seattle from smith tower (with needle)
The Dexter Horton building in the foreground, with the skyscrapers in the background. Photo by Jason Brackins

I have two thoughts about density I’d like to share, both more responses than anything* to ideas that I hear and read repeated frequently. I have made both of these points a few times in round-about ways in various posts and in the comments, but I would like to write a first-class post about them this time. The first point is a response to the oft-repeated notion that tall buildings and density are essentially the same thing, and the second is a response to the idea that what’s good for developers is perforce good for density and/or urbanism. Both below the fold.

The “tall buildings are density” refrain is so common that Matt Yglesias, a serious proponent of density, often uses “tall buildings” as a verbal substitute for density. Of course there is a correlation: for any given building square footage and plot square footage, there’s a minimum building height that is required. Something I think many people don’t realize is that very tall buildings are often not much more dense than shorter buildings, especially older buildings, when you compare floor-area-ratio (this is a pretty good visual explainer for FAR, if you are unfamiliar). I’ve made a chart of several historic buildings compared against the tallest** buildings in Seattle by both height and floor area ratio – the data is from King County assessors. I look forward to learning if anything is wrong with it in the comments (I mean that earnestly, no sarcasm).

Building Height (ft) Lot Size (sq ft) Total Building Area sq ft FAR
Dexter Horton 185 28,200 313,380 11.11
Exchange Building 275 21,240 295,515 13.91
Hoge building 200 6,480 96,076 14.83
Medical Dental building 261 24,820 292,038 11.77
Olympic Tower 148 12,217 116,651 9.55
Washington Athletic Club 232 21,000 274,070 13.005
Columbia Center 968 59,266 1,526,621 25.76
1201 Third Avenue 772 56,400 1,114,847 19.77
Two Union Square 740 89,950 1,137,666 12.65
Safeco Plaza 630 57,110 754,455 13.21
Space Needle 605 15,000 19,320 1.23
Russell Investments Center 600 54,021 940,000 17.40
US Bank Centre 580 57770 931243 16.12
Wells Fargo Center 573 56,400 940,648 16.68
Bank of America Fifth Avenue Plaza 543 61,440 904,784 14.73
Union Bank of California Center 536 49,755 558,595 11.13
Rainier Tower 514 47,500 538,529 11.34
Qwest 536 42,360 531,771 12.55

Certainly all of these “old” buildings are tall buildings, even the Olympic Tower. But none of them are what people think of as tall when they think of downtown, they think of the buildings on the second half of the list which is my point. I doubt anyone would have guessed the squat Dexter Horton building was as dense as the Bank of California tower (that shocked even me), or that the Hoge building is more dense than all but the most dense of the tallest buildings downtown, but that is the case. If you want more density you don’t  want taller buildings per se, you want more generous FAR. This distinction becomes pretty marked when you consider that more people live and nearly as many people work in South Lake Union compared to Downtown Bellevue, even though SLU doesn’t have a single “skyscraper” and is half the geographic area.

Of course developers want to build taller buildings, but it’s not just because of density: higher floors with better views are worth more money. Developers will take height without density as a matter of course, which is one reason height is so contentious in “South Downtown”. Many of the proposed taller buildings will not have higher floor-area-ratios than existing historic structures after the upzone. Without historic protections (different from character protections, but these get lumped together usually), these taller buildings may still be built on the sites of older ones because a new, 15-story building with parking is worth more than an old, 7 story building without it. Still, if they both have a 6 FAR, that’s not more density (in fact it could be  less because of the parking).

Which goes to my second point:  we shouldn’t take as a given that what is good developers is good for density. This was basically the premise of Martin’s post about the waterfront, at least as I read it, of course I hear it a lot and read it from other people as well. Why should we trust developers as the arbiters of urbanism or density? At one point a developer in Seattle tore out this to build a parking lot. Just recently a developer started building this, though admittedly that was a parking lot before the construction. Developers’ primary concern is to make money, sometimes that means building more density, sometimes that means building a parking lot. Why should we make “redevelopment” an urbanist goal in and of itself and fight for legal changes that may not add density on their behalf?

One last (bonus!) point. I work in a building in Pioneer Square that has a FAR of 5.8, which would be near the max allowed in the South Downtown upzone (at least as I read it). The building used to house some Microsoft offices, but they moved them out of the building to Bellevue because the building didn’t meet Microsoft’s parking and conference room standards. Because of this, now seven start ups occupy the 6 story building, and that’s after another moved out. When all eight were in the building, it housed about 440 workers in 33,000 square feet. When Microsoft was in the building it was 140 workers in the same space***. I imagine a developer would love to develop a building that would meet class A criteria and attract Microsoft or the like to the site. But would that be more dense? Where would the 11-person companies like mine go to?

We all love tall buildings, but there’s a lot more to think about than just height and what developers want when you are considering density. More height doesn’t necessarily mean more square feet, and more square feet doesn’t even necessarily mean more people. And that’s what we mean when we are talking about density, isn’t it? If we don’t pay attention, we might end up fighting on the wrong side of the issue.

* I would have written this in November, but I spent the month doing NaNoWriMo, and didn’t have the time or the appetite to write a lot at the time.

** I could not find records for the Municipal Tower. Maybe  the city doesn’t pay property taxes on it?

*** 140pp/33,000 sqft is actually significantly more dense in a worker/sq feet sense than the Redmond campus. I worked in a building that had 67,000 sq ft of office space (not including the massive parking garage) and had just 108 people working in it.

93 Replies to “Thoughts About Density”

  1. The really great buildings have a sense of space both on the inside and outside with exteriors that don’t start and end at the sidewalk and interiors with a grander view than the elevator bank. The library takes that to the far edge, but you get the point. That screws up the F.A.R. but who’s counting? Living and working in cubicle world isn’t all it’s cracked up to be.

    1. …with exteriors that don’t start and end at the sidewalk…

      Um, what does that even mean?

      Yes, there are some buildings that offer totally amazing foyers to extend the public space from outside to in.

      But there are many more amazing buildings with amazing exteriors that do, in fact, start and end where the sidewalk does.

      Do you really want to argue that working in the to-the-lot-line Trinity Building would be more drab and sterile than working in downtown Bellevue’s setback-o-ramas?

  2. I think most people here can agree that short and fat can be the same as tall and skinny when it comes to density. Especially those who have been to Quest and ordered a large beer. :D

    What comes out to me is that you are comparing old short and fat to new tall and skinny. Why is there no new short and fat? Is it economics, regulations? And just as importantly can anything be done to fix it? What? How do we do it?

    Because it seems to me that at the moment we have a choice between tall and skinny and short and skinny, and in that case I’ll take tall and skinny any day. Figure out a way to get short and fat one of the choices and then we can talk, until then any kind of height restriction is a density restriction.

    1. There are loads of short and fat, actually. The Amazon buildings in SLU are great examples of this. I picked old ones specifically because of comments like this:

      Are we going to run out of historic architectural examples in that area?

      Tall and skinny is all good and fine. When someone starts talking about tearing down historical buildings to kick start redevelopment, something both Martin and Roger (mostly in his blog, but all the same) have said my point is that you had better make damn sure you are actually going to get more density.

      Personally I’d take the Dexter Horton over most tall skinny towers, but if I’m talking dexter horton and you’re talking rainier tower, we’re talking about something other than density.

      1. If we can buld short, high-FAR buildings with our modern code, we can build tall, higher-FAR buildings. I buy your arguments about why to save older buildings – they just don’t build ’em like they used to, and it’s hard to build anything right to the sidewalks with no parking and skinny storefronts these days. But can you at least agree that all things equal (in this case, modern building codes), tall is better than short? Tall will always add more density if you don’t change the footprint.

        (and I hate the “skinny” vs. “fat” terminology. I love skinny buildings – if we’re talking about being physically narrow but still taking up all of their property. skinny buildings both make for interesting blocks and can allow for good natural ventilation and daylighting.)

      2. I love tall buildings. My point is just if you are going to make a principled position that we want density at all costs (as Martin and Roger have done), then you need to consider density and not height. You can’t just lazily assume taller is more dense.

      3. @Matt the Engineer

        “Tall will always add more density if you don’t change the footprint.”

        Yes mathematically that is absolutely correct, the thing is that zoning codes almost always don’t do that. The rational behind many zoning codes is if the city is going to let a developer build higher, for the benefit of that height the developer must build a skinnier tower, which doesn’t maintain the same footprint (at least on upper levels). Essentially, they think, we’ll give you more height if your buildings create less shading, are less bulky, preserve views better, and have more spacing between towers, etc.

      4. Sure. So I’ll put it another way. Tall buildings will always add density if you don’t change the footprint of the lower floors. Even when you deal with silly codes requiring skinny buildings up high, they don’t change the building down low. Take any building and add a skinny tower on top, and you’ve raised the density.

      5. @Matt the Engineer

        Yeah I totally agree, the Trace loft is somewhat along those lines. They added two floors on top of the existing historic structure. There aren’t however a lot of examples where a tall/skinny and short/fat building have been combined. Can’t be done, I’m just saying codes and precedence aren’t really on the side of a hybrid design.

      6. Taller buildings require more space for elevators, especially above 40 stories where you have to stagger them. That eats away at the usable space. There would be plenty of office space and housing if buildings were no larger than 10 stories or 40 stories, although of course some existing buildings would have to be taller than they are now.

      7. Slow down a minute. Catagorize new buildings by number of floors. Find any building with number of floors X where the average density is higher than the average building of floors X+1. I’ll bet this doesn’t exist.

        What I’m saying is that 41 floor buildings aren’t less dense than 40 floor buildings, despite having an extra elevator (the floor space in that one extra floor more than makes up for it). So why do we care? 41 is better than 40. And 42 is better than 41. (yes, and high FAR is better than height. but we’re now talking about height)

        “There would be plenty of office space and housing… of course some existing buildings would have to be taller than they are now” That’s not a small difference. Yes, Paris works. But we’ll never be Paris – we have wide streets, parking lots in buildings, short buildings, and parking structures. The time it will take to replace more than a vast minority of short buildings and parking structures will be on the order of generations.

        The common NIMBY cry of “but we have enough zoned density now” is a red herring. It assumes we can just bulldoze all of the less-than-maximized buildings and rebuild them – we can’t and we won’t. A tall building in the hand is worth three short buildings in the bush.

      8. I said 40 stories was acceptable. I didn’t say everything had to be capped at 6 stories.

      9. A bit of an aside, but the Trace Lofts have followed an age-old tradition of tastefully expanding a building upwards a floor or two at a time, with contrasting but complementary architecture, while not disturbing the integrity of the original structure.

        Manhattan is full of late-1800s buildings with obvious additional floors tacked on in the early 1900s.

        The 1927 Ballard Building (north side of Market) has an ugly mid-century additional floor on it, but it is inconspicuous enough not to look too bothersome from street level. (It’s still much better than if they’d tried to build a cheap faux-1927 addition.)

        The Trace Lofts, wishing to market the historic “loft-y” architecture, did a fantastic job of aesthetically merging old and new.

        But make no mistake, this is different from the facadist tendencies that most developers (especially ones looking to build high-end towers love to indulge.

        Facadism is a horrifying failure most of the time. What’s worse, it tends to blank-plate-glass over what used to be multiple entrances and storefronts, so the street level loses its activating features anyway. And even worse, cities tend to accept this as a valid preservation compromise.

        Zone to allow modifications and a few extra stories in Pioneer Square, and you might get more Trace Lofts. Zone to allow free-market-sculpted towers with no protections for anything but the present facades, and you get this. (Unless the economy stalls, in which case you get this).

      10. Adding one floor increases capacity. But we’re not talking about +/- one floor. If you add five or ten floors, you need a larger elevator or another elevator. That’s where it eats into the usable space. You gain some capacity, yes, but not as much as if the lowrise elevator were just extended. So a 40-story building has less than twice the capacity of two similar 20-story buildings. Above 40 stories, you need staggered elevators. Each elevator requires extra space above and below it for the drive mechanisms. So the optimal size is around 40 stories for a highrise downtown, and 10-12 stories for a midrise in Capitol Hill, the U-district, Roosevelt, Mt Baker (station area), etc. (Note the upzones that would be required to achieve this. Assume existing downtown skyscrapers would remain, and that 40-story zoning would be expanded as necessary for the demand.)

      11. But you’re not describing why this is “optimal”. Sure, adding 20 more stories doesn’t add as much space as the first 20 stories. So what? Do you have another empty piece of land to put those other 20 stories? In a land-limited situation (which we have), building up always gives you more capacity than not building up.

        Taking your logic to the extreme and optimizing for space efficiency, we’d have nothing but 1-story buildings – they’re always more efficient than 2-story buildings (no stairs).

      12. This is not just a thought experiment, Matt. The world is chock-full of examples of six stories as ideal efficiency.

        Six stories end to end is how Paris maintains a population density that averages 54,300/sq mile (even though much of its 40.7 square miles is non-residential space).

        Much-lauded seeder of tower-forests Vancouver only achieves that kind of density in one tower-forested neighborhood: the West End (a single square mile, with pretty much all built space taken up by residential use).

        As for “those other 20 stories”? There’s plenty of room for them next door.

        Most people, even urban-phobic Seattleites, can swallow the idea of 20 stories to the lot line in their commercial downtowns. But most Americans can’t swallow 40 stories without setbacks and plazas. Thus, the former can be built on smaller lots, leaving more small lots for developments of similar size and increasing density overall. The latter only get built on giant lots, giving them the frequently-lousy FAR that instigated this very thread.

        Even in New York, very tall on a tiny lot is the exception rather than the rule.

      13. (I cited population density rather than office-space density because, while FAR is tracked on a per-building basis, its numbers over large areas aren’t as easily accessible as pop-density numbers. But the principle of mid-to-high-rise providing more usable space per square mile than high-rise-with-setbacks carries easily between the two uses.)

      14. I firmly disagree that we have a land shortage. There is an astronomical amount of land in Seattle that is zoned for single-family, despite being immediately adjacent to major commercial centers.

        For example, large swaths of Capitol Hill and LQA are zoned for lowrise, residential small lot, or even SF5000. If we upzoned it all to midrise (like the Western Slope), or mixed-use (like SLU), we would have more buildable land than we know what to do with.

        And I think that would be way better than building a few blocks of Dubai, and leaving the rest of Seattle as the suburb it currently is.

      15. But I’ve heard we lack openspace! There’s a terrible, horrible shortage of openspace! I have to walk so far already to find openspace, and I never seem to pass anything of value on the way (must be because we have so little openspace)! Can’t we at least write into into the zoning rules that we must always add openspace? We must add infinite openspace!

        Won’t somebody please think of the children openspace!!

      16. Now we’re getting into strategy. Can we convince people to let go of SF zoning? Can we remove FAR limits? I don’t know the answer to either.

        We’re also wandering away from a general discussion (tall buildings or short?) and toward the specific – should we be expanding urban villages with mid-rise buildings, or should we be fighting to remove FAR limits? My choice would be both. But that discussion could fill a post (or 20) of its own.

        Re: Paris. Yes – 6 story buildings can work very well. But it’s not like we have a blank slate and are deciding how to build an entire city. We get individual projects in a built-up city filled with existing FAR-limited buildings, open space, parking structures, etc. Is it better to limit these new buildings to 6 stories, or would 7 be better? Or 10? Or 80? If we can’t convince people to remove FAR limits in skyscrapers, I’m fine with fighting for as tall of buildings as we can get away with without FAR limits. But let’s not limit ourselves to 6 stories.

      17. “Do you have another empty piece of land to put those other 20 stories? In a land-limited situation (which we have)…”

        We don’t have a limitation of land, we have a limitation of zoning. One option would be to make Seattle like Manhattan. Another would be to allow 12-story buildings in neighborhood centers; i.e., make Capitol Hill and Mt Baker like First Hill. A third would be to extend multistory zoning into single-family neighborhoods by a few blocks, perhaps a gradually-expanding increase over several years. A fourth (which I don’t advocate without further analysis) is to allow highrises in SODO. (My hesitation is that we’d fundamentally lose Seattle’s manufacturing capacity as Vancouver and San Francisco have done, and we may need it in the future when intercontinental trade becomes too expensive.)

      18. If we can’t convince people to remove FAR limits in skyscrapers, I’m fine with fighting for as tall of buildings as we can get away with without FAR limits.

        Do you realize how crazy that line of thinking is? So if we fail to educate the public and their representatives, if we just accept the endless “derp-derp-need-openspace” kneejerking-on-the-basis-of-nothing that afflicts even Seattle’s proudest self-identified liberals (see: any Stranger blog entry on the Waterfront), then we should instead allow more of our downtown to be decimated and replaced by inhospitable frontage-free wastelands à la the financial district?

        Do you want to live in Dallas?

      19. I’ve perhaps made the argument too simple. I think it’s a valid viewpoint to not allow sidewalk-to-sidewalk streetwalls to extend to thousands of feet in the air, everywhere in the city. I think it’s reasonable for people that are afraid of losing sunlight (for the 3 days we get any), to ask for setbacks after some number of stories – but the bottom stories should always be allowed to take up every square inch of their owned space.

        I don’t think open space is a good idea at all downtown, except in very controlled places (and for the next century or two, we have all of the open space we need in our existing buildings’ wasteful plazas). I should have separated the two arguments.

      20. Ah, but now we’ve completed the circle and have returned to Andrew’s precise point.

        A single, set-back tall building on a single gigantic base won’t necessarily have any higher FAR than the multiple mid-height structures that could (or perhaps already did) exist on the same footprint. Even if you started mandating full sidewalk contact, you are never in a million years going to get the activating multiplicity of street level uses that multiple buildings with multiple entrances provide.

        And good luck convincing developers to build a tower without a garage, so there goes some chunk of your frontage to ramps and fans and empty slabs no matter how much mitigation you try to require.

        Vancouver has succeeded in keeping the frontage on its major commercial streets as activated as any new-construction tower-forest could possibly be. You have no idea the amount of regulatory detail the city has required to make it so. None of that is ever going to happen in Seattle. And still, if you go half a block of Robson, street activity takes a nosedive.

        Even the tightest tall-construction regulations in the world can’t mimic what develops naturally from using many small lots!

  3. FAR isn’t the whole story either, though. Newer buildings make better use of support steel and need fewer internal structural walls, allowing open floorplans that older buildings can’t handle. Open floorplans can fit more workers comfortably into the same square footage (or not, depending on how cube farms are set up). This is one reason older buildings (e.g. the Cobb Building) get converted to residential — it can be hard to rent them as office space.

    On a side note: I wonder how much pressure Microsoft HR puts on the real estate group about needing Seattle offices. Almost all the young people I knew when I worked there lived in Seattle, and almost all of them eventually quit. The lack of Seattle offices probably wasn’t the only issue, but it was definitely important.

    1. Great piece. The stat about SLU was very interesting. At the rate they’re building new midrise there, it might eclipse Bellevue in a few years. And I know for a fact it already has easily the highest number of Tom Douglas restaurants per acre of the entire world.

      1. Thanks.
        I bet SLU will pass in terms of jobs soon, because there’s a lot more development going on in SLU than in DT bellevue at the moment, notably one more huge amazon building.

      2. SLU and Belltown have the *only* Tom Douglas restaurants in the world, don’t they? He’s a legend in his own town.

    2. We’ve got an open floor plan in my 120-year-old office building, so I’m not sure exactly what you mean. The cobb building was a medical-dental building, which means those walls were there because of the type of building it was, not simply because it was offices.

      As for the MSFT thing, during the time I worked there, they closed nearly every office building they had in Seattle and moved them to Bellevue or Redmond. I think there was a concerted effort to not have Seattle offices.

      1. Hmm. I’m not a structural engineer, but my understanding was that pre-1960’s buildings support significantly fewer tenant configurations. I thought tenant configurability was also part of the of class A/B/C office space designation (where class A is generally more valuable than class B and class B more than class C), though I’m pretty ignorant about the details.

        On Microsoft: that was also my sense when I worked there (1998-2004) – I know people who specifically joined the Visio group to avoid having to go to Redmond and who were pretty upset when Visio got moved to Redmond. In my exit interview, I got the sense that wanting-to-work-in-Seattle was a common story, and that the lack of Seattle offices was driven in part by facilities guys not wanting to have to deal with people on both sides of the lake. I think the downtown Bellevue office space was seen as a compromise, though I don’t think it’s really a good one. But I don’t know the internal story at all.

      2. That makes perfect sense about the configurations.

        That kind of goes to my point. You can imagine a world like this:

        Class Rent/sq ft Marginal Cost/sq ft
        A $30 $10
        B $20 $10
        C $10 $10

        In this world no one would build more class b ir c (which is a lot like ours) but developers might like to tear down class c to get class A, even if you don’t get more density, because paying $20 per square foot makes you more money than keeping the class c around.

      3. No. This is the same argument people use for the residential side: the thought that if you build a new building, it will be worth more and rents will go up overall.

        This is the exact opposite of true, as long as the new building has more square footage. Basic economics says that increased supply drops prices. Add more class A space, and some of the lower grade class A space drops in value.

        I’ll debate in other threads whether adding height adds density. But there should be no argument that increased density drops prices overall.

      4. This is basic economics. In my example, it’s profitable to switch Class C to class A as long as class A rents state above $20/sq ft.

        “rents overall” don’t have to go up, it only means that rents for Class A have to be higher than rents for class C.

      5. Actually, in this case rents overall would go up, because class A are higher than class b.

      6. Not necessarily Andrew. To increase those rents by building new, you’ll sink hundreds of dollars per square foot into it. Many building owners will prefer, often correctly, to keep the low rents and avoid a big investment that might not pay off. Or they’ll spend just enough to keep it in the $15 range rather than $10 for example, by keeping outdated systems working, maintaining the roof, doing a seismic upgrade at some point, etc.

      7. But your example isn’t real. You made it up.

        In reality, if you greatly increase the amount of Class A space, the rent you can get for Class A space greatly decreases. At some point, you run out of people looking for Class A, and you’re competing with those that typically look for Class B, at the prices they’re willing to pay.

        The best way to look at this is to look at the poorest company that can afford to operate in Seattle. Build a bit more space – of any class – and one more business can afford to move in. If you build it at the top end, the worst A class space effectively become B class space, since the richest B class tennant can now afford to move up. Etc.

      8. Not necessarily Andrew. To increase those rents by building new, you’ll sink hundreds of dollars per square foot into it. Many building owners will prefer, often correctly, to keep the low rents and avoid a big investment that might not pay off. Or they’ll spend just enough to keep it in the $15 range rather than $10 for example, by keeping outdated systems working, maintaining the roof, doing a seismic upgrade at some point, etc.

        I suspect that this is what happens most of the time in reality, which is why 2nd is still full of parking garages, and there are plenty of surface lots sprinkled over downtown. But if this is the case, then increased heights won’t do anything, so it’s pointless to talk about heights. Increased FAR might do something, but heights wouldn’t.

        Which puts me to Matt the Engineer’s point (which may not be made up but is significantly more unrealistic). If you are building at the same density to make more money, you are by definition raising rents. Right? That has to make sense to you. Why do you believe that will lower rents elsewhere?

        This is not a zero-sum game (obviously, or there would be no development). The markets for cheap office space, the sort of places start-ups go to are very very disconnected from the office space large companies go to. There are a lot of reasons for this, but one thing that is obvious is that the prices paid are very different.

        All of this is distracting from the main point of that paragraph. Rents are what developers go for, but Rent doesn’t necessarily equal density.

      9. You’re in the Cobb? My parents and I all went to the same dentist there, and I saw the older one retire, and the younger one go completely through his career from college to retirement, when he sold the practice to somebody else. They had to move when the building was converted. The elevators in the Cobb are something else. I wish I had an excuse to go into that building now.

      10. I agree that rents don’t equal density.

        “If you are building at the same density to make more money, you are by definition raising rents… Why do you believe that will lower rents elsewhere?”

        I’m just saying that increasing density drops rents, no matter what level you build that new density to. I agree improving a space without increasing it will raise the rent of that space.

        I certainly don’t advocate tearing down old buildings and putting in the same density.

      11. I certainly don’t advocate tearing down old buildings and putting in the same density.

        That’s what some developers have been advocating for, hence the zoning code.

      12. I’d like to see a bit more explaination on this. I’ve certainly heard the plan to increase heights, but I haven’t seen evidence that the new buildings in S. downtown would be less dense (though I admit I haven’t spent more than a few minutes with that massive PDF). Are there specific pages or examples you can point to?

      13. Microsoft has a building in SLU and a couple of spaces in Pioneer square. The SLU space and the “labs” space in Pioneer Square are there for the long haul. I believe anything else they have in Seattle is as the result of an acquisition and will be discarded as soon as the group is moved to either Redmond or Bellevue.

        Other than some specialized offices I don’t think we’ll see a big MSFT footprint in Seattle. Microsoft is nearly out of space to put new buildings on their main campus. Their longer-term strategy seems to be to spread out a bit on the Eastside with most of the new off-campus office space going in downtown Bellevue. I also wouldn’t be shocked to see them take office space in any redevelopment of Bel-Red, Midlakes, or Overlake Village (though this would be more of a case of extending campus South and West).

        The cross-lake commute is a huge retention issue for MSFT and it is only getting bigger.

      14. Microsoft is nearly out of space to put new buildings on their main campus

        There’s still loads of space in the old safeco campus they bought, and more room on main campus near building 37

      15. I think — well, I *hope* — that Microsoft will eventually notice the massive attrition to Amazon and other companies in Seattle, and will decide to create a Seattle office to attract and retain young talent.

        It’s a shame, since in many ways they’re a fantastic company, but they just don’t seem to realize the attraction of being able to walk to work.

  4. Great post, I’ll pass this around SLU. By the way it would be great to link the building names to the assessor’s site (e.g., 1201 Third Ave) if you have that handy.

    And speaking of the Amazon campus, one is at
    301 Boren Ave N: 12 stories, 282162 Building Net Square Footage / lot size 14400 (so FAR of 19.59)

    1. That FAR can’t be right. An FAR of 19.59 for 12 stories either implies a building with serious overhang or at least 7.59 stories of underground parking.

      14400 square feet seems pretty small for that lot — 14400 square feet is about 3 single-family housing lots.

      1. I’m no FAR expert. Maybe there are actually 2 or more buildings. It definitely has a section that’s open on the ground floor.

    2. That building actually sits over 4 different parcels, with a total lot size of 56,690 sf. That would yield a FAR of 4.98. (The building actually extends over the alley too, which is technically public ROW and so isn’t included in the lot size). FAR, by definition, cannot be greater than the number of stories in a building, unless (like Steve points out) you have overhangs.

      1. FAR, by definition, cannot be greater than the number of stories in a building, unless (like Steve points out) you have overhangs.

        You can have underground floors. Many buildings in pioneer square have semi-subterranean spaces. The buildings will look four (or five or six or whatever) from the street, but will have an underground space and often several set-back floors at the top you cannot see from the street.

        Many of the underground spaces are used for restaurants, or in the case of my building its bike lockers, server racks, showers, etc.

    3. Newer buildings are harder to measure FAR from the assessors’ data because there are complicated “height transfers” going on. With the 1918 building in the Denny Triangle, it’s 37 stories (or something) on about a third of the space those guys developed, but they got height transfers for like 12 stories one of the other thirds, and three or four from another.

      So one of the buildings is 37 (or something), the other is 3 (or something) and the last is 12 (or something) so in average it’s a 16 story building. Super weird stuff.

  5. Just an observation: There is a problem in providing transit infrastructure for districts with too many tall buildings. If you have ever ridden BART or Muni in San Francisco during the peak, or ridden the subways of New York City of the L in Chicago, you know what I mean.

    Another land use question is how much does any city want to set up a zoning height limit that results in all of the “office building profit” going to one location and one business owner. (I have heard that the Empire State Building glutted the New York City office market for around 15 years, and that the World Trade Center had a similar effect.) Consider too that building public transit facilities to serve mega-dense districts can easily be exponentially more expensive because it usually requires massive subway construction.

    Although the height limits are imposed for other reasons, the height limits in Philadelphia or inside the District of Columbia seem to do a nice job of creating an interesting American city that also has a high transit mode share. In the DC example, one result has been “new urban” office centers in places like Arlington, Virginia and Silver Spring, Maryland — places that are quite attractive and have a high transit mode share. Because these also get “reverse commuters”, WMATA benefits from two-way productivity at peak periods. From your FAR estimates and observing other cities, it appears that 10 to 15 FAR seems to be the sweet spot.

    1. Where you see problems, I see opportunities. Every ride on Bart represents one faster trip home, and minutes of productive life being resored to a commuter. Sure, subways seem expensive. But that’s a one-time cost, and not only does it yeild a constant return in saved time, it also saves massive road and parking infrastructure.

      I want to go further in this discussion, but I’ll stop. We’re close enough to the same side – I’ll save my breath for those arguing for a 5 FAR city.

      1. My point is simply that transit capacity ,ist be an issue in a high-rise density discussion. If there is not enough transit capacity, then transit riders suffer from not only over-crowded trains, but slower trains because the dwell times have to be much longer to empty the cars.

        I assure you from years of commuting on Muni Metro that it is like being inside a badly-ventilated sardine can that crawls for the last mile before the Financial District. I have even met people who have changed to driving and paying for parking in San Francisco simply because the transit rider experience is so unpleasant. There is a tipping point.

      2. I’ll buy that. The answer, of course, is to add more transit. (sounds easy, right?) I’d argue it will always be cheaper, overall, to have more density and provide an appropriate level of transit.

      3. A pretty strong argument can be made that San Francisco has under-invested in high capacity transit considering its density. The overloading of the Muni and BART Financial District stations and the crush-loads on buses in the Greary corridor certainly point to that.

        It could be argued Seattle is near maximum capacity for buses downtown during the AM and PM peak. However we’re in the process of building out Link which will ultimately result in a net gain in transit capacity (even with kicking buses out of the DSTT.

    2. Whyte also sites 15 FAR as the magic number for urban density. He was obviously a strong supporter of density, but felt that going above 15 FAR had too many negative impacts at the street level.

      He also felt that at street level, wider was better than taller, but that a few stories up it was better to narrow buildings into towers, to allow for more daylight at street level, and to avoid wind-tunnel effects. Old, traditional 3-tier highrises fit these rules very well.

    3. For what it’s worth, neither the Empire State Building nor the World Trade Center glutted the market for a decade per se. The Empire State Building was finished in 1931, if I recall correctly, and the WTC in 1974. Which happen to correspond with the beginning of the Great Depression and the beginning of the debacle that was NYC in the late 1970s. Had the Empire State Building been finished in 1925 or the WTC in 1988, either one would have been absorbed into the market quickly — neither one is all that massive.

    4. “There is a problem in providing transit infrastructure for districts with too many tall buildings.”

      But what are the alternatives? (A) Concentrate those buildings in another city, which just moves the problem, and probably means worse transit options for workers. (B) Dilute the jobs in exurban office parks, which would greatly increase sprawl and force workers to drive to work. (C) Let the jobs disappear entirely, decreasing employment and dampening commerce. Maybe tall buildings and an extensive but still crowded HCT are better than the alternatives.

      Do SF and NYC really have too many tall buildings? Overall the residents like it that way, and they have built enough HCT to make it work. I said above that super-tall buildings (above 40 stories) were unnecessary, but it would still be an “urban downtown” even at 40 stories. The problem in America is that the “downtown zones” were not expanded and replicated to absorb most of the employment demand: instead jobs went to low-density office parks and strip malls. In most cases the downtowns were forcibly prohibited from expanding; i.e., SLU and Belltown could have been built up in the 1950s and 60s, and Capitol Hill and Beacon Hill could have been built like First Hill at the same time.

  6. Maybe zoning needs to change so the even when building skyscrapers the base of the building occupies its lot with street level retail and at least 5-6 floors covering the whole lot, and it only gets narrower after that. That’s more like the empire state building and older skyscrapers everywhere. Right now because of FAR limits if you want to go tall, you have to be skinny at the bottom, which often leads to useless plazas and lack of retail.

  7. Why do skyscrapers get built if they’re significantly more expensive to build yet have less rentable space?

    1. They obviously don’t have less rentable space.

      Why skyscrapers get build is because land in a city is expensive. If a block of space costs you $10M, you can build a single story for very cheap, but good luck making enough rent to pay that back. If you build 20 stories on the lot, however, the land price goes down to $0.5M per floor.

      Here’s a bit of math I ran on this a while ago. You’ll see that the premium for building up isn’t actually that much. For apartments, it’s $116/sf for high-rise vs. $107/sf for mid-rise.

      1. If you managed to average 20 stories, you’d have the second densest building in Seattle after the Columbia center. You’re not allowed FARs that intense.

      2. The land price is the most fundamental reason, but it goes beyond that. Some developers are obsessed with building the tallest building possible, like Martin Selig and the Columbia Center. Some businesses place a high value on three-digit addresses and 50th-story views, partly because it impresses their clients. This leads to buildings being taller than necessary.

      3. [Andrew] If you’re arguing that we should reduce/remove our FAR restrictions, I agree. But the math doesn’t change much even with a restrictive FAR. You just have to build higher to build enough square footage to share the high land costs.

      4. No you don’t. That’s the whole thing. You can either build tall or build squat. If you are making a principled argument for density, then all you should care about is density. You shouldn’t care about tall buildings or 1 story buildings with 100 floors underground for the magic mole people.

        There’s this mad Ayn Rand fascination with skyscrapers. Yes they are cool, but no, they aren’t they definition of density.

      5. You’re right – I was thinking of FAR a bit wrong. Of course, a maximum FAR limits your building, no matter how you build it.

      6. I probably should have labeled the graph more clearly (floor, area, and ratio ). Sorry about that!

    2. Building the Dexter Horton building today would cost you much more than it would cost to build a comparable sized high-rise building today. Also, zoning rules wouldn’t allow that intensity on that sized lot.

      I imagine there are economies of scale in construction, but I know nothing about that stuff.

  8. Aesthetically, at the same FAR I would prefer a shorter building. It means more street frontage, assuming street-level retail. And in terms of good development, going with a building with more low-level retail is a better choice unless the neighborhood already has as much retail as it can support. Even in some cases with lower FAR, if it means sufficient density covers a larger area then it’s probably better than maximizing density on one block. I’d like to see minimum FAR requirements, FAR averages for neighborhoods, and a requirement that new developments have higher FAR than what they replace. But I suspect this would actually mean fewer towers.

  9. Pulling your table into Excel, there’s another number that’s quite instructive. FAR*10/height makes the old vs. new buildings really break into two groups. FAR*10/height is an approximate volume of the occupied space inside of the owned envelope (and coincidentally a measure I just made up). Basically, the percentage of space inside the bounding box owned by the building owner that is part of the occupied building. The 10 is a rough estimate of an average floor-to-floor height.

    The average of your old buildings is 58.4%, and the average of the new buildings (excluding the Space Needle, which was just 2%) is 24.5%.

    1. Could this be due in part to additional infrasturcture in the newer buildings? Bigger elevator shafts, more utility closets, more space for HVAC, etc.? Perhaps also higher ceilings.

      Also note that added height probably means less usable space per floor. More elevator shafts, fatter pipes for water and fire standpipes, bigger wire ducts. More structure to hold the building up. Consider also the impact of modern earthquake standards.

      1. I’m not sure about the floor-to-floor heights in older buildings. In modern buildings the largest impact in floor-to-floor heights would be ductwork. But I can’t imagine that taking up more than a foot or so. That explains 10%, not the entire 42% change.

        The more likely explaination is our land use codes. These days FAR restrictions make builders put in (boring, useless) planters and setbacks. Older buildings were built right to the sidewalks (the way it should be).

      2. I’ll bet it’s more than 10%. Every couple of floors there’s a mezzanine for running computer and phone lines that someone has to be able to actually move around in plus all the HVAC and plumbing for the building. I don’t remember at what height they have to start pumping water but in the skyscrapers several floors are nothing but equipment rooms. You can’t put all the A/C heat exchangers on the roof of a 100 story building.

      3. At one atmosphere pressure (15 psi) you should be able to support a 30 ft. column of water. So, if you’ve got a 100 psi supply line, you’ll need pumps above 200 feet. You better make sure there’s enough pressure to flush the executive toilet on the 45th floor. You’ll need a pump and maybe a water tank on the roof. Oh, and you’ll need some steel to support that water tank.

      4. We generally don’t do water tanks – it’s all done with pumps. We do put a mechanical space every 20 floors or so, to keep our ductwork chases from getting too big. And we leave space on the roof for cooling towers and sometimes chillers. But again, even losing a floor or two every 20 doesn’t get you close to 42% difference.

      5. I guess I’m thinking old school. I lived in an old building in Chicago on LSD. One night I started worrying about my neighbor when the shower was running for going on an hour. Walked out into the hall (next to last room leading to the stairs) and it sounded like a river. Evidently the tank on the roof burst and it was 3″ deep cascading down the stairs. A building built to be residences probably has a much higher water demand than office space. You have to count on everyone taking showers all at the same time in the morning. In the HP Lake Stevens building they had a mezzanine between floors 2 and 3. But HP is probably more concerned with connectivity than standard Class A office space would be.

      6. I’ve seen that a lot in NYC as well. It turns out they were added back before booster pumps were common, which makes sense. Apparently they’re still required in Tribeca whether or not they’re being used, for aesthetics.

  10. A note of clarification. Zoning in Pioneer Square allows unlimited residential density. Commercial density is limited by FAR. This is consistent with neighborhood goals for new residential development.

    As well, the City’s Land Use Code provides protections for historic structures in both Pioneer Square and the International District along with Landmark structures throughout downtown.

    In Pioneer Square, development is complicated by a high water table, which makes parking underground very pricey. This means that higher value space–luxury units in very tall structures that come with at least one parking space–would feature parking above ground, reducing overall density in the structure.

    Thanks for the observations about density in new development.

  11. I thought I’d run some numbers on our maximum allowed FAR by building type. Specifically, I used the Height*10/FAR that I’m calling BVR (Building Volume Ratio) – it’s roughly the percentage of the envelope a building is allowed to take up.

    The highest BVR buildings allowed by code are:
    3 and 4 story commercial near stations (100%)
    14 story highrise residential (100%)

    Nothing else came close. As you get over 14 stories for residential your BVR quickly drops, since you’re limited to a max FAR of 14. For example, a 100 story residential building has a BVR of 14%. The same happens for commercial – at 12 stories you’re already down to 50%, and at 16 stories you’re at 44%.

  12. IMO we need more tall bulky buildings.

    I love the Monolith, for example (1001 Fourth Ave). Could such a building even be constructed today, code-wise?

    1. I don’t see why not. I believe it has a podium, meaning it doesn’t fill its entire property beyond a few floors.

  13. Your average MS building in Redmond also has hundreds of vendors working “some” of the time in them. There simply isn’t dedicated space set aside for them, so they’re not counted in the numbers.

Comments are closed.