Building affordable housing across DC: it's more than a question of equity, it's about how

The Post has an article ("Bowser continues push to get D.C. residents on board with plan to build low-cost housing in all neighborhoods") about a recent public meeting on the city's proposal to build more affordable housing across DC, more importantly--in every ward, including higher income areas of the city where such housing hasn't traditionally been constructed, in particular Ward 3.

They're positioning the initiative in terms of housing equity, by referencing past practices of mortgage redlining, which made underwriting of mortgages in mixed race or African-American neighborhoods virtually impossible.

D.C. residents arrive for a discussion on affordable housing at the Howard Theatre in Washington. (Astrid Riecken for The Washington Post)

It's a complicated issue. I regret I wasn't able to attend the meeting, in order to look at the various materials they presented at the meeting. Fortunately they are also online.

-- Undesign the Redline
-- city press release

But while it's great to position this in terms of equity, I am much more interested in whether or not they can build anything.

It's not that it's impossible to right the wrongs of segregation and structurally racist policies "from the past."  But to build housing at scale in a community where land is mostly "used up" albeit less intensively than today's development practices warrant is almost impossible because the process of adding significant amounts of housing to places that are already developed is practically and politically difficult.

It requires a sea change in thinking about what kind of housing can be realized, and in a city where most of the land is already developed, wrenching decisions must be made.

They aren't making more land: multiunit versus single family.  Basically for cost reasons it has to be: multi-unit ("apartment buildings" either as rental or owner-occupied).  But most of the residential districts in the city are zoned for single family housing.  Most of the housing in these areas is single family, mostly detached but sometimes in various runs of rowhouses, but as few as two attached units.

In a situation of limited land resources, traditional single family houses, even in the urban context where lots tend to be small, and lots/houses are squeezed together, and/or attached, a traditional single family house is no longer an option if you want to add units at scale.

-- "Let's Quit Fetishizing the Single-Family Home," New York Times
-- "Millennials want a single-family house, even if it means a long commute," Housing Wire

Duplex/triplex legalization elsewhere.  There has been a lot of press on how Minneapolis and other communities are legalizing the conversion of single family housing districts to districts where lots can be converted into duplexes and triplexes.

This is great for allowing very large rowhouses to be broken up into more units than would typically be allowed by zoning.

But it won't do much for traditional single family detached houses, which are not easily subdivided the way that rowhouses can be.

As a result, this won't result in the creation of a lot of new housing units, unless neighborhood are in super high demand.  And even with conversions, the cost of each unit won't be particularly "affordable" because relatively speaking, the cost of land is pretty minimal as a total cost of unit in terms of labor, materials, assemblage and holding, etc.

There isn't vacant land except in rare situations.  Most of the city's property for housing has already been developed, even if in some areas the lots are relatively large comparatively speaking, and could accommodate greater density.

Most of the new multifamily unit housing has been built in commercial districts, by reusing land more intensively.

Where new single family housing has been constructed is on interstitial institutional land that has been deaccessioned.  Mostly it has been in the outer city, for example on the site of the old Methodist Home on New Hampshire Avenue almost in Montgomery County, or on the sites of closed Catholic institutions in Greater Brookland.  Mostly the housing developed is zero lot rowhouses, but sometimes it may include detached houses as well.

Plus it's market rate housing so by definition it isn't "affordable" and usually in response to residential opposition it ends up being less dense than it could be ("More Than 80 Townhouses Planned Near St. Joseph's Seminary," Washington City Paper).

Single family houses can be a underutilization of land.  Ironically, earlier in the week I had cause to review testimony I submitted in 2006 on an earlier proposal to develop rowhouses on the Takoma Metrorail site.  I said that it should be multiunit housing and about 8 years later, they agreed (although nothing's been built).

As it is, the Fort Totten Station building with a Walmart on the ground floor, a few blocks from the Metrorail station, should be a couple floors taller.

This is the case still today.  For example within a few blocks of the Fort Totten Metrorail Station, across the street from the multiunit mixed use building on Riggs Road that includes a Walmart on the ground floor, they'll be constructing mostly rowhouses ("Restaurants, Retail and 185 Townhouses Proposed for Fort Totten," Urban Turf).

Larger lots capable of redevelopment aren't usually easily transit accessible.  Besides the fact that most potentially available land is already developed, a lot of the land that might be available for more intensification is not well situated in terms of transit and community amenities.  Most often it's on the outskirts of the city.

That means that most buyers will also choose to own one or more cars, thereby adding to area traffic congestion and increasing their cost of living.

Transit efficient locations mean that a household can get around without owning the typical two cars that most households (in the US, not in the DC) own.  Using the rough estimate that maintaining a car costs about $8,000/year, that supports almost $200,000 of monthly payments on a mortgage.  Two cars equal to about $400,000.

Ward 3...  The article states that the biggest target for new housing is Ward 3, the city's most exclusive ward with neighborhoods like Georgetown, AU Park, and Hillandale, and other spread out areas, where car use is particularly high relative to the rest of the city.

Besides redeveloping commercial land along Wisconsin Avenue it means redeveloping existing lots as apartment buildings (or courtyard housing, see Ross Chapin's Pocket Neighborhoods).

Not to mention that acquiring property capable of such redevelopment isn't easy, especially if you want to do this in a way that doesn't encourage car dependence--so you want to focus on locations within one half mile of transit.

Moving from homogeneous housing districts to mixed forms.  It can be done, but it wouldn't be easy politically because it requires people to accept a change from single family zoning to mixed residential zoning. That's how housing was developed in the core of the city in the late 1800s, but people seem to have forgotten this.

Crap design.  Plus there is the issue of new construction design being compatible with existing housing.  The city doesn't have mandatory design review in areas that haven't been designated historic, and most of the infill and replacement housing that gets constructed is execrable.

DENVER, CO - MARCH 18: A single family home, on Lawrence St. between 25th and 26th Street, is sandwiched between two newly built multi family developments, March 18, 2015. Issues surrounding dense housing development are being fought in many neighborhoods throughout the Denver area. (Photo by RJ Sangosti/The Denver Post)

Granted it's SFH replacements that aren't quite McMansionized (see the McMansion Hell
blog) but built bigger in place of smaller houses, but each of the three instances of this on my block are pretty bad.

It's possible to do it right, but without any design requirements or consultation required whatsoever, the likelihood of such a process producing quality design in new construction is remote.

Courtyard housing image from Fine Homebuilding, "Designed for Community."

Courtyard housing as an intensification device.  As an aside, on my block and similar blocks in Greater Takoma/Manor Park, the north and south sides of the  block have long lots--our house is 1/6 of an acre, and you could easily take two lots and construct courtyard housing or apartments.  (Salt Lake has a bunch of exanples.  But think the apartments for the Melrose Place tv show too.)

You could definitely fit 4-5 units on each side.  But we are 0.8 miles from the Metrorail station (but three blocks from a bus line) and likely most of the households would want a car.  There's no way to build 8-10 units in this space if each is required to provide parking.  Adding garage or underground parking would boost the costs astronomically.

It wouldn't be too expensive to do this relatively speaking.  Buying the property would be maybe $1.2 million per lot.  Then there would be costs of demolishing the current houses (or moving them) + other infrastructure requirements including boosting the capacity of area utilities.  Constructing each "carriage house" without providing parking would cost no more than $300,000/unit.

That's at two floors of 400 s.f. each, which is the limit for ADUs.  Maybe they'd change the rules and allow three stories, which would add another $150,000 but increase the appeal.  You could have a ground floor apartment for each, plus the upstairs unit, providing 20 units total on 1/3 acre.

Writings on ADUs, but extending the concept from one unit to infill construction, should be referenced for ideas, along with the aforementioned Pocket Neighborhoods book, which is beautifully written and illustrated.

Auburn House Courtyard Apartments, Brookline, Massachusetts.

Courtyard apartments.  Then there is the option of adding small apartment buildings, either as courtyards, or not. 

This would be a type focused on infill redevelopment in traditional single family neighborhoods. 

Think the building type typified by the apartment complex in the "Melrose Place" tv show, which was set in Los Angeles, or the movie "Singles," set in Seattle.

Coryell Apartments, Seattle. 

That'd cost more but would add more units compared to courtyard single family housing.  Not necessarily for ownership. 

Smaller units would yield even more units.  But then you have the car problem too.

And you'd want to require a mix of larger units, to accommodate families.

Regular apartment buildings.  I see courtyard apartments as a type more for infill locations within existing R1 and R2 single family detached neighborhoods. 

There's also the regular apartment building, ideally aimed strategically, within 1/2 mile of Metrorail stations, where "forbidding" car ownership is more practical.


These smaller apartment buildings immediately abut the Columbia Heights commercial district on 14th Street NW, which when they were built, was much less intensively developed, where two story buildings were the norm.

The Gables apartment building on Blair Road "in front of" the Takoma Metrorail station.  The reason this building is "prettier" than a typical newly constructed apartment building is that it was originally intended to be condominiums.

The opportunity costs of not building taller.  A big issue though is density.  Today the nation's population is more than twice what it was in the 1930s, by which time most of the city had been built out.

Most of the new buildings constructed today, such as at Fort Totten Metro (four stories right on the station grounds) or near to the Takoma Metrorail Station (a three story building constructed behind the CVS, a couple blocks from the station) are in my opinion not big enough.

Adding a couple more floors would add units, and over time and with volume would create a brake on price appreciation.

While existing residents typically agitate for such buildings to come with lots of parking, studies of actual behavior find that such buildings generate 50-75% fewer car trips compared to traditional single family housing in the same neighborhood.

Identifying intensification opportunities.  In addition, there are many apartment buildings that were developed in the past, when the city's population was much smaller, that have the opportunity to be rebuilt or otherwise made bigger.  This could add a lot of units.

3 story apartments on Randolph Street NW in Petworth, about two blocks from the Georgia Avenue Metrorail station. 

Conclusion.  While I think the initiative is admirable from a moral standpoint, by not focusing on the most practical and basic questions of how and what to build, I can't see the initiative accomplishing very much, especially as it is being launched on the tail end of the likely last term for Mayor Bowser--three years could be enough time, but the practical issues:

• buying land for redevelopment at market prices
• getting residential buy in on new infill development of multiunit buildings
• doing decent design (which frankly won't be a priority of this administration anyway)
• huge huge huge subsidies required to make the housing "affordable"

are daunting even in the best of circumstances.

I'd say too the issue isn't getting people to feel like it's their business to right the wrongs of inequality, it's equally if not more important to focus on showing examples of already existing denser housing and how it makes neighborhoods better, by extending the range of available housing, and doesn't overtax the road infrastructure when it is proximate to transit stations and lines.

DC has plenty of examples, but they need to be codified, explained, and promoted.

Still, getting property owners to agree to selling their house/lot so it can be redeveloped will be very difficult, as will getting area residents immediately impacted to be agreeable to the changes.