Transit Score: Buses Matter

Changes to Seattle's Transit Score underline the impacts of buses.
King County Metro, facebook

King County Metro, facebook

They’re out! The cool kids over at Walk Score just posted their all-new 2014 Transit Score rankings with data on transit service in more than 200 cities across the US. And in good news for the Northwest, Seattle’s Transit Score ranks 7th among all large cities, trailing only New York, San Francisco, Boston, DC, Philly, and Chicago. Portland, meanwhile, ranks 10th. (Note that Transit Scores only rank transit within city limits, and don’t cover suburbs or surrounding municipalities, and that I’m counting cities with at least half a million residents as “large.”)

There’s good news beyond the Northwest, too: on average, Transit Scores have inched up a bit since 2012—suggesting that more people are living closer to frequent transit service.

The bad news, though, is that both Seattle and Portland saw their Transit Scores fall over the last two years.

I was lucky enough to get an early peek at the raw data, which provided a fairly fine-grained look at the trends. As it turns out, Portland’s drop was so small that it probably doesn’t mean much. But Seattle’s Transit Score fell by more than a full point (technically, 1.3 points, but it winds up looking like 2 points because of rounding). That’s one of the largest drops among the 80 cities that had Transit Score rankings in both 2012 and 2014.

So what gives? Why did Seattle’s Transit Score dip?

The Transit Score methodology can seem a little confusing. But basically, adding transit routes or increasing the frequency of nearby transit service boosts a location’s Transit Score, while eliminating routes or reducing service frequency lowers the score. Rail, cable cars, and ferries count more than buses towards a location’s Transit Score. To score a whole city, the Transit Scores of individual locations throughout the city are weighted by population, then averaged.

With all that in mind, and looking at the neighborhood-level changes in Transit Scores, there’s one likely explanation for Seattle’s dip. In late 2012, King County Metro made some significant changes to Seattle bus service—closing and consolidating some bus routes, while boosting service on the new “RapidRide” C and D lines. And on net, the reshuffling of neighborhood transit service reduced the city’s overall Transit Score.

Some parts of the city clearly benefited from the changes. For example, the Fauntleroy, Roxhill, and East Delridge neighborhoods in West Seattle are all on the route of the new C line, and all saw a nice bump in their Transit Scores since early 2012. Just so, Interbay, East Ballard, and Broadview all saw a Transit Score boost from the D line.

But at the same time, neighborhoods like Mt. Baker, Columbia City, Beacon Hill, Brighton, and Holly Park had particularly steep declines in Transit Scores. That’s where Metro eliminted of “duplicative” routes along the Central Link light rail lines in late 2012. And, as shown by Eric Fischer’s nifty map to the right, that same corridor also happens to contain the city’s most racially and economically diverse neighborhoods—suggesting that the bus service changes had a disproportionate impact on the lower income, racially mixed neighborhoods in Southeast Seattle.

It’s possible that light rail previously had given Southeast Seattle a nice Transit Score boost. Still, it’s clear that the 2012 bus changes reversed a good share of any gains from light rail.

Just to be clear, I’m not trying to critique Metro here. As far as I can tell, they’re doing their best to provide efficient, frequent service throughout an entire county, not just Seattle proper. And besides, it’s not clear how Transit Score stacks up to actual transit ridership. Indeed, even as Seattle’s Transit Score fell, total transit Ridership on both Central Link light rail and King County Metro bus service rose.

Still, the Transit Score trends for Seattle neighborhoods may reveal a powerful lesson: don’t underestimate the impact of bus service. Even as the city’s rail system expands, buses will remain the mainstay of Seattle’s transit system for the foreseeable future. That means that big changes to bus service can have big impacts on overall transit accessibility—and that even in a light rail corridor, transit-dependent communities can face a big hit when their bus service is trimmed.

We are a community-supported resource and we can’t do this work without you!



  1. Lori Ahouse says:

    Clark, you are more kind than I would be. As a 15-year resident of SE Seattle, I can assure you that it is not a mere coincidence that the neighborhoods dropping in score are from our racially diverse areas. Same story applies to so many of our public services.
    I appreciate your reporting transit scores though. Important stuff – and all the more so with another round of Metro cuts looming.

    • Clark Williams-Derry says:

      What I wonder is whether SE Seattle’s Transit Score is now lower than before the light rail line opened. There’s probably no way to know, unless someone happens to have saved the transit data from 2009 in a format that Google Maps can read. But if it were true…well, that’d be extra bad news for SE Seattle.

  2. Anita Woodruff says:

    I would like to see transit scores reflect service in surrounding suburbs rather than just focus on the city limits. Many of us live in the suburbs surrounding Seattle because we simply cannot afford the city’s ridiculously high rents. We would still like to be able to rely on public transportation but service is so infrequent that it makes this difficult. I believe the transit score would be significantly lower if the experience of those of us living in the nearby suburbs was factored in.

    • Clark Williams-Derry says:

      Good point!

      When you search for an address on the main Walk Score page, it will give you the Transit Score too — provided that there’s a reliable “feed” for transit data in the area. So Walk Score can give suburban Transit Scores…but hasn’t yet published comprehensive Transit Scores that aggregate those scores across every municipality.

      My guess is that, outside of a few transit-oriented hubs (Bellevue, perhaps Redmond) suburban Transit Scores will be low. And I agree that it’s a shame that bus service in suburbs can be hit-or-miss! That’s one of the problems with the suburban form: because it’s so spread out, it can be prohibitively expensive to provide reliable, high quality transit service.

    • Kenny says:

      This seems especially important given the set of cities that they are comparing. For some cities, like San Francisco, the distinction between being in the city and in one of the suburbs makes a lot of sense. But for others, like Los Angeles, some of the densest parts of the urban core are actually in other cities, like West Hollywood, Culver City, and Santa Monica, while some parts of the actual city are a completely separate region, like the San Fernando Valley, or the San Pedro harbor area. Similarly for New York, it seems odd that remotest Staten Island and Queens count against transit score, while Jersey City doesn’t count towards it.

  3. CJ Clark says:

    Forever (as in the past 30 years) there was a bus line (# 39) that went from Seward Park/Lakewood/Rainer Beach to downtown. In the AM and PM there was even an express service. This all ended in the spring of last year. To go from these areas to downtown by bus one has to change in SODO, almost doubling the transit time, OR take the new service (# 50) to the Light Rail, doubling the cost. For some reason, known only to Metro, the new service (# 50) runs from West Seattle (Admiral) to Seward Park via SODO. I say “known only to Metro” as neither I, nor anyone I know, has ever expressed a need (or desire) to travel between these two areas. Note also the # 39 route was well utilized, especially during the rush hour times.

    • kk says:

      CJ – how does transferring from the 50 to Link double the cost? That’s a free transfer.

    • d.p. says:

      “doubling the cost…”

      No. Enough with the fictions.

      Get an ORCA. Stop holding your fellow passengers up with your bills and quarters. Congratulations, the transfer to Link is now free, and your ride downtown is both faster and easier.

      Because replacing infrequent, laborious routes from everyone’s front door to downtown with intuitive transfers to frequent, quality trunk routes is how a system becomes better over time, and scales to meet the needs of a growing population. Many of the best transit systems on Walkscore’s list run NO buses from their outer corners all the way to downtown.

    • Chris Arkills says:

      I use the 50 often to transfer to Link from West Seattle. I know others who access the VA from WS on the 50 as well. We needed a SW-SE connection.

  4. kk says:

    I think you’re missing a key detail on South Seattle’s score. Link boosted their score (while all service was maintained) then the score receded after duplicative service was removed. The net effect could easily still be more service due to the existence of Link. What you present doesn’t address that at all.

    • Clark Williams-Derry says:

      I think that’s exactly right, kk — as I say in the post, “It’s possible that light rail previously had given Southeast Seattle a nice Transit Score boost.”

      I just don’t know how today’s transit score matches up to transit scores prior to Link, or how today’s transit accessibility (which Transit Score tries to measure, but may do so imperfectly) lines up with transit accessibility prior to Link.

      All we know for sure is that the bus service changes in 2012 gave a real wallop to the corridor’s Transit Score. Whether they’re “better” or “worse” than before Link…I really don’t know.

  5. Chris Arkills says:

    A little context. First, I love the folks at Walk Score. But their model doesn’t always recognize some details about transit. First, Link+Metro provides far more service hours in SE Seattle, which has a higher than average level of service for ridership than most of Seattle. Second, Rapid Ride has average stop spacing of nine blocks, rather than 2-3 blocks. Adds speed and reliability, but may reduce walk score. Third, the KC Council mandated that Metro restructure 100,000 service hours to more productive corridors, which they did with Rapid Ride. Metro added core service to busy routes like the C line, 120, and 21 in West Seattle, but cut service to less productive areas like Alki and Arbor Heights except at peak hours. So Metro, carries more people in West Seattle, but may offer less geographic coverage. Finally, the Times linked this story to a poll on Rapid Ride which is misleading. Rapid Ride C line is up 55% over the previous 54, and the D line is up 15%. The A line in South King is up well over 50% and the B line in Bellevue is also up about 20%.

  6. ad says:

    Agreed on the issue with Transit Score. The methodology is pretty rudimentary (multiply rail routes by 2, bus by 1) and the distance decay function utilized is the same regardless of route, when all the research out there shows that people will walk more to high capacity transit. Additionally, the normalizing step may adjust scores and the so called drop of 1.3 could easily be within the margin of error. That said, they are the only ones out there doing a comprehensive scoring for all agencies, so it’s at least something, but I wouldn’t get too deep into trying to interpret the results as it could really just be noise.

    • Clark Williams-Derry says:

      One thing, though, is that multiplying rail by 2 affects the distance function, by making rail twice as attractive as a bus to someone who’s got a half-mile walk.

      Still, I agree that Transit Score is a rough-and-ready measure, and could benefit from refinement…though in practice, it’s not clear that there’s a single “right” answer about how to measure transit access, since transit can serve so many different functions.

  7. Manuel Soto says:

    Hi Clark,
    So how does it work? I haven’t yet explored the inner workings of transit score. Are they counting cumulative frequency (of bus routes and transit modes) on a block by block basis and then weighting by population in that block? If so, are they counting bus frequency for all blocks along the line, and light rail only along a walk distance from stations? Since link stations would be at least 0.5 mile or more apart, there could be “frequency gaps” along the lines, and then a higher frequency count around stations and lower count along the line. If that is the case, the methodology would be favoring concentrations of people around light rail/subway stations (they have a higher weight than bus) than along the entire transit corridor. So denser development in the form of TOD, around BRT/LRT/subway stations, would help boost the score in the future? Still, I like your point about the importance of underlying bus service. It opens up a debate between favoring development in TOD nodes as opposed to Transit Oriented Corridors or TOC.

Leave a Comment

Please keep it civil and constructive. Our editors reserve the right to monitor inappropriate comments and personal attacks.


You may add a link with HTML: <a href="URL">text to display</a>