Affordable by Design

Meeting notes applied to the agenda

Feb 18, 2008

Key Conclusions from the Meeting

• Because of high costs (attributable in most part to regulatory and land use policies), building new market-rate rental housing in PA or SF is impossible. 
• "Costs are way up compared to the last RHNA round. All cities are going to do a much worse job of meeting ABAG affordability targets."
• There’s affordable housing and there’s market rate housing, and there’s a $500,000 PA price “gap” in between. The gap is at an all-time high. The housing market is broken.  In PA, there’s currently no way to create product in-between $350,000 (subsidized affordable) and $850,000 (Hyatt Rickey’s and East Meadow projects).  “Nothing is built in the middle.” 
  • Castle Group is a technology leader in factory-built housing, where sections of homes are built off-site (with cheap labor) and then trucked in. At 60 DU/acre, factory housing can save compared to on-site construction, but not enough to un-break the housing market.  Claims that pre-fab technology can make a significant dent in housing prices are untrue.  There is no silver bullet. 
• That being said, SPUR shows a way to unbreak the housing market. SPUR’s affordable by design vision of 5-story stick frame construction over podium parking really is the way for PA to go. For SF, the ambitious plan is to split the DUs as follows: 60% market rate, 40% moderate affordable at 133% or 150% of area median income (AMI).  This is a re-definition of “moderate” to a pricier level than the traditional 120% AMI moderate definition, but it is also more realistic. The vision is for 800 square foot (sf) two-bedroom for-sale condos as workforce housing, priced at $450,000.  Hard core residential TDM is a key part of this – less parking really matters.
  • Hence, we must remove PA’s 50’ height limit and follow all of SPURs recommendations. 
  • And we should build tall within walking distance of Caltrain. 
• PA has much higher land costs than SJ, so the product that may make the most sense? 8-20 story steel frame housing (try Avalon Towers in MV).  Not popular, but PA’s 40 years of “drawbridge mentality” have bid up land values to create a distorted housing market.  This is truly PA’s own doing. 

Who:

• Lydia Tan, BRIDGE Housing, huge affordable developer, wife of Councilmember Barton, http://www.bridgehousing.com/Default.aspx?DN=115,22,2,1,Documents
• Bruce Fairty, The Castle Group, 60DU/acre factory built homes in SJ

o         1800 DUs in the pipeline, all TOD within 2,500 feet of CalTrain/VTA LRT stations.  Mostly in SJ and Milpitas. Around 1998, built Whisman Park in MV at Whisman Station LRT. Proposal to  build next to San Mateo Caltrain station: http://www.sanmateodailynews.com/article/2007-1-5-sm-condos 

• PA resident
• Modular multistory chunks of housing are built off-site, trucked in, and placed with a crane.  The modules are 80% finished off-site: drywall, electric, plumbing, light fixtures, A/C, windows, exterior cladding. Above 60DU/acre, the cost advantage increases.  At 30DU/acre, there is little cost advantage.  Product is of uniformly high quality. Factory process increases quality. 
• For the state of CA, the state is responsible for modular housing building code and inspection. It would be exceedingly difficult to build standardized modular housing for multiple cities at an off-site location and having multiple local city inspectors be responsible for plan check and inspection. IE you can’t subject standardized product to randomized building codes and inspections. Building inspection has to occur at the factory site. Not all states are this advanced. 
• San Jose has been terrific to work with on this factory housing. I would expect affluent cities to slow down factory-built housing entitlements.   
• City Architecture Review Boards could present some challenges for factory-built housing
• Scott Ward, Mozart Development / Classic Communities.  20DU/acre homes in PA.
• Steve Raney

Annotated Agenda

0. Background

Palo Alto has been given a RHNA allocation of 2,860 housing units for 2007-2014 by ABAG.  About 50% of these units are moderate, low, or very low income.  Given 15% inclusionary zoning, how do we build these affordable units?  Palo Alto Housing Corp and other affordable builders struggle to build 100 subsidized DUs every 5 years in Palo Alto. Can we build a substantial portion of the affordable allocation using "affordable by design?" 

Other affluent Santa Clara County and San Mateo County cities have very similar challenges, so this challenge applies far beyond Palo Alto.  I met with ABAG and I was disappointed that they really didn't have any sort of a solution as to how affluent cities with high land costs could build large numbers of affordable DUs.  ABAG recommends looking at pre-fab housing, but does not share any insight about it. 

MEETING NOTE: I want to emphasize that ABAG and HCD are dropping the ball about figuring out how to build lots and lots of affordable housing. They truly are dumping an unfunded mandate on cities. In contrast, SPUR is making an important contribution towards solutions.   

1. The big question:

“What would it take for Santa Clara, Sunnyvale, Cupertino, Saratoga, Mountain View, Los Altos, Palo Alto, Menlo Park, San Mateo, Redwood City, San Carlos, Belmont, Foster City, East Palo Alto, Burlingame, San Bruno, Millbrae, and South San Francisco to build 7,000 standardized mid-rise affordable units in 7 years (at a substantial buyer cost savings compared to current costs)?"  Would sub-regional collaboration, standardization, and large economies of scale help? 

ANSWER: sub-regional collaboration would not be a silver bullet. 

2. Pre-fabrication / manufactured housing / panelization

• Says Scott Ward: There is no silver bullet on pre-fab, or else everyone would be doing it.  For 3 story stick, surface parking, a $700 per sf (buyer price) home only has $125-$150/sf in building construction cost.  HUD’s claim that you can cut that $150/sf by 50% is dead wrong!  Even if HUD was correct and construction cost could be cut by $75/sf, that’s only a 10% bottom line savings to the buyer. Land cost, impact fees, etc add too much to housing costs in PA.

MEETING NOTE: There are many costs that are apportioned to each DU: land costs, site improvements, podium parking, city fees.  Building a housing structure on the cheap is only a small part of the battle. 

LONG MEETING NOTE: Housing cost at different densities. 

PA builders typically obtain a 12% to 17% profit margin.  I’ll explain a table below that shows that the market rate units have a much higher profit margin, but 15% to 20% of DUs are BMR.  Builders lose money on BMR units.  So, really, builders don’t make exorbitant profits, despite what residents think.  

 

The builders’ overall South Bay cost is roughly $480 per sf.  This includes land acquisition, site improvements, construction, parking construction, fees, financing, carrying costs, etc. San Jose land costs less than PA land.  Intuitively, 60DU/acre should cost less per sf than 20DU/acre, but this is not the case.  Denser housing has more economies of scale, but this is offset by higher per unit parking costs and higher land costs.  Land enabled to be entitled for 60DU/acre in PA would sell for more than $6M/acre.  “High density multifamily (60DU/acre) is a relatively low-margin, high-volume business.”

 

Denser housing lowers the square footage, that’s why the buyer pays less per unit.  In PA, the condo buyer’s retail cost is $700+ per sf. 

 

Here is market price housing cost at 20 and 60 DU/acre. 

 

Density	sf per DU	Bldrs overall cost per sf	Bldrs overall cost per DU	Buyer's price
20 DU/acre	1,600	480	$768,000	$1,120,000
60 DU/acre	1,100	480	$528,000	$730,000

 

15% (20% if 5 acres or more) DUs are BMR.  BMR condos are sold for roughly $350,000, though price varies with sf and other considerations.   

 

More about builders’ margins.  Builders have to buy land on speculation, hold onto land with high carrying costs, go through bruising, lengthy public approval processes, and ride out industry downturns. Cost of sales (commissions, closing costs, fees, etc) take an additional 5% out of the buyer’s price, reducing builder margin further. Development is not a job for the faint-hearted.  Just look at the recently insolvent US Top 10 builder Lennar. 

 

PA has a “drawbridge mentality,” so housing is artificially scarce.  This increases builder profit margin compared to other cities.  Developers need a risk premium to justify doing business in as unpredictable and difficult a place as Palo Alto.  Comparatively, typical developer margin in more predictable cities like San Jose and Milpitas is 12% above cost.     

 

MEETING NOTE:

• ConXtech in Hayward produces snap together steel frame components for 6-10 story steel residential. Plan is also to go with a “podular” concept, IE euro pre-fab style kitchen/bath walls with appliances and shelves. CEO is Robert Simmons. 
• His steel system is used at 800 High Street. 
• http://www.bizjournals.com/sanfrancisco/stories/2006/02/27/focus4.html
• http://www.thefabricator.com/Robotics/Robotics_Article.cfm?ID=1052

3. What would it take to build non-subsidized, market-rate moderate-income apartments in Palo Alto?

MEETING NOTE: can’t do it.  Lydia: SF needs lots more rental housing, but building rental doesn’t work anymore. 

Except for "luxury apartments," apartments generally have monthly rent that qualifies for moderate income designation.  Would for-profit developers build apartments in PA?  Says Scott: "No way.  It won't pencil.  For-sale market rate residential will pencil, apartments won't."

A related point, even if market rate developers could build product for a very low cost per sf, they have no incentive to keep buyer prices low.  IE, they'll just sell at $700/sf, no matter how much it costs them. IE it would be silly for for-profit developers to sell a unit for $500K when they could get $750K for it.   They wouldn't pass on the savings.  Hence, "not for too much profit" developers would be required.  Lydia Tan runs Bridge Housing’s BUILD division, one such example.  This almost argues for a "cost plus" model. 

4. Five-story stick over podium (SPUR’s Affordable-by-design Report)

• SPUR has a report on affordable by design.  Sara Karlinsky was the lead SPUR staffer.  Lydia Tan was on the report’s panel.  See: http://www.spur.org/documents/2008.01Interiors_FINAL.pdf. (Pages 2-9.) SPUR advocates 65’ tall 5 story stick ontop of a 1 story podium as hitting a cost sweet spot.  San Diego and Seattle have modified fire codes to enable this.  SPUR also advocates for some commonsense modifications to zoning and planning codes, arguing to enable more DUs rather than larger DUs for bigger families. 

MEETING NOTE:

• SF is targeting the Eastern neighborhoods: Dog Patch, Bay View, parts of Portrero.
• Ownership housing is part of SF’s strategy to keep families in town by providing workforce housing. 
• Plan is for mostly 5 story stick above podium, not super high rises (like Rincon center). 

4C. What is the state of the art of squeezing down square footage per person.  Can we eliminate bedrooms, replacing them with futons or Murphy beds?  It works in Japan. 

Berkeley's Gaia Building has been held up an example of nice Dan Solomon small sf design. 

4D. John Barton: We have to reduce parking requirements

The Opportunity Center has two stories of underground parking, mostly empty.  Each parking space is very expensive, so we have to reduce the number of required spaces.  20% of the Opportunity Center cost was in parking construction, much of it wasted. 

5. Other meeting notes

• The state is providing BILLIONS to fund in-fill TOD.  Alas, PA has no projects in the pipeline that would qualify for this. Oakland and other cities are going after this funding aggressively. 
• “Arlington was a very comparable, snooty city like PA, but transformed itself with an extremely dense, vibrant TOD spine within 20 years. Arlington thrives.  Arlington did darn close to perfect in those 20 years.  They got everything right.”

• Here’s an important theoretical concept – politically, we can’t implement the proposed solution. The concept is the “zoning tax” from top Harvard housing economist Edward Glaeser.  Cost of housing is a function of land use and zoning regulations. The more restrictive, the higher the cost.  IE in an unregulated (truly free) housing market starting from 1960, PA would have triple the number of homes and homes would cost $350,000.  (No drawbridge equals many cheap homes.) Back when land was still cheap, SFH neighborhoods would have been transformed into dense, multifamily neighborhoods.
  • Hence, to lower the cost of housing, remove regulation and let the housing market be flooded with many, many new homes. (But, you have to do this before the cost of land is too high, or wait for a titanic housing market crash.) Once home values are high, voters will keep them that way. It’s impossible to deregulate PA’s housing market. 
  • Why should Orlando housing be 1/3 the cost of the Bay Area?  It’s because they don’t restrict residential development like the Bay Area does. 
  • Zoning’s Steep Price, Regulation, Fall 2002: http://www.oaklandnet.com/BlueRibbonCommission/PDFs/Zoning's%20Steep%20Price.pdf
  • Randal O-Toole provides a good summary: http://www.heartland.org/Article.cfm?artId=10635 .  Anaheim, LA, and NYC are really bad, and SF is twice as bad as them.  In ROT’s own work he villainously (and incorrectly) argues that smart growth is bad and sprawl is good.  Glaeser, our hero, is a huge advocate of urbanism.