Townhouses – Part 2 (The Problem)

Here’s the juicy part.

Everyone who lives in Seattle has some kind of opinion about the townhouses that have sprung up like weeds during the last ten or so years. Most of those opinions fall squarely in the negative on some level, and a significant minority of them can’t be expressed in polite company.

And yet, they sold in large numbers at increasingly eye-popping prices: enough that they kept being built even as everyone complained about them and wondered what the buyers and builders were thinking.

With some hindsight, we know that a non-trivial amount of this building and buying activity was driven by the gigantic credit bubble inflated by the Greenspan Put. But that only explains the magnitude of the problem, not the form. After all, most cities in the USA were affected by runaway asset inflation driven by loose credit. Very few of them wound up with anything that looks like this:

parking-diagramParking Court for a Typical Seattle “Four-Pack”
notice the unit “front doors” (if you can find them)
(from the
West Seattle Blog)

So, while the credit bubble may explain runaway land speculation and home prices justifiable only with the aid of recreational pharmaceuticals, it certainly does not explain why Seattle got so many atrocious townhouses. The excesses of financiers and their unindicted political co-conspirators may be responsible for many evils in the world, but it’s a stretch to heap responsibility for crappy townhouses in Ballard on their narrow shoulders. Other cities don’t have townhouses that look like this. They’re unique to Seattle.

Obviously, there was some other set of factors at work here. What were they?

Having watched it happen first-hand, I’ll argue that the Seattle townhouse boom was driven by a rogue wave of several negative undercurrents rising together in an immense and singular excrescence of bad taste:

Uncontrolled Condominium Litigation
Builder-driven Development
Poorly-conceived Land-Use Regulation

So, while loose credit certainly opened the door to speculative land development, the development itself was shaped primarily by the factors listed above.

Did you know that there are 45% more dogs living in the city limits of Seattle than there are children? If you’ve ever experienced how truly weird Seattleites are about their canine pals, this factoid probably doesn’t surprise you.

That statistic, though is indicative of much deeper demographic issues in Seattle:

Only 40% of Seattle households qualify as “families”
Fewer than 20% of Seattle households have children
Single-person households outnumber “families”
Average Seattle household size is unusually small

hh2000_LatestReleased_DPDS_006610Seattle Household Composition
(US Census Data)

Seattle households overwhelmingly tend to be small, white, affluent, and childless. In fact, there is only one city in the United States where this demographic trend is more pronounced: San Francisco. Yet, a comparison of San Francisco’s urban housing vs. Seattle’s shows a stark contrast. Seattle resembles nothing so much as a big suburb, while San Francisco is clearly a city.

Demographics define the market for housing, and Seattle’s demographics demand “Smaller Dwelling Units” and “higher density.” We have too many big, unaffordable houses, and not enough smaller, more affordable options available for ownership (rental is another story):

roomsMetro King County Housing Stock by Size and Type

So, demographic pressure toward the creation of smaller dwelling units for individual ownership was strong and increasing as the 1990s tech boom got into full swing. A lot of condominiums were built to satisfy some of this demand, but not nearly enough (besides, condos rarely have things like yards). Despite an intense cultural bias against them, townhouses started to get built by the hundreds to fill the gap.

Why didn’t we get more condominiums instead of townhouses? You can build many more condominium units on a site than townhomes, after all, so isn’t there an incentive to go condo instead? The simple answer to this question is: lawsuits.

In 1989, the Washington State legislature re-wrote its Condominium Act, introducing a whole host of new provisions intended to regularize the legal creation of condominium residences. Among the provisions of that Act was the seemingly innocent RCW 64.32.445, which stated in part that any constructor of condominiums implicitly warrants to buyers that the building and units are free from defective workmanship and materials as well as being designed to industry standard quality.

Sounds reasonable, right? A lot of poorly-constructed condos had been built during the 80s, and the developers pocketed their profits and walked off scot free…leaving condo owners holding the bag on expensive repairs.

In practice, though, this obscure but well-meaning legal provision was a slowly-building disaster.

By 2003, condominium construction defect litigation in Washington had become Big Business. Law firms began to specialize in taking down condo developers and arm-twisting condo associations. Awards to litigants were escalating. Condo Homeowner Associations (HOAs) were ripe targets, since the HOA itself would be the plaintiff in a suit and become a sort of mini-class-action against a developer with presumably deep pockets.

One of my clients, on completing a condominium project in 2004, moved into one of his own units and began serving an initial term as HOA president (partly as a self-protective measure). Within months of completion, he had received a stack of alarmingly-worded solicitations from attorneys (as far afield as Texas) that measured more than two feet thick in a pile. Some attorneys even began to threaten to sue condo HOAs themselves for negligence if they in turn did not act to sue developers (with the attorney’s help, of course) before the statute of limitations ran out on potential claims, no matter how vaguely defined.

Needless to say, condominium lawsuits proliferated and the lawyers got rich. By 2004, condominium developer liability had become truly gigantic in response to litigation risk. The cost of insurance alone (calculated into the project “wrap policy”) was adding as much as $50,000 to the cost of building each residential unit (“per door”, as the developers say). Developers tried to pass these costs through to condo buyers, but there were limits on how far that could go.

Would you pay an extra $50,000 for a condo just so the developer could be insured?

In the end, the liability risk of building condominium projects became uneconomic in the extreme. By 2003-2004 developers had quit building small condo projects entirely.

In place of condos, townhouses became the smaller-lot multi-family residential projects of choice for the development community. Because each townhouse sits on its own piece of land, its ownership is held “fee simple” rather than in condominium as air rights. Rather than having a collective entity representing common property interests, which could then easily serve as common plaintiff or defendant in a lawsuit, townhouse owners would each have to sue a developer individually in any defect claim. While that doesn’t do anything to prevent the lawsuit liability, it does remove all the incentives the lawyers had to drum up business by scaring HOAs into suing.

Townhouse developers tended not to get sued, while condo developers did. Case closed.

The situation had got so ridiculous that the state legislature finally revised the Condo Act in the mid-Aughts. By then, it was too late. The credit bubble was well underway, money was flowing into real estate, and risk-averse developers had attained a preference for building townhouses by the dozen.

There is a negative stereotype of Americans that goes something like this (paraphrased from Oscar Wilde): We know the cost of everything, and the value of nothing.

Nowhere is the reality of this stereotype more evident than in the world of speculative housing construction. Spec builders are focused on one thing: putting up housing as fast as they can for as little cost as possible, then selling it as fast as they can to their target buyers for as much as they can get. Churn ‘em and burn ‘em, baby.

In practice, there are many market and legal forces which prevent spec builders from building firetraps which self-destruct as soon as they’re occupied. Aside from the obvious regulatory standards embodied in building codes and tort liability, buyers have a strong incentive not to buy poorly built houses (would you?), and mortgage lenders have a strong incentive not to loan money on houses that are going to fall apart quickly and lose their value. Insurance companies also play a large part here (and are the original creators of building codes in the first place, a role later usurped by government). If an insurance company won’t cover a house, it’s unlikely to sell and thus unlikely to be built that poorly.

(Do keep in mind that, even with these incentives in place, building codes are minimum standards intended to prevent only the worst offenses, and mortgage lenders typically don’t care if your house spontaneously disintegrates 30 years and one day after it’s built. All of which has a lot to say about the quality of 20th century housing construction, none of it good.)

As cheap money flooded the economy and housing demand increased, buyers became significantly less concerned with building construction quality, focusing instead on surface shine and the need to own anything. Similarly, lenders become much more focused on the near term profits to be had from origination fees, figuring all those loans were going to get refinanced in five years or less anyway (LOL Subprime). Building departments were overwhelmed and couldn’t effectively police burgeoning volumes of construction. Worst of all, everybody who could scrape together the miniscule licensing fee got a contractor’s license and got into the construction business. Construction project oversight was often poor, and the majority of the workforce barely spoke English (if at all…a common hiring requirement for construction superintendants was fluency in Spanish). The quality of construction workmanship crashed even as demand for subcontrators drove labor rates sky-high.

With changed incentive structures come changed results.

For homebuilders looking to make a quick buck, townhouses were the perfect vehicle for their ambitions. Historically, townhouses have usually been very well built (as evidenced by the large number that have survived the centuries). Those historic examples, however, were typically built from long-lasting materials (like brick or stone) at a time when standards for craftsmanship were much higher. Not so many of the Seattle spec townhouses. Built hastily from wood framing at minimum quality standards, down-market fixtures, and a spartan approach to detailing, complaints about the quality of their construction and appearance became common even before they were finished.

But they were cheap and went up fast. The more people bought them, the more got built…by people who were only out to make a quick buck. Reputable builders were still out there (and I was fortunate to work with some of them), but the market became saturated with crap.

I’ve already briefly mentioned Seattle’s ongoing problem with suburbanism. More than any other major American city, Seattle is dominated by single-family homes and the parochial attitudes that go along with them. Visitors often remark on how low-density Seattle appears, and the appearance reflects the reality.

Single-family homeowners, and the mossbacks who admire their quaint little neighborhoods, are a powerful force in Seattle politics. Wealthy developers and mighty institutions cower in fear at the sound of their approbation.

But what do the single-family homeowners fear? This:

ballard-seattle-homeowner Worst Fears Realized in Ballard
(never mind that this is in a
Commercial Zone on a major street,
this is what keeps Seattle homeowners awake at night)

This fear should not be underestimated. Only 30 years ago, Seattle was still very much a small town with lots of sleepy suburban neighborhoods close to the city center. Many of Seattle’s long-time residents desperately want it to stay that way…and they vote. Some of them file lawsuits.

At the same time, Seattle has been growing by leaps and bounds. As new growth has generated new building activity, we’ve implemented a series of reactive ad hoc measures and regulations to prevent new development…and failing that, try and force it to mimic the appearance of single-family houses.

Which leads us to the biggest issue of all…

All of the forces discussed here already certainly conspired to decrease the quality of townhouses built in the Seattle, but none so dramatically as our land-use code.

As much as Seattleites have complained about the poor quality of townhouse developments and lamented the ugliness of the “four-pack” and “six-pack” townhouse projects, these projects are exactly what the Seattle zoning code tells developers to build.

In the industry, we call this DBZ: Design By Zoning. Whenever zoning restrictions become so detailed, numerous, and onerous that there is only one possible way to satisfy them, you will inevitably get a whole lot of whatever that one possible way dictates. In Seattle’s case, that means hundreds of “four-packs” and “six-packs”.

Let’s take a quick look at how that works with a typical Seattle building site:

0. Typical 50’ x 100’ Seattle Lot


Here we have your typical 50-foot by 100-foot urban Seattle lot. For the sake of clarity, we’ll assume that it’s flat and empty. Our goal is to build high-quality townhomes on this site, since the zoning code encourages this, but first we must satisfy numerous zoning restrictions on development. For now, we’ll just assume that this is an “L” zoned lot (for low-rise multi-family), but won’t specify what kind (LDT/L1/L2/L3/L4).

As the area of the lot on which we’re allowed to build shrinks, the medium-grey area in the diagrams will shrink with it.

1. Building Setbacks


First, we have setbacks. The medium-grey area is the buildable lot area left over after setbacks are applied. Setbacks are a form of development restriction used extensively in suburban development to enforce minimum space between buildings and ensure that they never touch. Setbacks are also typically used to establish front and rear yards.

In this case, the minimum setbacks require a 10-foot front yard, 20-foot rear yard, and 5- to 7-foot side yards on both sides. These setback requirements are nearly identical to those required for single-family detached houses in Seattle.

Note that the width of the area we can build on for a 50-foot site is actually only 36 to 40 feet after taking away the side yards. That becomes a critical dimension when trying to put townhouses on the property.

2. Maximum Lot Coverage
codes_03_lot coverage

Although we’ve figured out the required setbacks, there are further limitations we have to deal with. For townhouses, we’re limited to a maximum of 50% of the total lot area to be covered by buildings (less for apartments and less in the lower-density zones). So we lose a little more developable area to lot coverage limits.

In practice, though, lot coverage limits are not as important as parking and open space requirements…

3. Parking Access Limitations & Requirements

The zoning code requires that all development project provide parking on site sufficient to serve the needs of the land use proposed. In the case of townhouses, that’s defined as one car per townhouse. Because the buildable width of site was reduced by the side yards, we’re limited in how many cars we can park on the site and still have room left over for things like front entrances and stairways.

The zoning code also requires that the site only have one driveway access to the street (called a “curb cut” for the place where the driveway cuts across the sidewalk). All driveway access to parking spaces is required to occur on the site, and may not be located in a front yard setback. These requirements still apply where a site is located on an alley, but the driveway is required to connect to the alley instead of the street.

From the diagram, you can see the large impact this particular requirement has on the site (and why Seattle townhouse developments have been described as “parking diagrams with houses on top”). Nearly half of the buildable area is now taken up with driveway and parking spaces. In practice, this is why most Seattle townhouses have their main living areas on the second floor. They just don’t fit on the ground level with the garages.

4. Required Residential Open Space
codes_05_open space

The zoning code also requires that every ground-related dwelling unit (i.e. townhouse) have an exterior open space (yard) connected to it at ground level. In the medium- and low-density “L” zones, the open space for a unit may not be smaller than 200 square feet, and all the open spaces on the site must average to no less than 300 sq. ft. Individual open spaces must measure at least 10 feet in any dimension, and never be smaller than 120 sq. ft.

Required open space must be private: we can’t count entrance walkways, or any other place that isn’t essentially a completely fenced yard.

In the case of our hypothetical project, you can see that the big back yards for the rear units allow the front yards to be smaller due to averaging, but they still may be no smaller than 200 sq. ft. The yards then push the street-facing wall back further from the street and make the building footprint even smaller.

At this point, we’ve established the two-dimensional planning restrictions imposed by the code. Notice how little of the site is actually available for building after we’ve gone through this exercise. No wonder the results are strange-looking and inherently suburban in nature.

Now, let’s look at how the zoning restrictions affect the three-dimensional shape of the buildings:

5. Base Height Limits
codes_06_base height

Zoning codes typically set limits on building height. The height limit is specified as a maximum number of feet above the ground surface to which any building may rise. All Seattle zones have defined base height limits associated with them. By “base,” we mean that this number sets a baseline limitation and there may be additional height bonuses available beyond that (which we’ll get to in a minute).

For the medium- and low-density “L” zones, the base height limit is 25 feet. Every floor level we try to squeeze under that height limit is going to require at least 9 feet at a minimum, and probably more like 10 feet if we don’t want the ceilings to be too low. How many times does 9 go into 25? Less than three. Fractions don’t count, since a floor level with a 5-foot ceiling is useless, so that means we get at most a two-story building out of this.

It’s probably worth noting that the base height limit for single-family zoned lots is actually 5 feet higher than this: at 30 feet. So, we’re being forced to build at a lower height than the suburban-style houses that might be next door. L-3 zones (considered higher-density) are allowed to match single-family in height with a 30-foot base height limit. L-4 zones (the highest-density low-rise zone) has a base height limit of 37 feet, but before we get too excited about that we should keep in mind that there is very little L-4 zoned land in Seattle.

Also remember that most of our ground level is already taken up with parking, and that another large part of it will have to be devoted to a stairway in order to get to the upper levels. That means most of the actual living area has to be above the first floor. If we’re in a higher-density “L” zone, we’ll get two floors to work with. It’ll be tight, but it can work out. In the medium- and low-density “L” zones, we get one.

In the diagram, I’ve simply taken the ground-level footprint left over after setbacks, parking, and open space, then extruded it up to the height limit. That makes for unusably small floor plans. Something’s got to give, which takes us to…

6. Building Width & Depth Limits
codes_07_width depth

There is another restriction in the code which is intended to prevent multi-family lot developers from building large, monolithic “bread loaf” style buildings: width and depth limits. The combination of all structures on the lot may not exceed 40 feet in width without “modulation” (essentially, dents and protrusions), and may not be any deeper than 60 to 65 percent of the depth of the lot. Townhouses are allowed an exception here: given a free pass to go to 90 feet in total depth.

So, using some creative structural engineering, we’re going to take the upper levels of our proposed design and push them all the way out to the setback lines. They can’t actually touch in the middle, since we’re required to maintain a 10-foot separation between them, but no matter. Now we have a little more space to work with when we do our floor planning. The project might work out after all.

7a. L-1/L-2 Building Height Bonusescodes_08_slope bonus

7b. L-3/L-4 Building Height Bonuses
codes_09_L3 height

I said we come back to height bonuses, and here we are. The code says that we can get some extra height on our buildings if we do something the writers of the code think will add to the lovely appearance of the structures. In this case, if we put sloping roofs on the top, we can make the buildings taller.

For the medium- and low-density zones, we get 10 additional feet if we make the roofs slope at a pitch of 6” rise in 12” run or steeper! All is not lost! We might get a third story after all.

For the higher-density zones, we only get a 5-foot bonus, but we started from a higher base height limit, so we wind up in the same place (35 feet total). (L-4 goes higher, but again, there are so few L-4 zoned areas in the city that could could count them on both hands and have plenty of fingers left over.)

In fact, single-family zones get this roof height bonus too (an additional 5 feet on top of the 30 foot limit for 35 feet total). So, if we compare the way our zoning code treats single-family lots vs. low-rise multi-family lots we immediately see that except for two provisions (lot coverage limits and number of dwelling units allowed) the zoning restrictions result in exactly the same buildable envelope on the site.

This is not in any way accidental, and knowing it we should not be surprised that Seattle townhouses look like ridiculously overgrown tract housing.

8. Unit Density

In fact, by now this conceptual planning diagram should start looking eerily familiar. This is what “DBZ” means. When the zoning code restrictions force us into an awkward configuration with little or no flexibility, you get a lot of whatever the zoning code restrictions stipulate. Every project starts to look like a zoning diagram. Hopefully that’s a good thing, but in reality it almost never is. Zoning codes are far too blunt an instrument to enact good design by legal fiat.

Though they happen to be very good at generating bad design in bulk.

The final piece of the puzzle is the unit density limit. After all, these are multi-family zones, not single-family (which have an inherent per-site unit density maximum of one). Here is where the distinctions between the various low-rise zones become more pronounced:

LDT 1 Unit per 2,000 sq. ft. lot area
L-1 1 Unit per 1,600 sq. ft. lot area
L-2 1 Unit per 1,200 sq. ft. lot area
L-3 1 Unit per 800 sq. ft. lot area
L-4 1 Unit per 600 sq. ft. lot area

(These density limits, by the way, are imposed by the Washington State Growth Management Act via the Seattle Comprehensive Plan).

So, if our 50’ x 100’ lot (5,000 sq. ft. lot area) is zoned LDT, we can only build two (2) dwelling units, whereas if it’s zoned L-3, we theoretically are allowed to build six (6). That’s a big difference, and the land value of LDT-zoned property is lower than L-3 zoned property as a result.

But hold on a minute…

After we’ve finished stuffing all that parking and open space onto the site, there isn’t really enough room left over for six individual townhouses on this site, let alone the hypothetical limit of eight should the site be zoned L-4. In fact, it’s a mathematical impossibility (add up the areas, assuming a very compact 350 sq. ft. per car, and you’ll see what I mean).

That’s why the graphic diagram for unit density above is only showing four (4) dwelling units (indicated by the red doors). That’s the practical maximum, given all those other limitations, and that’s why the “Seattle Four Pack” has become so ubiquitous (and run so afoul of the GMA).

There are lots of qualitative problems with the oh-so-uniquely-Seattle Four Pack, of course. How do we hate them? Let us count the ways:

courtThe rear units have no relationship to the street and the central parking courts are dark, dismal spaces.

(This is the view facing the street on a corner lot, by the way. Lovely, isn’t it?)

1_29101929_0The mushroom shape forced by the small ground level and cantilevered upper levels is visually unbalanced and appears ugly to most people.

side To qualify as “private” the postage-stamp yards are typically fenced in with the code-maximum six-foot-high solid fence…often right at the sidewalk.

front The shape and configuration is boring and suburban-looking, and the units are cramped.

df_says_ugliest For projects limited by code restrictions intended to make them more friendly to single-family neighborhoods, they seem to hulk over everything around them (the “ten pounds of shit in a five-pound bag” problem).

fence The “lowest common denominator” design (using the term extremely loosely) and lowest-cost construction methods, make them look cheap and crass (Google “ugly townhouse” and 5 of the 21 results on the first images page are located in Seattle).

Shall we compare them to a summer’s day? I think not.

However, in our next installment, we will compare Seattle’s townhouse development to some significantly less hideous examples from around the world, and then try to draw some conclusions about how to make Seattle townhouses better.

Stay tuned.