|
|
A printer-friendly PDF version of this report is also available.
AFFORDABLE BY DESIGN
A proposal to create middle income housing in San FranciscO
A SPUR REPORT
Adopted by the SPUR Board
November 21, 2007
Alongside all of the formal programs to create affordable housing in San Francisco exists an under appreciated, but potentially important,component of the housing market: units that cost less because they are small and efficiently designed, and in many cases do not come with a parking space.
Housing that is affordable “by design” could become a more important part of San Francisco’s middle-income housing strategy, something to been courgette to supplement nonprofit, 100 percent affordable projects subsidized with public funds and inflationary housing units subsidized by private developers. This is a housing type that benefits the lower middle class rather than the truly poor, but it is precisely in this middle income stratum of the market that San Francisco has been least successful in serving. Given that San Francisco’s median home prices have hovered between $750,000 and $800,000 over the past year, and that the City’s inflationary housing program creates units priced between $200,000 and $250,000 , there is a need to create units priced somewhere in the middle that don’t require public subsidy.
Currently, the Planning and Building Codes make it extremely difficult to build housing that is affordable by design. This is primarily due to density controls, overly prescriptive courtyard and unit exposure rules, and minimum parking requirements in the Planning Code, as well as Building Code limits on the number of stories allowed for wood-frame construction. The result is that most of the city’s moderate income housing stock is in older buildings, constructed
before current code provisions went into
effect. We believe that some private developers
would build affordable “by design” units in
some locations in the city if doing so were not
essentially prohibited, as it is now.
In 2007, SPUR convened a task force
comprising architects, developers and policymakers
to explore strategies to reduce the hard
“bricks and mortar” construction cost of new
housing. In general, this means designing units
that are smaller, more efficient, or have fewer
amenities. The goal is to enable housing that
can be brought to market at prices affordable to
households earning between 120 percent and
150 percent of San Francisco median income.
That translates to between $77,000 and $96,000
for a family of two and between $96,000 and
$120,000 a year for a family of four.
This article, resulting from the work of the
task force, contains SPUR’s recommendations
to enable the construction of housing that
is affordable by design. In particular, SPUR
recommends that the City should:
> Regulate building density by height, bulk
and setback requirements, not by limits on the
number of units allowed.
> Stop requiring parking in new buildings.
> Stop regulating bedroom counts.
The Landes is an example of a
five-story woodframe building over
two stories of concrete, a common
construction type for Seattle, but
not San Francisco.
> Enable a greater range of wood-frame
buildings to be constructed by allowing housing
at the ground floor of podium buildings and
greater flexibility in the code to facilitate a fifth
story of wood-frame construction.
> Allow developers to fulfill their inclusionary
housing requirement by providing a greater
percentage of their units at middle-income price
points.
> Modify requirements for courtyard widths
and rear-yard setbacks to allow for greater
design flexibility in locating common open space.
Recommendation #1: Regulate
building density by height,
bulk and setback requirements,
not by limits on the number of
units allowed
The right way to control the size of buildings is
to rely on height, bulk and setback requirements,
not by limits on the number of units allowed.
This manages the impacts of buildings on the
streetscape and the skyline. However, in much
of San Francisco, we also regulate building size
through limiting the number of units that can be
built within that building envelope. Very often,
it would be possible to fit more units within the
allowable height and bulk, but because the total
unit count is restricted, there are strong financial
incentives to build larger units in order to fill
the allowable zoning envelope, which results in
larger, more expensive units.
Again, the main way to make a unit cost less
is to make it smaller and more efficient. The
simple change of regulating building size by
height and bulk instead of limiting the unit
count would encourage the development of
more small units instead of fewer, much larger
units. This has the added benefit of dividing (or “amortizing”) certain fixed costs, such as the cost
of land, by more units, further driving down the
cost of each unit in the project. That being said,
we understand that smaller spaces need to be
designed to be as livable as possible. In general,
higher ceiling heights and good exposure help to
improve the livability of smaller spaces.
Recommendation #2: Stop
requiring parking in new buildings
SPUR already has written extensively on the
relationship between parking requirements and
housing cost. Because parking costs so much
to construct (between $40,000 and $75,000
per unit in San Francisco), it adds to the cost of
the housing unit. Requiring units to include a
parking space increases the cost of construction
by that amount. By eliminating the requirement
to construct parking (and by selling or renting
parking spaces separately from housing units),
greater affordability can be achieved. Clearly, if
a developer were trying to create a housing type
with smaller units targeted to moderate-income
households who would otherwise not be able
to afford to stay in the city, many of those units
would not include parking.more bedrooms. While SPUR strongly supports
the policy goal of retaining and attracting
families to the city, we do not believe that
mandating the construction of multi-bedroom
units achieves that goal. The requirement for
multi-bedroom units adds to housing cost,
while not necessary resulting in the housing of
more people.
Recommendation #3: Stop regulating bedroom counts
Many people are rightly concerned about
the loss of families from San Francisco. Some
advocates believe that one way to keep families
in San Francisco is by requiring the construction
of “family sized units” — that is, units of two or more bedrooms. While SPUR strongly supports
the policy goal of retaining and attracting
families to the city, we do not believe that
mandating the construction of multi-bedroom
units achieves that goal. The requirement for
multi-bedroom units adds to housing cost,
while not necessary resulting in the housing of
more people.
The truth is that the city is filled with multibedroom
units — most of the traditional
Victorian building stock — but those units
often are occupied by unrelated adults. There
is no evidence that multi-bedroom units in new
developments are being occupied by families, so
by requiring multiple bedrooms, the City simply
may be requiring singles and couples to have
offices and guest bedrooms. In other words,
from a policy perspective, requiring multibedroom
units fails the basic test of targeting:
The benefits do not accrue to the intended
beneficiaries. At the same time, this strategy
raises the cost of housing for everyone. The only
way to make the city more family friendly, from
a housing perspective, is to solve the aggregate
housing problem. The government should reverse
course on the current trend of forcing units to
have more rooms, and let buyers and renters
make their own trade-offs between location and
space consumption.
One option to increase the availability of
affordable family housing is to change the
inclusionary housing requirement to one that
would require 15 percent of the total square
footage (not of total units) of a project to be
priced at below-market-rate levels. The BMR
square footage could be concentrated in all
two- and three-bedroom units within a project,
instead of applied proportionately to the unit mix
within a development. In other words, fewer units
in a project, but more two- and three-bedroom
units, would be offered at below market rates.
One additional possibility for adding new
non-traditional, multi-bedroom units to San
Francisco’s housing stock is to look more closely
at “one plus” units, where the second bedroom
does not share all the characteristics of a full
bedroom but could be used for sleeping. These
unit types employ sleeping alcoves or offices that
function as a separate bedroom. These onepluses
come in all shapes and sizes and are being
developed throughout the Bay Area. They are
already in place in many existing buildings,often
described as “junior twos.” The City could
consider allowing one-plus units to count toward
some of its two-bedroom count requirements.
Finally, through many of the changes
recommended in this paper, more compact two bedroom
units could be created and brought to
market at more affordable price points.
It is possible to develop compact
units for couples and families in
under 800 square feet.
Method #1: Housing at the Ground Floor of Podium Buildings
One of the most common building types for
new projects in San Francisco is four stories
of wood frame housing over a concrete ground
floor with parking and retail (a “podium”). The
building code section that allows this type of
construction does not permit housing in the
ground-floor podium. Some jurisdictions,
Oakland included, have developed local
guidelines that allow and actually encourage
housing at the ground floor.
In addition to allowing more units on the same
area of land, there are other obvious benefits
to this strategy: In locations where mixed-use
development that includes ground-floor retail
may not be appropriate, it would activate the
street edge and bring “eyes on the street” to
increase neighborhood safety.
Method #2: Enable an Additional Story of Wood-Frame Construction
The Building Code requires relatively onerous
measures if a builder wants to add a residential
story to the typical four-over-one configuration
described above. Again, other jurisdictions,
namely San Diego and Seattle, have adopted
code standards to make an additional story
of wood-frame construction feasible while
maintaining the equivalent fire and life safety (see the modification to Type 3 construction
recommended below).
As of 2007, the new building code will allow
four-over-one buildings up to 60 tall. A five-over-one would fit within this height, but only by
having a 10-foot tall ground floor. Ten-foot-tall
ground floors are common in new buildings, but
we believe that taller, 15-foot ground floors make
new buildings more graceful — in fact, more
similar to well-loved older buildings.
In many parts of the city the Planning Code
allows buildings to be 65 feet tall. By amending
the building code to facilitate the economical
construction of five-over-one buildings in
neighborhoods already zoned for this height, the
City would encourage more affordable density.
Additionally, in order to facilitate the
construction of additional stories of housing
in a wood-frame building, SPUR recommends
modifying the Building Code to make Type 3
construction more flexible.
Type 3 construction allows additional building
height and an additional story of occupied
space compared to Type 5 construction, but
without the large construction cost premium
associated with Type 1 construction. The code
definition of Type 3 construction presumes a
masonry building exterior with a wood frame
interior structure, which is not a cost effective construction technique in a seismic zone.
Modification of the requirements of Type 3
construction to allow the use of wood framing
in exterior walls, as has been done in some
California jurisdictions, would allow builders
to cost effectively take advantage of the 65-foot
height limit proposed in many areas of the city
by providing an additional floor level of housing.
SPUR has analyzed the affordable “by design”
strategies discussed here in the context of the
historically industrial areas of San Francisco.
We found that in places where some additional
height was conferred, unit size was reduced and
density controls were eliminated, many more
inclusionary units than the standard 15 percent
could be brought to market at prices affordable
to households earning up to 150 percent of the
San Francisco median income. One of the central
findings of this analysis is that the pricing of
units is highly sensitive to unit size, meaning the
bigger the unit, the less affordable it is to middle-income
people. The analysis also is sensitive to
mortgage interest rates and the percentage of
household income used to cover housing costs.
Giving developers the option to make more of
their units available at middle-income levels
(as opposed to fewer units, but ones that are
affordable to people with even lower incomes, as
is necessary under the City’s current inclusionary
housing ordinance) is something that the City
should explore throughout San Francisco.
Recommendation #6: Modify
requirements for courtyard widths
and rear-yard setbacks to allow
for greater design flexibility in
locating common open space
The Planning Code specifies the amount
of open space that each dwelling unit should
have. That requirement can be met through
provision of common open space that is available
to other residents. Common open space can
be provided in an interior courtyard, provided
that the courtyard is a certain minimum area,
which must be increased as the height of the
building area facing the courtyard increases.
Starting with a four-story building, the size of
the courtyard is increased by some percentage
over the minimum requirement. One way of
providing more building area would be to keep
the minimum area requirement of the courtyard,
but reduce the amount by which the courtyard
must be enlarged as the building gets taller.
The City also could allow the requirements for
common open space to be met elsewhere, say on
the roof. Another alternative would be to move to
a performance standard that requires reasonable
privacy separation while still providing design
flexibility.
Rear yards generally are required to represent
25 percent of the total lot area at the rear of the
lot. This sometimes serves to support existing
mid-block open space, but in other instances it
does not. In some locations it may be desirable
to replace rear-yard requirements with a lot coverage
maximum, allowing project sponsors
the ability to do courtyard buildings or other
designs that are appropriate for the building’s
location. This is commonly found in planned unit
developments and could be used successfully
on smaller parcels as well.
Conclusion
Affordable “by design” housing has been an
important part of San Francisco’s housing stock
for decades. By making a few key changes to San
Francisco’s planning and building codes, we have
the opportunity to encourage the creation of
new middle-income units — without additional
public subsidy. 
|
|
