A question I have heard a lot lately is “why can’t developers build housing for the people who need it most instead of for the rich.” Let’s look at what a typical multi-family development project in a reasonably central part of San Francisco would cost to build (in a very simplified way). I’m assuming an 800 square foot apartment in a five story 100 unit wood-framed building over a concrete first story (very common in San Francisco):
Land cost per unit of housing: $120,000 (current average)
Construction cost per unit (hard costs), which could be up to 30% higher in an all-concrete building: $300×800 s.f.: $240,000 *1
Permits, city fees and professional services fees at 20% of $240,000, but this could easily be doubled on some projects:  $48,000
Subsidy to build affordable BMR units (12% of total unit count) based on a $200,000 per unit subsidy x 12 divided by the remaining 88 units: $27,000 *2
Total cost so far: $435,000
Selling expenses (marketing, legal fees and real estate commissions) at 8%: $34,800
Total bare-bones cost per 800 square foot unit: $469,800
[EDITED ON 1/16/2014 TO CORRECT AN ERROR IN COMPUTING THE TOTAL]
This is VERY SIMPLIFIED and does not include construction financing expenses, contingencies, or developer’s profit, among other things, and on projects that are difficult to get permitted the cost could skyrocket. Google around for sample proformas if you want to get a better idea of project costs.
You’re looking at nearly half a million dollars for an 800 square foot apartment, and that is not even including all the costs that would actually go into a project like this. This is not affordable at all when the median household income is $73,000 in San Francisco, which qualifies for a $310,000 mortgage (play with the numbers here).  How could we bring this down?
-
A clearer entitlements process could bring down the $48,000 figure for permitting/fees/professional services significantly. Remember, this number could easily be $100,000 on some projects.
-
Finding another way to fund the BMR program would reduce the $27,000 subsidy, although arguably the BMR program would be unnecessary if market rate housing were cheaper.
- Lower real estate prices would mean construction costs would be lower. The cost of expensive San Francisco labor (construction workers have to deal with the high cost of living too) is a big part of that $300/SF figure and real estate cost also factors into the cost of storing materials and construction staging.
- An easier entitlements process would also mean that there would be more sites available for development, which would presumably drive the cost of land down.
1. This might even be a bit low now, costs are escalating quickly right now. Most of the construction cost goes into things like the structure, foundation and HVAC systems, elevators etc. not the finishes, so the construction cost per square foot is similar across different segments of the market.
2. The Mayor’s Office of Housing handles these sales, you can see what is for sale now here: https://sf-moh.org/index.aspx?page=299 Alternately, a developer can choose to pay a fee to the Mayor’s Office of Housing which will pay for affordable housing elsewhere in the city. These projects are typically built by a nonprofit like Bridge Housing or the Tenderloin Neighborhood Development Corporation.
One question I have is at what point (# of units) do you reach economies of scale, bringing down the cost per unit?
That is an excellent question and it’s very hard to answer. I do know that the hard costs per square foot of a concrete building at some size become cheaper than a wood building. Also, with extremely repetitive units costs become lower for things like cabinetry and fixtures. Many of the other costs like fees etc. don’t really come down with economies of scale, and often go up because larger projects are more difficult to get approved.
Hi Mark, thanks for writing something about this. I feel like it’s a hugely important question that gets at the *cause* of SF’s affordability crisis, and one that usually gets ignored in favor of heated debates about symptoms like “techiesâ€Â, Google buses, etc.
I think it’s also worth mentioning the cost of providing parking as a major block to the development of affordable housing. I came across a pretty solid writeup with good examples + images recently: https://daily.sightline.org/2013/08/22/apartment-blockers/
In addition to the things you mentioned above as possible ways to reduce the cost of these new units, what about increasing density? Both in terms of covering more ground or building higher and decreasing unit size. Though 800 square feet per unit is not huge, it is probably much larger than it needs to be in a city like San Francisco, where there is a vibrant collection of civic and entertainment venues and a culture and climate that facilitates and encourages their use. Particularly as family sizes decrease and the number of singles increases, making more compact units the norm would make a lot of sense, no? Just look at how much great functionality IKEA is able to fit into one of their mocked up sample “apartments” of only ~375 square feet!
Terrific post. It takes a very complicated issue and strips it down to its most salient, and most misunderstood, elements.
Great article Mark. One thing, if the 800 Square foot is cost for the building overall the apartment size you are talking about is more like $640, Excluding corridors, stairs, etc.
Excellent point. In the quest to make things as simple as possible I left out the efficiency of the building as part of the calculation, but it’s obviously a big factor in the cost.
Just curious – why are the things that contribute to the quality of the city – such as design review, community process and affordable housing fees thought to be a “burden” on new development, but land costs that have risen at unprecedented rates in the last several years are thought to be an indisputable, market driven, unchangeable input? Aren’t land costs defined by development potential? Shouldn’t developers in SF anticipate all these costs (inclusionary and processing costs) and negotiate lower land costs?
Developers do anticipate these costs when assembling a proforma for the project. I don’t think I used the word “burden” anywhere, but they have to be acknowledged in this calculation. I should also point out that San Francisco has the most notoriously difficult to negotiate planning process of any large city in the US and the costs are higher as a result, both for affordable housing and market rate.
Real estate is priced based on development potential- the height limits and possible density are taken into account and the land is priced accordingly. In the current market developers have no ability to negotiate land prices down, in many cases the land is bid up above the asking price (similar to the current market for single family homes).
You have combined city fees with “professional service fees” (presumaby, architect, engineers, etc) at 20%. What is the breakdown? How much could reducing the review times realistically reduce the project cost? More importantly, how much would any reduction realistically be reflected in the sale/rental cost? Are developers not getting sales prices and rents that the market can absorb, no matter the cost (assuming developer profit)?
I grouped all of this together because the professional services fees are related to the amount of time it takes to get approved, as the fee for the entitlements phase is often on a time and materials basis (at least for the architects). All of this depends on the specifics of the project. It also depends on how many consultants the developer may need to employ (or in the case of 8 Washington, whether a political campaign needs to be waged). I put 20% in as a conservative figure, I have heard much higher figures from developers in the past (as much as $90-$100k per unit).
The main reduction in cost would be from the reduction of time in moving a project to completion. If the approvals process ends up taking two years, the cost of carrying the land and the escalation in construction costs drive the overall price of the project up further over that two year wait. Many other cities make it easier for permits to be granted over the counter (as I’m sure you know). This was the topic of a recent forum at SPUR.
Developers are clearly getting the prices/rents the market will absorb. However, I was trying to illustrate the minimum housing could be built for under the current system as a way of showing that even when profit is taken out of the equation a typical new built unit is still above the budget of the average San Franciscan. These high costs also make housing construction less likely when the market is less than booming, because projects stop penciling out.
If you have any other feedback I would love to hear it.
So the delta between $460k construction cost and $800k – $1.2M asking price for these units is the developer’s overhead and profit, or am I missing something? Innovative cost-saving measures to make cheaper buildings serves the profit margin, but how does cheaper construction help build better cities, or even lower housing costs when demand is so high?
It’s actually a 640 square foot unit once building efficiency is factored in (see David Baker’s comment and a later post for the correction), so I think saying they are priced up to $1.2 million is a bit of a stretch. That would mean new units were selling for something in the vicinity of $1800/square foot which would be extreme even for San Francisco. Innovative cost-savings means that when we aren’t in a tech bubble (as I’d argue we are now) housing development is still feasible. Lower costs would also mean we could build more subsidized affordable units with the same pot of money we have now. I touch on this a little in a new post: https://markasaurus.com/2013/11/05/housing-costs-controversy-and-the-math-of-bmr-ownership/
All this assumes development by the private sector. Why ignore the fact that only the people can solve the peoples problems, and that is done through the institution of the people “government”. To expect private for profit companies to act in any way that helps slow the market rate cost of housing is deluded.
However the city has the ability to not only streamline the development process, thus bring down cost, has land that can be re purposed, and can build on a much larger scale, further reducing per unit cost.
The only thing lacking is the political will of the people to elect a government that is not corrupt, and in the pocket of those who profit most from maintaining the status quo.