Houses being demolished for Yesler Terrace housing project, Seattle, 1940
Houses being demolished for Yesler Terrace housing project, Seattle, 1940. Photo by Seattle Post-Intelligencer, courtesy of IMLS

Update: The current Yesler Terrace project is 22-acres. The redevelopment is expanding it to 30 acres. So I’ve updated the math. Point still holds.

The Seattle Times ran an article about concerns over the Yesler Terrace replacement project. What jumped out at me:

[Hired land use consultant Matthew] Gardner is skeptical that Yesler can capture 10 to 15 percent of new residential development citywide, which he says the plan calls for. But Heartland points out that Belltown grabbed 22 percent of the city’s new housing in the 1990s.

Yesler Terrace is 22-acres, or about 3.4% of a square mile[1]  30 acres or 4.7% of a square mile. With 5,000 units and assuming only one person living in each unit[2], that’s a residential density of 145,455 106,667 people per square mile, more than twice one and a half times the density of Manhattan. The project will also have a good deal of office space, some open space and a community center.

The city can plan on putting 10-15% of new residential development citywide[3] on 3.4 4.7% of a square mile and 0.04% of the city’s area, with mostly mid-rise buildings, through the magic of density. In Black Diamond, 6,500 units are being built on 1,500 acres. I know I’m preaching to the choir, but this is why dense construction near transit is so important. Though we are building rapid transit access to only a small portion of a the city, it clearly could be very easy to accommodate all new housing construction in just the planned Link station areas. Sadly, we are mostly getting new parking lots and vicious fights.

1. It is possible the development project is larger than this, though I cannot find evidence of that.
2. Obviously there will be more than one person per unit.
3. I am not sure what time frame that is supposed to be over.

57 Replies to “Yesler Terrace and Density”

    1. What are you talking about? Currently Yesler Terrace is among the least-dense part of Seattle.

      Clearly this is proof that anti-density leads to slums and poverty. /sarcasm

    2. You seem confused, Bailo. Currently Yesler terrace has about 1100 residents, so that’s about half the density of Manhattan, and a bit less than LQA or Belltown.

      1. 2010 Census Map

        The five most densely populated census block groups in Seattle are:

        1. 80.01 Group 5 (Belltown): 110,810 ppl/sq. mi.
        2. 80.01 Group 4 (Belltown): 67,489 ppl/sq. mi.
        3. 74.02 Group 3 (Capitol Hill): 65,855 ppl/sq. mi.
        4. 53.01 Group 1 (U District): 64,672 ppl/sq. mi.
        5. 53.02 Group 1 (U District): 61,854 ppl/sq. mi.

      2. I guess City Data is egregiously wrong…although maybe something like “Queen Anne” and “Belltown” are ambiguously defined regions.

        The map is interesting, but it’s still a secondary source.

        I’d have to go back to the regular 2010 Census data to be assured.

      3. John, you are trying to compare whole neighborhoods with a tiny 22-acre development. That’s why the numbers cannot match.

      4. John,

        Manilla is a city.

        Belltown is a neighborhood.

        Yesler Terrace is a low-income housing project that is part part of a neighborhood.

        Apples, Oranges, Mangos.

      5. It’s interesting to compare developments. The 43 story Bellevue Towers for example is 540 units and sits on a little over 2 acres. If all the units were full and averaged 2 residents per unit (probably would be closer to 1.5) it works out to ~13,000 people per square mile. High density for sure but not even close to mega-density numbers like Paris at 54,000. It’s a good example of tall does not equal dense. The reason for the height is to build larger units and with a view so that they can be sold at a fancy price. At a 1/2 million to 5 million each it’s nobodies idea of “affordable” housing. The proposed 36 acre Spring District in Bell-Red would be less than 8,000/square mile if built out as planned. That’s about the average for all of Seattle and often considered the minimum for effective public transit.

      6. @Andrew

        But in your own post you make the comparison [Yesler Terrace has] a residential density of 145,455 people per square mile, more than twice the density of Manhattan.

        So aren’t you doing exactly that same thing — comparing a small neighborhood to the borough of a large city?

    3. John, the statement was regarding density after the planned redevelopment, not now.

      1. Ok, so after the planned redevelopment, it will make sense to compare Yesler Terrace with Manhattan.

        Great…I’m so glad you clarified that.

        Before the “planned redevelopment” Yesler Terrace and Manhattan Island had nothing to do with each other. It was “apples and oranges”.

        But afterward, Yesler Terrace will be ideal to compare with Manhattan.

        How could a rational person not have see this?!?!

      2. I think it’s a perfectly legitimate comparison – you can still take the average density of a representative chunk of Manhattan and compare it to any given slice of any neighborhood. As kind of a above average/below average sort of thing. We are looking at area averages, after all.

        However, I think the comparison is a bit skewed, because Yesler Terrace is bordered by I-5. Either some of that unusable I-5 space should be included in the calculation, lowering the YT density figure, or we should compare to the average Manhattan density with their freeway land disregarded, incressing the Manhattan density figure.

      3. It’s apples and oranges. Manhattan density counts freeways, parks, schools, churches, museums, opera houses, and all of the less tall/dense areas we don’t think about when someone says “Manhattan”. The comparison was clearly made to give some sense of density, but it doesn’t hold up as a real analogy. John was called out on his comparisons because he was taking them too seriously.

      4. I think talking about population density in Manhattan has got to be a slipper fish if there ever was one.

        Unlike in 1900 when Manhattan was more a place were people lived, and it had twice as many people before urban flight over the next century, much of its population is transient.

        Also the building are highly heterogeneous. From sky scrapers in midtown, to brownstones on the Upper West Side to high rise luxury apartments — sometimes with only one unit per floor — on the Upper East.

        And then there’s impoverished areas on the north end.

        Oh, and don’t forget Central Park — a huge track of land that is not home to anyone in particular except probably for some homeless people.

    4. I take issue with the term “slum” as you use it here. Yesler Terrace is simply not a slum. Most residents are low income, but the community there is actually much more vibrant than many places in the region. There are fantastic examples of micro-agriculture, entrepreneurship and business incubators, and real community inside Yesler Terrace that far outweigh the stereotypes.

  1. Somewhat related, short but great read over on CityTank today. Putting subsidised housing near transit makes it much more affordable. It would be nice if there was something better than a streetcar serving Yesler, but it’s an easy walk to the Pioneer Square Station.

    1. Yes that is completely related. By putting units next to say, Northgate, instead of near, say, Aurora, we are making housing more affordable.

    2. An easy walk to Pioneer Square Station, but not an easy walk back. That’s Profanity Hill you’re talking about there. It’s also pretty far, and walking across the freeway is nasty (though walking through the ID is pleasant).

  2. Your picture is interesting, because it is from a time that rarely gets discussed when talking about Yesler Terrace. Yesler Terrace, like a lot of public housing projects, was built on top of an already-existing, functioning low-income neighborhood. Remember that this was before the freeway; and while some of that hillside was unbuildable because of steepness and slides (numerous buildings in the area collapsed over the years before they gave up on the lot or developed better techniques), a lot of it was occupied, both above and below the freeway.

    Above, you had a number of blocks of classic Central District housing — overcrowded but solid wood frame houses, some brick apartment buildings, and a street network. You also had shops, and a long-standing historical connection between the neighborhood down below, which once was filled with…Japanese people, who all mysteriously disappeared not long before Yesler Terrace was built. In addition to Yesler Terrace itself, you had the Boren Avenue Extension, which demolished a large portion of the neighborhood in order to bring that street through to Rainier. This project also destroyed the street grid, which was deliberately not reconnected; notice that that stretch of Boren has no cross streets.

    Then of course you had the freeway barrel through there after that.

    The end result was an almost deliberately ruined urban fabric, further cutting off a vulnerable African-American population from the city and a deliberately removed Japanese-American community taken away to prisons in the desert. Yesler Terrace provided rooms to live in but none of the other features of a thriving community.

    I rather dispute the density figures for Yesler Terrace because I think they deceptively leave out the vast tracts of waste ground that surround it and are part of its legacy even if they don’t officially count as part of it.

    1. If Wikipedia’s to be believed, YT was completed in 1941. The Japanese-Americans weren’t evacuated until early 1942.

      1. The location was chosen specifically because the Japantown buildings were considered to be substandard. During construction, they had to move out of the neighborhood. And while in theory existing residents had priority to move back in, those of Japanese descent couldn’t take advantage of that from 1942-1945.

  3. @ Bernie

    Your Bellevue Towers density numbers have an error.

    540 units x 1.5 persons = 810 people
    810 people / 2 acres = 405 people/acre.
    640 acres = 1 square mile
    405 people x 640 acres = 259,200 people per square mile. Not sure how you came up with 13k. :)

    1. Good job of fact checking. My error was assuming 3.4% was a per acre fudge factor and not allowing that it was a 22 acre fudge factor. I’m not sure I’m comforted in knowing that the Spring District proposal is for a density almost 4X that of Manhattan.

      1. The 4x math isn’t right either:

        The Spring District, a 36 acre environmentally sustainable, mixed-use urban neighborhood is planned for the old Safeway Distribution Center in the Bel-Red Corridor. The emerging urban neighborhood will have over 4 million square feet of office space with supporting retail amenities, up to 1,000 multifamily residences, and a 16 acre open space plan that will include a turf athletic field, a large park and several plazas and green spaces

        1000 units/36 acres = 27.7778 units/acre
        640 acres/sq mile * 27.7778 units/acre = 17778 units/sq mile.

        Obviously more than one person per unit. So this is very dense, but not Manhattan density. Seems quite comfortable to me.

      2. The 4X was for Bellevue Towers, not the Spring District and was assuming 1.5 people per unit. That yielded the 260,000 people per square mile which in round numbers is 4X all of Manhattan. If you take the whole superblock rather than just the lot size for the Towers the density would be 1/5 or 52,000 people per square mile. Not Manhattan but pretty close to Paris. What’s the population density of Rome?

        From 1986 to 1995 the city annexes land totaling one-third of its present size, reducing the population density from over 1500 people per square mile in 1986 to just over 1000 in 1995 Rome has a population of 2,546,804 (2004)

        Population in 1995 was 2,650,000 and is currently ~2.7 million which, assuming no more annexation wouldn’t make much of a difference, ~2%. But that same source says the city density is 4,698 inhabitants/sq. mi. It pegs the 1,000 number as people per square kilometer for the metro area.
        But even if this second set of numbers is right and you counted all of the current population and don’t account for the increased land the density would only be ~7k per square mile. Rome is a huge ass city, 579 sq.mi. but less dense than Seattle???

      3. It’s super weird comparing things like a block to a city. Cities have things like parks and roads, blocks usually don’t have parks, roads and buildings. Neighborhoods usually do, so it’s less weird to compare the Spring District to Yesler Terrace, for example.

        It’s always weird to compare different geographies when you are talking about cities, because political boundaries are very different than human boundaries. 1st Ave NE and NE 146th isn’t that different from 1st Ave NE and NE 145th, but one is in Seattle and the other in Shoreline.

        In this case, 561 sq miles is just too massive to make sense compared to, say, Bellevue or Seattle. San Antonio and Phoenix are like that: they have huge limits that include tons of suburban areas, and even rural and wilderness lands in the city limits.

      4. The reason I went with a superblock is that it’s unlikely that any other 40 story buildings will be sited there. Of course there will be lots of other low rise construction that may include housing units. Bellevue’s pretty big on maintaining air space and avoiding the concrete canyon feel. Almost half of the Spring District is slated to be parks/public space.

      5. Yeah, 16 acres of open space for the spring district. Sounds lovely, to me. Exactly the sort of thing that shuold be going up around Northgate.

      6. I think the Spring District is a better master plan than Group Health’s clear cut proposal in Overlake:

        Health representatives highlighted the project’s mix of residential, commercial office and retail space. Their plans include 1,400 new residential units, 1.18 million square feet of commercial office space, at least 25,000 square feet of ground-floor retail, and an 180,000-square-foot 180-room hotel.

        Plans also call for a 2.67-acre park and a central pedestrian walkway running from the Microsoft campus to the light rail station.

        Although I do think it’s a better balance of office vs residential. I blieve the only building that will be more than 6 stories is the hotel. That’s why you only get ~10% for parks instead of almost 50%.

  4. Andrew – You are spot on that we can house most of our population growth on a tiny fraction of urban land. Last year I wrote a comparison of Yesler Terrace and the Cascadia planned community on my blog (click on my name above to see).

    A note about the population densities: Yesler Terrace is listed as 30 acres on the SHA website. This value produces ahousehold density of 106,700/sq.mi. For comparison, one of the census block groups in Belltown had a population density of 110,800 ppl/sq.mi. in the 2010 census. So population densities at this scale are possible and not unprecedented in Seattle, for small areas. It is difficult to keep population densities so high over larger areas, such as Manhattan, due to unpopulated areas such as parks, hospital complexes, and commercial and industrial areas. However, it is quite feasible on a few intensely developed blocks.

    1. We’re both correct about the size, and both wrong. Yesler Terrace is 22 acres today. The new version will be 30 acres.

    2. Well, and this might be key.

      Maybe a natural way for humans to live is in archipelago type densities.

      Essentially towns and villages.

      But these require smaller sizes — not the end to end density of a large East Coast city — and give way to a lot of green space.

      So, not suburban sprawl, but not vertical density either.

      Just a kind of medieval town structure.

      And this would play well into a transit type infrastructure, with TOD at each station.

      However, the trade off for making people cluster would be to surround them with nature, parkland, beauty.

  5. From my read of Gardener’s comments I think you’re misinterpreting his opinion. I’m not sure he thinks it is physically impossible to fit the planned housing and office on the Yesler Terrace space. I think his concern is that it will be difficult to attract the necessary private development to make this project work within the 20 year time frame.

    TO be honest if the facts in the story are correct (which is entirely debatable knowing the ST) then I think its a valid concern. Yesler Terrace may not be a slum, but it’s no Capitol Hill. Its no LQA, Fremont, SLU, Belltown, U-District, or Ballard for that matter either. Even with huge upzones and perhaps some public incentives, I think it will be a hard sell to developers that they can make reasonable returns on these properties. Especially since its going to be a mixed income community, which unfortunately limits your ability to sell luxury.

    Belltown may have attracted 22% of the development in the 90’s but today I think the urban market is much more diversified. Capitol Hill is probably the hottest neighborhood right now and it probably isn’t attracting much more than 10%. For YT to get 15% is going to take some real marketing savvy. Or at least some very strategic anchor projects in the beginning. I mean, as urban as it is, the place is a virtual desert when it comes to walkability.

    1. Gardner and other bearish real estate commentators are underestimating the epochal shift from suburban single-family to urban multi-family housing. The number of urban apartment units in the construction pipeline, even excluding Yesler Terrace, is unprecedented for Seattle (projected for 6,000 units in 2013 and 2014). However, for the past 20 years the Seattle metro area has consistently added at least 20,000 housing units per year (except since 2008). Seattle multi-family has increased its market share of Seattle metro area development from 9% in the 1990s to 30% the past few years. If Seattle multi-family’s market share increases to 40% or 50% (10,000 multi-family units per year in a good economy), then there will be more than enough demand for every Seattle multi-family project proposed, even Yesler Terrace.

      If Yesler Terrace’s 5,000 units were sold over a 10 year period, that is 500 per year. If the construction rate of 6,000 multi-family units per year continues, Yesler Terrace only needs to capture 8% of the Seattle multi-family market each year. And a lower percentage if it is built out over 20 years.

      I understand why the naysayers are worried about apartment overbuilding, but I believe that the market fundamentals have shifted. We won’t be going back to constructing 80% single family homes (like in the early 2000’s) even when the recession is over.

      1. Where are you pulling these numbers from? Quickfacts Seattle’s population only grew by 45k between the 2000 and the 2010 census. If 20k housing units were added every year for even 6 years there would be an excess of 100k housing units (2 people per household avg.). Seattle only has 300k housing units total. Are you smish smashing Seattle, King County and Puget Sound? I find it really hard to believe 80% of new housing units built in Seattle in 2000 were SF homes.

      2. Historically there were 20,000 housing units built per year in the Seattle metro area, defined as King, Snohomish and Pierce Counties. I am comparing total housing construction in the 3-county area to City of Seattle multi-family construction. My numbers come from the Puget Sound Regional Council (PSRC) and the multi-family projections come from recent press reports.

      3. Mixing construction in the 3-county area with City of Seattle multi-family construction isn’t very meaningful. The Seattle Comprehensive Plan goal was 47,000 households over 20 years. The 5,000 units in Yesler Terrance would be 10% of all housing in the City over that whole time frame. It doubles if you have to sell over a 10 year period.

        Net single family housing in Seattle proper is on the decline. You pretty much have to tear down a SF home to build a new one and even if you redevelop some areas with smaller lot sizes, dulplex on one lot for instance, that’s more than offset by the number of SF lots rezoned for MF. As you push further out it’s essentially the opposite. Of the 6,500 units recently permitted (Master Plan approved, not active permits) out in Black Diamond it’s almost all SF. Ditto for Redmond Ridge, Snoqualmie Ridge, Monroe, Marysville, et al. The 2010 census indicates a continuing dispersion of population in the Seattle metropolitan region. All of Seattle’s MF housing is only accounting for 9% of the regional growth. Seattle and inner suburbs only 39% and there are still plenty of SF homes being built in Kirkland, Redmond and Bothell/Woodinville. So 70-80% SF construction for the region sounds about right. over the next 10-20 years.

    2. Josh, when SHA first started talking about Yesler Terrace that was one of my big concerns. Knowing the history that Fnarf identified above (particularly the destruction of the street grid), it seemed obvious that any rebuild/density increase of Yesler Terrace needed to involve re-knitting the grid. I recognize Boren and I-5 aren’t going away anytime soon, but better pedestrian connections to the ID/Little Saigon to the south would be a huge help. I’m not sure there’s much hope for the I-5 crossing at Yesler, and headed east involves crossing both Boren and 12th. At least 14th will be calmer once the streetcar project goes in.

      1. The walk down Yesler is quite scenic, actually. And the stop sign on the west side keeps traffic fairly calm. But to really connect Yesler Terrace to the International District, I’d replace the stairs next to I-5 with a public escalator. The slope up Jackson is quite walkable, you just need a little boost to make it up to Yesler.

      2. Yeah, Matt, the views are nice. But it’s awfully loud :) I thought about an escalator too, a la Hong Kong. Don’t think SHA has that in their plans either.

  6. Here’s something interesting…while you’re trying to figure out how to make Seattle into New York City — NYC is stealing your thunder by becoming a high tech, web startup center!

    New York City: The Nation’s Second Leading Tech Hub

    New York City has emerged as the nation’s second leading high-tech hub, according to a new report by The Center for an Urban Future. “In 2006, I wouldn’t have put New York anywhere on the map,” Silicon Valley high-tech and venture capital expert, Vivek Wadhwa, told the authors of the report. “If there is any second to Silicon Valley, it’s now New York not Boston.”

    Since 2007, 486 of New York City’s’s high tech start-ups have received venture capital investment, mainly in the fields of the Internet and mobile technology. Of these, 15 have raised more than $50 million in venture funds, 27 have attracted at least $25 million and 81 have raised $10 million or more, according to the report.

    When I was back in NYC last year, I heard that Brooklyn was taking some of the area around the old shipyards, for example, and turning it into high tech offices. Very cool and an interesting place to be and work.

    So, Seattle – you’ve got competition for the whole “hipster techie” thing.

    1. So you’re saying that NYC is trying to copy Seattle’s startup culture? :)

      NYC was inevitably going to come out near the top when it set its mind to it, because of its large population and infrastructure. Seven million people in the city and ten million outside means a significant number of them will be tech-entrepreneurs even with average percentages. And with its extensive transit infrastructure, even when things are broken people can still get around better than elsewhere. Extensive industrial districts in Brooklyn and Queens allow companies to locate next to their partners and suppliers (both traditional manufacturing and new tech).

      Regarding Silicon Valley-like startups in New York, what I’ve heard is that the financial sector is so large that it has dominated the computer industry in the region so that’s where most of the tech jobs have been. Tech-wise it’s very traditional because banks are fussy about numbers and proven technologies, so that’s what most of the tech jobs are like. Silicon Valley is totally different, its economy based on Internet startups and the like. But if the number of startups in New York has recently increased enough to be noticed, that’s not surprising given the region’s long history as an incubator of new industries.

      The Battle for Gotham” is relevant to this. Roberta Gratz writes a third-party account of the Jacobs-vs-Moses years in New York, and her own life is a concrete example of things Jacobs asserted in generalities. Gratz’ father had a drycleaning shop on the west side that was displaced by a Moses project, and the family moved to Connecticut in the early days of suburbanization. (She says a lot of people moved to the burbs not because they loved tranquility to but because they were forced out of their neighborhoods by urban renewal.) When she grew up she married a metalworker who had his own manufacturing company in Manhattan, but again they were displaced by another Moses project. This time they relocated to Brooklyn, where she says the synergy of having several companies on different floors of a building or in neighboring buildings, all close to the subway, created lots of opportunities for her husband to make metal things for the surrounding companies (theater groups, retailers, etc), and conversely to find things his company needs right in the neighborhood, and for people to collaborate on new ventures, etc. This is the kind of environment that leads to startups and small businesses, so it’s not surprising that high-tech companies have finally grown there too.

      As for turning obsolete industrial blocks into high-tech urban villages (or manufacturing urban villages like Gratz’), that has been happening for years in New York and other places. Hers, and my, main concern is that some of the manufacturing capacity remain, even if it’s converted to cleaner industries. San Francisco and Vancouver have completely replaced their industrial districts with condos and retail. That’s good for walkability and non-sprawl but it makes their economies dependent on a narrower range of sectors, and it gives no space for growth in case local-manufacturing or urban-agriculture has to be urgently revived someday (if long-distance shipping and food transportation comes to a halt).

Comments are closed.