heck is up with the Hend/Rainier stop at the top of the left pane. Data from terminal stops always seems to be messed up. I should probably have excluded that data point from the chart.

      1. Thanks Bruce, that all made sense – even to me!
        It’s a result of too many routes that have too many variants, trying to do everything for everyone all the time.
        My favorite name for these route variations are “Route Mutations”.
        How many riders, except the transit geeks, can figure out all the variations of a route, then apply that knowledge daily, without ever getting caught on a bus that doesn’t really go where you think it’s going?
        I think it cost Metro ridership in the long run, but they and the computers are convinced it saves money. There probably right, but it’s still a source of irritation to some riders.

      2. It took me months as a commuter to optimize my bus ride home – sometimes by trial and error (I admit to confusing an 8 for an 18 once, looking up from my book halfway down Denny – that cost me a long walk). Then Metro changed the routes around on me.

        And I’m someone who takes the time to figure these things out. I’d guess I saved 15 minutes a day using my method. Assuming 1 out of every 10 commuters in our region could similarly benefit but don’t know it, that’s a waste of about 850,000 man hours a year*. Assuming a median wage of $15/hr, that’s $12.8M a year wasted simply from lack of information.

        * With the wild assumption of 200k commuters. I’ll bet there’s a lot more bus commuters than that, but I don’t know the number off the top of my head. Doesn’t matter much anyway – the uncertainty is dwarfed by my 1 in 10 commuters guess.

      3. Matt: What a Hoot.
        I rest my case about Route Variants.
        BTW, has the program ever faulted and you get stuck in an endless loop?

      4. One of the interesting side effects about the long-term trend to move towards a few frequent trunk routes, rather than tons of infrequent routes each doing slightly different is to reduce the advantage people like Matt have, who have months to optimized their commutes, over everyone else.

  8. Problem of “bunching” is nothing new. Fall shake-up 1994, I had an afternoon run on the Route 7, including a turn on the 9. Five days a week, picked work- I really liked the 7.

    However, infuriating thing about that run was not the lack of breaks- I looked at trolley-driving like a motor sport in those days, in itself pretty stupid but more common than generally known.

    Really maddening thing was spending an hour northbound on Rainier aboard an empty sixty-foot bus between two identical buses, with only my leader carrying passengers. Natural response would’ve been to request control to tell leader to blank signs and leave me passengers.

    Or, drop back so bus stops could “refill.” But couldn’t do that, because my follower’s schedule required he leave Downtown on time.

    Remedy? Bus-only lane the whole route and signal pre-empt would have solved a lot of the problem. In San Francisco, supervisors occasionally take late buses out of service and turn them back, re-loading passengers on follower. Requires much turnback wire.

    Failing that, however, might help to give drivers means to space themselves out-maybe their own radio channel, or wayside signals.

    Problem is worth some addressing: empty buses not only waste tax money, but even worse, give anti-transit people something to be right about.

    Mark Dublin

    1. Agree absolutely. The current situation doesn’t serve riders well, doesn’t serve taxpayers well.

    2. Or, you could make a game out of it. See how many minutes you can leave Henderson LATE, and still catch your ‘slow-pokey’ leader by the time you hit IDS. Last run of the day, of course!

  9. Bruce,

    As usual, your charts and analysis are really fun to read.

    If you need ideas for future posts, I’d love to see a chart of combined ridership on Jackson — it seems like “killer ridership on Jackson” is a frequent theme.

  10. Some of the data points here seem to be in conflict with the proposed solution of cutting the 7 at Othello. The graphs here indicate that Rainier and Henderson is one of the largest on/off spots outside of downtown. Cutting the 7 at Othello means that all of those riders would have to transfer.

    I don’t see the advantage of moving the 7 to MLK and Othello given that the 8, in conjunction with Link, serves that corridor well and that these data show the high numbers who rely on Rainier and Henderson.

  11. One thing to keep in mind with the 7 is that it’s heavily used for trips within the valley. So it shouldn’t be split between Henderson and at least Mt Baker, or it would divide the neighborhood like an Othello Freeway would. No more one-seat rides from Cloverdale Street to Columbia City or Oberto’s. Those kind of trips are what make the Rainier Valley community a thriving entity. South Seattle is severely handicapped in east-west connectivity due to hills and waterways, so north-south connectivity is the only thing they have to keep their neighborhoods together. The 7 is one of Metro’s star routes: heavily used and well-liked. I’d really hesitate to break something that works so well. Redirecting the north end to Boren or Broadway, or reassigning the Prentice segment to another route, would be better than breaking up the middle segment.

    1. I agree, splitting the 7 between Henderson and Mt. Baker would be a bad idea.

      One suggestion I keep seeing repeated is to split the 7 at Mt. Baker and make the segment to Rainier part of the 48. The segment between Downtown and Mt. Baker would become a separate route.

      I kind of like this idea, but I don’t think combining the Southern section with the 48 would be a wise idea, due to reliability issues, even if the 48 was split in the U-District.

      In fact splitting both the 8 and the 7 at Mt. Baker (along with splitting the 8 at Madison and the 48 in the U-District) would make a whole lot of sense, improve reliability, and allow service hours to be matched with ridership.

    2. The TMP has a corridor 5 as a 48S/7: going UW, 23rd, Mt Baker stn, Rainier, Rainier Beach, Rainier Beach stn. Corridor 3 as a 60/36: UW, 10th E, Broadway, 12th, Beacon, Othello stn. Corridor 4 as a short 7: downtown, Jackson, Rainier, Mt Baker stn. That wouldn’t be bad because it keeps the easternmost routes east and the middle routes middle. It would separate south Rainier from north Rainier, but that’s better than a split within south Rainier. It assumes the corridors will be upgraded to HCT standards, which should take care of the Montlake bottleneck that kills the 48’s reliability.

  12. “Service workers going from the RV to the Eastside?”

    I can’t think of any other reason for I-90/Rainier’s existence.

    Every single I-90 bus going beyond Mercer Island stops at this freeway station that’s a terrible place for TOD because of the especially wide footprint of the freeway here, not helped by the freeway station itself. So if the station isn’t a destination in and of itself, it must be a connection to somewhere else.

    The 7 is the only frequent route on Rainier that serves the freeway station, which is great for people in the Rainier Valley commuting to the Eastside, but doesn’t exactly connect job centers, and so isn’t terribly useful for the numerous one-way peak-only express buses that serve the station. Unless you look at the route north of the station, where it goes to the International District and downtown… the exact same places all those buses are going.

    What happens if we reroute the 7 to skirt downtown as multiple people here have proposed in the past? If we send it up 12th Ave we’re hitting Seattle University and not much else unless you consider Capitol Hill a job center or revive the old 9 to the U-District. Once U-Link goes in, it won’t be necessary to go to either of those two places, especially once East Link makes it a one-seat ride.

    But if we send it up Boren? Suddenly we’re hitting two major job centers in First Hill and South Lake Union.

    I didn’t like Zach’s proposal to send the 7 to Boren a while back, preferring 12th based on the current service structure of the 9 and 60, and I say this both as someone that would benefit greatly from a 7/9 on 12th and with the caveat that the 9’s current kink along Jackson and 12th could negate the geometric attractiveness of an all-Boren/Rainier route… but considering that the 7/9 hits I-90/Rainier and the 36 doesn’t, sending the former up Boren could produce massive ridership that could easily justify 15-minute if not 10-minute frequencies on a corridor that isn’t currently served at all except by peak-only expresses.

Comments are closed.