Monday, April 5, 2021

San Francisco bans affordable housing

"San Francisco bans affordable housing," is the spot-on conclusion of a lovely post by Vadim Graboys (link to twitter). 

The post is titled "54% of San Francisco homes are in buildings that would be illegal to build today" with an interactive graph of those homes. 

Or, put another way, "To comply with today's [zoning] laws, 130,748 homes would have to be destroyed, evicting around 310,000 people."

The latter statistic is fun, but actually severely understates the damage of San Francisco's (and Palo Alto's!) zoning laws. The only reason current homes are illegal is that they were built under slightly less restrictive zoning laws. So that measures how much zoning laws have gotten stricter over time. It does not measure the much larger number of homes and apartments that were never built.

Now, how does San Francisco, ground zero of progressive governance, and a city whose politicians can't get out of bed in the morning, or sign permission to build anything without saying "affordable housing" about 10 times, in fact "ban affordable housing?" 

If your goal is to make housing as expensive as possible, the best way to do it is to require lots of land per home, because land in cities is very expensive. San Francisco does this in a multitude of ways, but I decided to focus on 3 of them:

Minimum lot size: San Francisco mandates that a lot must be at least 2,500 square feet (or 4,000 sqft in some cases) in order to build a home on it. Since a decent single-family home could be built on a 500 sqft lot, and a very nice single-family home can be built on a 1,200 sqft lot, the minimum lot size requirement greatly reduces the amount of homes that could be built in San Francisco and increases the price of each home. 65,974 homes would be illegal to build today because the lot size is too small.

Density limits: San Francisco mandates explicit density limits, from prohibiting apartments entirely (in 76% of the city) to mandating a minimum lot size per apartment (almost everywhere else). The sole reason for these limits is to make housing more scarce and therefore more expensive. San Francisco limits density in many other ways, such as floor-to-area ratio, setback requirements, etc. but to keep things simple, I looked only at the minimum lot size per apartment for this map. 69,499 homes would be illegal to build today because of density limits.

Height limits: When land becomes expensive, developers can still create inexpensive homes by building multiple apartments on the same plot of land. Each apartment has to pay for only a fraction of the land cost, which can make apartments significantly cheaper than single-family homes. The taller the apartment building, the less significant the cost of the land becomes, so San Francisco limits this savings by limiting the height of new buildings to 40 feet in most of the city. 9,066 homes would be illegal to build today because of height limits. 

Whenever San Francisco creates a new restriction, homes that already exist are grandfathered in rather than demolished. That's where all these "homes illegal to build today" come from. 

That also is why so much of San Francisco has an ancient housing stock. Many of the houses were built cheaply in an earlier era, and don't match current needs. If a house is grandfathered, you can "remodel" it within existing perimeter but you can never tear it down and build something appropriate to today's realities. 

Graboys asks 

How can you say San Francisco purposely bans affordable housing? All its leaders constantly talk about building more affordable housing!

and concludes 

Sorry to break it to you, but politicians lie. 

I think a more nuanced view has better power to explain behavior.  Politicians like government-constructed or government-controlled new housing marked as "affordable" and rationed by politicians. They just don't like "developers," i.e. people in the business of building houses, to build houses and sell them without paying proper homage to politicians. Later: 

I notice there's a building built last year that you claim is illegal to build. How is that possible?

Well-connected developers can apply for a conditional use permit to get a per-project exemption from zoning laws. The city does not need consistent reasons for granting or rejecting these permits, so this is an easy way for politicians to grant special favors to friendly developers.

I gather from a friend who knows developers in New York that "affordable" units in new apartments are negotiated in this process. 

More importantly,  politicians in a democracy are sensitive to the demands, however incoherent, of their supporters. Talk to the voters of San Francisco. They want some virtue-signaling government mandated "affordable" housing, little museums of the poor, but they absolutely don't want any such housing in their neighborhood.  In my neighborhood, the same people with the BLM / no hate here signs in their front yards also have no new development signs and show up to city council meetings to make sure nothing gets built. 

A separate post documents that "Apartment buildings are illegal to build in 76% of San Francisco" with another lovely graph.

Why? Graboys offers

These buildings ... are only illegal due to regulations called zoning laws that San Francisco passed to restrict the density of housing in certain areas, originally to keep poor minorities out of white neighborhoods.

...Because minorities were poorer than whites and dense housing such as apartments is usually cheaper than single-family homes, the goal was to ban the construction of housing that poor minorities could afford.

These days, many people who don't have explicitly racist motivations support zoning laws because they like the status quo, even though the status quo bans the construction of affordable housing in most of the city....

... even if the intent of zoning laws isn't racist, their effect has racist impacts. San Francisco's Black population declined from 13% of the city in the 1970s to 5.5% today. In contrast, zoning-free Houston added 100,000 Black residents from 2010 to 2015, and Black people are now 23% of Houston's population. 

This is a little superficial. While I think it is useful to remind our progressive friends, as they look to the uncomfortable racial past of American institutions, perhaps they should look to the uncomfortable past and unintended-consequence present of institutions they support, (next up: unions) I also think we should all step back from the "you're a racist" witch hunts of the current moment. 

It is also likely that allowing more construction in San Francisco, will not today directly create "affordable" new houses. It will create affordable housing via the musical chairs effect: fancy new apartments and condos will be built for techies, "affordable" because they only cost $1 million not $3 million. They will vacate the existing housing stock, which will then become affordable for people of modest means. 

When zoning laws were introduced, they had a myriad of motivations. People who had bought new houses were aware that neighborhoods, like the ones they left, could fall in to decline. Yes, they wanted to keep Black people out, but they primarily wanted to keep poor people out, and Asians, and low income whites (grapes of wrath) and so on. This fundamental desire was tinged with the unthinking racism of the time, but I don't think it's a useful reading of history to say that anti-Black racism was the only motivation, as fun a rhetorical point as that may be. 

The irony is that laws to keep poor people out are now keeping wealthy young techies out, and stopping the same neighborhoods from becoming wealthier, and thus better serviced by business and taxes. The counter-irony is that this is exactly what many people in San Francisco now want. They don't want 300,000 new residents. Zoning laws are as handy a tool against gentrification as they are against intrusions of struggling low-income minorities. 

This all may be moot, and they may get their wish, good and hard as H.L. Mencken used to say. 


"Well, you wouldn't want someone to build a 100 story apartment next door to you, would you?" No, I wouldn't. Of course as usual this is an extreme example. Neighborhoods change slowly. Even in Houston, nobody builds 100 story apartments in the middle of a residential block. 

But let us take it seriously. There is an externality. How do we handle externalities? (Quiz question). 

Property rights. (Quiz answer). For example, as a transition measure, each house could have the transferable right to impose current zoning restrictions on the immediately adjacent houses. Then, sure, you can build an apartment building, but first you have to buy the right to do it from the neighbors. What if they say no? Then build it somewhere else where someone will sell the rights. 

Perhaps better, each homeowner can be granted the right to trash zoning laws on his or her property, but anyone including the neighboring houses can buy that right. Now if you don't want an apartment going up down the street you can put your money where your mouth is, without having as now to buy the whole house. Per Coase, it doesn't matter to the allocation who gets the right. 


  1. All of these housing issues were true 10 years ago and people came anyways as the rents continued to skyrocket. I know it's Chic to say only software engineer salaries could support it, but the rents got so absurdly high that people were taking pay cuts to live in the city. On top of that, Sf was and is plagued by chronic homeless and drug abuse + car and domestic break ins. It seemed people didn't care, suggesting an extremely inelastic curve. However with zoom and a taste of life away from SF I'll be very curious to see if people ever return.

    I will also be curious to see how much progressive politics will continues to win the day when the tax base disappears and never comes back.

    As an aside, I lived in San Francisco for close to 4 years only a few years back. It's interesting to see there there are clear points of demarcation where homeless people congregate and don't congregate, suggesting that while progressive politics is the official City line; there are definitely some quiet but firm orders about where something is and is not allowed.

  2. 250 square meters is not that big. I really don't know how you could build a single family home on 50 square meters. Even 120 square meters sounds barely possible unless the house is 3-4 storeys tall. Of course apartments make more sense to increase density but 250 square meters is netirely reasonable limit for single family homes IMO.

    1. Yes. The writer has no idea what he is talking about. I live in Australia where we use square metres, but I've just built a 4 bedroom home on 220 square metres (2368 square feet), and it is tight. I'm not sure what you would get on 20% of that, but it sure wouldn't be a comfortable family home. Perhaps there is some confusion with what you could do with apartments.

  3. The owners of these properties pay taxes and fees. They purchased their properties on the expectation that they will have quiet enjoyment of those properties for decades, if not centuries. Zoning bylaws provide assurance that changes, if change occurs, will be conducted in an orderly and organized manner and that the current residents' voices will be heard and weighed in the decision to change such zoning bylaws.

    Judging from Vadim Graboys's other posts, along with parts of the post (quoted above) that are not included in the blog page, strongly suggests that Graboys is merely a quant with a superficial comprehension of the real world beyond the techie hinterland.

    Vadim states that a single family house can be built on a 500-square foot plot--really? Let's see. Single family lots are characterized by the lot setbacks for front, rear and side yards. Front yard setback = 10 feet, min. Rear yard setback = 20 feet, min. Sideyard setback = 5 feet, min.; combined sideyard setback = 10 feet, min. In this case, in order to maximize the available building area, the lot dimensions will be 12.909944 feet (frontage) by 38.729833 (depth). What, then, is the size of the building area? Answer: 25.4 square feet (8.729833 feet by 2.909944 square feet). This is also the optimal maximum building area for such a lot. Conclusion: Vadim is unknowledgeable about the subject matter. No surprise there--Vadim is a computer geek and a desk jockey--a whiz at the computer console. Does that make Vadim a good source for real estate policy ideas? Evidently not.

    The current property owners of San Francisco are under no obligation to make Vadim's life easy, or for that matter, anyone else's, either. Rotten luck, eh?

    Additionally, Vadim's notion that the price of housing is in some way related to the commodity prices of the commodities and labor used in constructing the housing is naïve. An example suffices. Single-family lots with 'old-timer' single houses on them located in a duplex-zoned neighborhood sell for $1,000,000, say. What is the selling price for one duplex unit in the new duplex house that replaced the 'old-timer'? Answer: $1,000,000 ($2,000,000 for both halves of the duplex house). Construction cost? $300,000-$400,000. Inference: The coder has not done his homework.

    Vadim is not a new phenomenon. His ilk were around in the late 1960s and early 1970s. Same point of view, same naïve utopian ideals, the same detachment from the real world, the same level of ignorance albeit on a different subject matter. The same indoctrination. The medium is the message; the message isn't worth the candle.

    1. You take setbacks as a given. There are numerous treatises on why setbacks don't make sense in urban environments, and why almost no buildings were built with them before post-war suburbanization.
      Indeed, bringing buildings closer to the street actually has a positive effect on the built environment, and is the reason why almost all "historic districts" are so popular, even in sprawling places like Southern California. The intimate sheltering of pedestrians within narrow streets and alleys only enhances the neighborhood character.

    2. Dear Old Eagle Eye;
      I notice you did a great deal of math to disprove a point that the author made with a link to a house on a 527 square foot lot. He did in fact state 'could be built" and showed it:


    3. Given the context of "land restrictions including setbacks bad", I expect 500-sq-ft is specifically excluding setbacks. Imagine them as rowhouses, two-storied, 20x20ft with 5ft between the front and back. They're not *nice*, sure, but it's doable.

    4. Yes, 415 Ivy St. SF, is an example of zero lot line planning in an older neighborhood in SF. There is no front yard. The entrance to the garage at ground level is set back by 4 feet. This example would be classified as a box car floor plan. Windows at front and rear; none on the sidewalls. In the case of fire in an adjacent building, reliance is on the abutting walls to slow the spread of the fire. Side yard set backs are desirable for two reasons--reduce fire hazards to or from adjacent lots and for windows for cross ventilation. The example house fits its older neighborhood. But it isn't a prototype for housing in other established neighborhoods with detached single family dwellings.

    5. "Zoning bylaws provide assurance that changes, if change occurs, will be conducted in an orderly and organized manner and that the current residents' voices will be heard and weighed in the decision to change such zoning bylaws."


      I suspect anti-zoning libertarians are guilty of a "Chesterton's Fence" violation: they've experienced the downsides of the current zoning system firsthand (and to be fair to them, NIMBYism and overly tight zoning codes are a big problem), but the far greater failings of the alternative nuisance law regime are just a vague, theoretical idea from ancient history.

  4. Excellent post.

    This problem of property zoning is national, even global.

    If you want to fight inflation AND raise living standards, then fight property zoning.

  5. This seems to follow a strange self-organization principle, with the exception that those with wealth and/or political power do the organizing. If Vadim's stars are correct, then this is a great way to force the "undesirables" out and force other municipalities to deal with the same problem, perhaps even in the same way. I suppose you could call it self organization by wealth/power, organizing along the lines of preferences. It's producing a condition where the equilibrium is really messed up.

  6. I want to see this diagram for every American City. Ben Cole and I are members of the Lost Cause Zoning Law Disruption Committee. But I'm glad this issue is getting more press in the last decade. I worry that it's a political dead end because every property owner in the country---regardless of professed beliefs---has a vested interest in the regulatory capture effects of idiotically restrictive zoning codes.
    Oh, and to respond to the Aussies who commented earlier. The architectural model for a 2000 s.f. house is the row house. It's an old solution to the density imperatives of scarce urban land. Can be bad in fires, though.

    1. I worry that it's a political dead end because every property owner in the country---regardless of professed beliefs---has a vested interest in the regulatory capture effects of idiotically restrictive zoning codes.--David Stuhlsatz

      As I always say, "There are no atheists in foxholes, and there are no libertarians when neighborhood property zoning is under review."

    2. I agree and am guilty of this myself, though my NIMBYism really only extends to the property directly attached to mine. I would incur significant loss if I had to move because the parcel behind me was cut up into N units. They could put nothing but beautiful 4k sqft houses there and it still ruins the value of my property for me.

    3. You don't need any special new committee... just join the actual zoning board of the place you actually live. Zoning boards are usually not THAT hard to get on for those so inspired.

      I suspect that with some exceptions (e.g., the Bay Area), you'll find the work being done is not what you expect, and the NIMBY people have some decent points.

  7. Many are confusing the INSTITUTION of zoning with the INDIVIDUAL PREFERENCES of local zoning boards. The entire point of "Euclidean" zoning is to make those individual preferences transparent to the broader community: it is very much premised upon a "sunshine is the best disinfectant" theory of government. Zoning boards must specify a consistent set of rules, and make any granted exceptions known to the public (and in most places hold a public hearing on exceptions).

    But somehow a century plus later we've forgotten all this. Zoning works perfectly, revealing the many biases and failings of local officials to the world, and we think the problem is the zoning system. No! Zoning is HOW you know the bad things are happening. The answer is to get involved and fix it, not scrap zoning.

  8. My concerns with densification of our neighborhoods, cities, or region all hinge on the infrastructure available to support it. In the Santa Clara county (San Jose, CA area) we suffer from traffic jams, water rationing, and power outages. The more crowded it becomes, the more anti-social behaviors affect us. Cramming in more people into resource-limited regions isn't making life better for the current residents.

    1. Oh yeah. I agree with you 100% this is a real problem.

    2. As with most things, the answer lies in the middle. Santa Clara can indeed add new residents; it just needs to also add the infrastructure to support all those new people, which it's perfectly capable of doing (if Riyadh can find enough water, so can Santa Clara). Most of the NIMBY "we already have too many people!" stuff is just not true.

      But on the other hand, the libertarian "eliminate zoning and just let people build!" approach would be an unmitigated disaster. Road capacity, water, electricity and schools can all be expanded, but nowhere near as QUICKLY as high-density apartment complexes can be built. Developers are really good at building apartments, while cities/towns are really bad at building infrastructure.* So growth must be limited to a level cities/towns can keep up with (which is not zero).

      *A good libertarian would fairly point out "but that's a self-imposed limitation! The sole reason we cannot build infrastructure quickly is overregulation!" And they'd be 100% correct in my opinion. But the overregulation needs to be solved FIRST: there's no "chicken or the egg" conundrum here.


Comments are welcome. Keep it short, polite, and on topic.

Thanks to a few abusers I am now moderating comments. I welcome thoughtful disagreement. I will block comments with insulting or abusive language. I'm also blocking totally inane comments. Try to make some sense. I am much more likely to allow critical comments if you have the honesty and courage to use your real name.