Hacker News Reader

I'm an ex-SF'er and a "just build more supply" person generally who is basically just not there for arguments about how tech workers are ruining SF, and this is the first formulation I've read of this argument that resonated for me. I get the argument now: the long-term health of SF is an externality to tech workers. Yeah, that does sound right.

Thank you! I really appreciate that.

Totally. This is why both the state and the companies locating in California needs to be held accountable for improving the situation. Not the workers.

It sounds like something that could be trivially implemented by taxing high income brackets, and directing that money into subsidized low-rent housing.

Prop C baby! Although definitely not “trivial” to pass or implement. Took Marc Benioff to pour in millions and own Jack Dorsey on his own website which was awesome. And a lot of grassroots activism too.

That targeted homelessness, the most urgent issue, and did so by taxing (tech) companies as opposed to wealthy individuals. More generally the approach the city uses is to negotiate with developers for affordable units as a condition of project approval. Taxing the wealthy is an interesting idea but very far from trivial.

>Prop C baby! Although definitely not “trivial” to pass or implement. Took Marc Benioff to pour in millions and own Jack Dorsey on his own website which was awesome. And a lot of grassroots activism too.

And all that to raise a paltry $300 million a year, enough to build ~700 subsidized housing units, or a drop in the bucket.

>negotiate with developers for affordable units as a condition of project approval

A scheme in which affordable housing activists are motivated to minimize total development to preserve their negotiating leverage, and we "address the housing crisis" by encumbering development to make sure it will never catch up.

So you're suggesting that this isn't enough money? Then we'll pass another prop C :)

I’m not suggesting it. Solving housing affordability with BMR requires provisioning a BMR unit for every struggling renter household, i.e. almost every renter household.

Prop C is self evidently not enough money to duplicate every rental apartment, or even half of them. You’re really unlikely to find that much tax revenue anywhere. But you’ll easily find that much private capital ready to build market rate projects with BMR set asides, if it’s allowed to.

This is a token effort, not a serious attempt to grapple with the problem. I’m happy for the handful of people who will be helped by it. But to the extent that it convinces others they are “on the right track” or “helping” it is doing more harm than good.

I think you're convincing no one by saying that it will be doing more harm than good. At this point anything would help the situation. SF is unlivable now.

It's yet another election cycle with the focus anywhere but the colors on this map [0], which are an explicit and unambiguous choice to have precisely the situation we have. This problem is not going to get fixed a few thousand units at at a time.

[0] http://default.sfplanning.org/zoning/zoning_map.pdf

> And all that to raise a paltry $300 million a year, enough to build ~700 subsidized housing units

You cite no source and are grossly misinformed about prop C. It will build housing for 4k homeless with just half the budget and the rest goes to services specifically targeted at homelessness.[1]

[1] https://www.ourcityourhomesf.org/

$300m/year: https://ballotpedia.org/San_Francisco,_California,_Propositi...

Cost per subsidized housing unit: https://ternercenter.berkeley.edu/construction-costs-series

I presented a more optimistic view of what it could accomplish (produce 700 units a year), where your link indicates it will house 4,000 people in perpetuity. After 6 years, my “grossly misinformed” interpretation would have been a lot more bang for buck.

There are so many more than 4,000 people who struggle with housing that the framing of Prop C as some kind of victory is just bizarre. It’s like winning an increase of $5 per retiree in the pension system.

Nope, it’s less optimistic. The 4K of permanent units are funded by 50% the budget over 8 years; you’re suggesting 100% allocated to 5.6K units over the same time, a 30% worse bang for buck.

And you’ve cut the homeless out of the picture. The other half of funds is dedicated to mental health, addiction and other needed services for the homeless population. You’ve completely misunderstood the aims of Prop C.

Public housing costs more than $400,000 per unit to build [0], assuming you can find a site that the neighbors won't kill.

Increase the city income tax from 1.5% to 15%. You'd need around 9 $300k households, each contributing $45k, to build low-rent housing for one family.

[0] https://sf.curbed.com/2018/5/22/17380702/farrell-affordable-...

It's surprising to me that you'd jump on board that argument, based as it is on a lazy stereotype that's a poorly disguised variation on the "shallow self-absorbed millennial" trope.

It's clearly not true for me. I just bought a house here and plan to stay for the long term. Why would I not care about the long-term health of SF?

I didn't say you didn't care, I said that for many developers in SF, the issue is an externality. That's not synonymous with not caring.

A city of 880k shouldn't regard demand as inexhaustible after grudgingly tolerating 4400 new units. They should try again with at least one extra zero.

> 5. In order for the market rate price of housing to become affordable to cooks and servers, it would first become dirt cheap for tech workers making 2-4x their compensation. Therefore all of that worldwide demand for SF tech worker housing would need to be satiated before the market solves this problem.

Suppose you take a fraction S of city's residential surface area and build it up by H times average residential height. Housing volume increases by S * H. Most housing in SF is merely 2 stories high. Let's assume average is 3 stories. Now, pick 1% of city's residential area and build it up by adding 30 stories, say from 3 to 33. Note that 33 stories is actually smaller than a few existing tall residential structures in SF [1]. This would give relative increase in housing volume by S * H = 1% * 10 = 10% which given SF's population of >880k [2] amounts to >88k increase in housing capacity. This is more than the entire global workforce of Google [3].

It's a lot of construction work, of course, but it looks feasible. Many cities build a lot higher and on a wider area than the figures above.

[1] https://en.wikipedia.org/wiki/Category:Residential_skyscrape...

[2] 884,363 (2017), https://en.wikipedia.org/wiki/San_Francisco

[3] 85,050 (Q1 2018), https://en.wikipedia.org/wiki/Google

Edit: formatting and correction

I mean I don’t want to sound too discouraging but this is the problem with “plausible sounding” armchair urbanism. 1% of the surface area of SF is over one hundred city blocks of 33 story residential skyscrapers. And as you note, that’s not even enough to house the global workforce of the current tech companies, many of whom would move to HQ if housing became cheap along with crap tons of others.

Well, sure, go ahead. The point is market rate construction alone will not solve the service industry housing problem. There needs to be affordable price-controlled housing to support a healthy local economy.

>1% of the surface area of SF is over one hundred city blocks of 33 story residential skyscrapers.

You're writing that like it's ludicrous. Manhattan's ~2500 blocks are much more than 1% covered in high-density residential structures. So it's not even close to touching the examples we already have.

> market rate construction alone

I'm not sure why you keep bringing this up. Market rate construction alone is not a thing in San Francisco. The requirement is 18% BMR. If we zone for more capacity, it will be 18% BMR.

I am tempted to draw a parallel with NYC. And London.

NYC cleaned up and it used to be as gnarly as SF is now. SF is like 80s NYC.

Both London and NYC have excellent networked public transport. The population is all kinds of diverse. We don’t. People are free to move around. We are not.

SF politics is a stumbling block. Any kind of high density building plan will fail unless infrastructure and public transport network is bolstered BEFORE building more.

London has a way better public transit than nyc. It's not even comparable.

This is true. But there are good ‘bones’ in place for nyc. We don’t have that here in the Bay Area.

Demand is not inexhaustible, just pent up. On the other hand, there is a lot of pent-up demand and market YIMBYs are certainly in denial about just how much building it would take to satisfy that demand.

It would involve tremendous upheaval. Traffic would get worse. Many people would be evicted and have to find new housing before rents come down. Many people would find themselves underwater on their mortgages. We'd have to improve water/sewer/electric/gas infrastructure. We'd have to spend billions on transit. It would take at least 25 years.

I would love to see that happen, but it won't. Too many people are too invested in the status quo. Instead, they'll make SF so unattractive to tech workers that they'll leave. As you say, they're more invested in the tech industry than SF.

I think we’re saying some of the same things, but I’d argue the term “pent up” is not really the best fit for the condition of price inelasticity. That usually is the result of some extraordinary event. Boomtown economics qualify. We’re creating some of the most valuable companies here and are still in the thick of it. That’s just enormous demand for SF housing from all over the country and world.

Any affordable housing for the service industry would be dirt cheap for tech industry workers. If SF was dirt cheap again and the tech boom was still underway, who in the industry wouldn’t move here or start their next company here? That’s how deep the demand goes. That creates price inelastic conditions. It’s Economics 101 for those who actually paid attention.

If we want a vibrant restaurant scene moving forward, or housing for teachers and homeless and whatnot, we need to look beyond just building more market rate supply. There needs to be means tested and/or price controlled housing for the working class (or homeless, etc), and that’s where the market YIMBYs and affordable housing activists come to blows.

That doesn’t equate to making SF unattractive to tech at all. Marc Benioff came around on a local issue here about funding housing for the homeless (prop c). He saw it as an existential threat to Salesforce. Maybe Salesforce is just a more mature (longer term) company than most startups. But he saw that homelessness was actually threatening their bottom line in SF.. in addition to being this huge moral imperative. It’s kind of the same thing with restaurants and that type of quality of life: you can optimize for lowest taxes and maximum housing for tech workers, or you can see the longer term macroeconomic picture.

We fund new means tested housing as a percentage of new market rate housing. People who want to substantially increase the availability of either type have aligned interests here: as much development as possible.

In my experience, affordable housing activists want to make sure that if development happens it's price-controlled, but are fine with zero or a token level of development. It's the YIMBYs who are pushing for volume/scale for all kinds of projects. That's where the blows come. Percentage vs. quantity.

This is a false dichotomy. I call BS on your “fine with zero development” claim based on your “experience.” The most development in SF has happened in Jane Kim’s district who is a leading affordable housing activist.

Under inelastic pricing conditions simply building more market rate housing alone will not alleviate the crisis of the working class. The only solution is to build and protect affordable housing.

It's not a false dichotomy because it's not a dichotomy. Getting more housing of either type requires building more of both types. We don't build market rate housing alone, we build it in projects that are 10-25% price-controlled. Concerns about "simply building more market rate housing alone" are irrelevant here. What gets us more price controlled housing is more market rate housing.

The question is not which type we will build, but how much, and where. All San Francisco, including District 6, saw a net growth of 4,441 housing units in 2017 [0]. It needed more like 10-20x that just to tread water. So effectively zero. The battleground now is to increase the zoned capacity across the board, so that net-new hybrid housing stock has somewhere to go at scale.

[0] https://sf.curbed.com/2018/4/17/17247340/san-francisco-housi...

“Percentage vs quantity” is explicitly a dichotomy and you misrepresent what affordable housing activists are comfortable with.

We can agree that height limits are an opportunity to negotiate for affordable housing. I’d further suggest we prefer development in neighborhoods with less rent controlled housing so as to minimize the impact on existing affordable housing.

So we built a paltry 4,400 in the past, let's shoot for 44,000 units next year. At the already-negotiated 18% BMR, that's 7,920 price-controlled units, a massive increase in affordable housing capacity.

This line of reasoning is too simplistic. You have to look at the economic effects of the other 36,080 new high income earners, who will create new service industry jobs and thus place more pressure on afffordable housing. The BMR percentage may need to be substantially higher than that to relieve the restaurant crisis.

And that is exactly the volume vs. percentage schism I'm talking about. You're taking an increase in volume as something the other side wants, a concession that might be made in exchange for something you want.

Market rate housing does not spontaneously generate high income workers. Industry recruits them, regardless of housing construction. Those tens of thousands of workers are coming either way; the variable that can be controlled here is whether they gentrify existing buildings or divert to new ones. They're going to create more affordable-housing need by displacing people in existing units than they would in infill.

And if the BMR ratio goes up so high that projects don't pencil out anymore, then what?

Do you agree that there is a limit at which developers will just refuse to build more housing, affordable or not?

> The most development in SF has happened in Jane Kim’s district who is a leading affordable housing activist.

Yeah, until Jane Kim raised the affordable housing percentage so high that the pipeline is now virtually stopped, in a bid to (unsuccessfully, I might add) try to win an election against Scott Wiener.

Why can't they pay their chefs more until the chefs too can afford to live closer?

At least in part because restaurant margins are very slim, and customers are paradoxically price sensitive to menu prices despite being largely insensitive to rent (probably because everyone feels like they can go across the street to get a better value for food, but nobody feels like they can trivially move to a higher-value neighborhood).

There’s no paradox. Rent is a necessity, given that you want to live in and get your career going in SF. Restaurants are a luxury. Thus there are those who will spend as much as possible on rent (esp. if they think it will help them earn more later), taking it out of their going-out budget.

This. You can always get cheaper food and find it acceptable. Cheaper rent in SF is often no great...

Is this not obvious? Seriously, this question always comes up in these discussions. The answer is it’s a macroeconomic problem: SF restaurants have already increased prices to $34 for a pork chop and $13 for a cocktail (Nopa.. which I attest is a very good value by SF standards). Restaurant margins are low. If they triple their cook salaries it will result in a net decline and they’ll close.

What we’re seeing is a shift towards “fine casual” formats like Souvla that have lower labor costs. But there are just fewer midrange restaurants opening due to labor and rent and the ceiling on what they can sustainably charge. The high end and low end / fine causal are doing ok.

This. Restaurants in SF are already quite expensive !

So then why not let them close? Seems like the supply-and-demand problem will eventually reach a new equilibrium.

Because then chefs would be even worse off because half would lose their jobs? A better solution is to provide them affordable housing.

This was my first reaction too, especially after the line about “figuring out how to pay him more without having him lose access to different low-income programs for which he currently qualifies.”

But they get into the actual wages below, and while it is absurd to me that servers are making almost double what the cooks are, what’s even more insane is that both of them are paid an absolute shitton by restaurant standards nationwide.

50k per year puts you at the top end of Cooks in the entire country. Hell, anywhere but SF, that is still solid middle class wages regardless of job, and in a field where customarily only the truly high end even bothers to pay more than minimum wage.

I would’ve donated a spare kidney to make that kind of money when I was still cooking. And the idea that that is still not enough to survive even rent should be a massive alarm bell that something is deeply fucked with your economy.

From a european perspective it already seems mind-boggling that chefs and service staff make 80-90k and still can't live comfortably. I don't think that doubling down on that trend will help as long as there isn't enough housing available.

Thank you - it’s very refreshing to see these points here. So many comments here miss the main points of the article or default to tired lines from Econ 101. Yes more affordable housing is good an important but it takes a long time. And there are a lot of folks suffering in the meantime. To me the core problem is your point 6 - despite our power and wealth, not enough of us give a shit about long term community goals.

> 5. In order for the market rate price of housing to become affordable to cooks and servers, it would first become dirt cheap for tech workers making 2-4x their compensation. Therefore all of that worldwide demand for SF tech worker housing would need to be satiated before the market solves this problem. (See point 2.)

No, that is oversimplified. The "worldwide demand for SF tech worker housing" is actually limited. Why is this supposedly uniquely true for San Francisco and not for other places with Google offices that are more affordable, such as Atlanta?

> 6. Most tech workers don’t give a shit. They are here to hopefully make their fortune or at least get their career started in the next 5-10 years. A vibrant restaurant or working class community is not a priority. If they burn through it, it’s fine. Long term community development goals are a “nice to have”.

Don't speak for me. I absolutely care about a vibrant restaurant and working-class community. I donate to local charities and patronize local restaurants and shops. I just bought a condo for the long term. Why would I not care?

It is exceedingly annoying to be demonized as somehow less of a San Franciscan by your side because I happen to work in tech. You have no evidence to back up the assertion that we don't care; it's just a lazy stereotype.

> 7. Only reserved means-tested housing for the working class will solve this problem. This is what divides the “market YIMBY” and “affordable housing PHIMBY” political groups in local SF politics.

That is a very uncharitable way to describe the YIMBY movement. The YIMBY movement wants to maximize the amount of affordable housing, not just the percentage. The "PHIMBY" movement, by and large, refuses to adequately engage with this argument.

> The "worldwide demand for SF tech worker housing" is actually limited.

No evidence, just an "actually."

>If you want a vibrant city rich with interesting restaurants and cafes then you need to provide housing for the service industry

That sounds like a testable hypothesis. If we examine cities with interesting restaurants and cafes, we will find that most/all have substantial means-tested housing set aside for the service industry?

You’re forgetting the all important factor of tech-boom fueled demand. Pretty major omission there.

Also, means testing is not the only way of providing service industry housing. In SF what cooks and servers are left are hanging on to rent controlled units. As those are eliminated through gentrification (Ellis Act evictions, etc), the restaurant crisis worsens.

The runaway success of an economy bringing lots of people to an urban area is not some sort of world-historical anomaly. Economic growth built practically every city on the map, and endowed many more than San Francisco with large professional services sectors.

We should be interested in how they've dealt with these issues. Across the US, massive and centralized public housing complexes, for example, became catastrophic sociological failures. We probably shouldn't repeat those. Which is why we intersperse affordable housing in luxury buildings instead of doing towers in the park.

In New York, price controls fell apart, but the region stays affordable through an effective train network serving places with lower land prices. Could that be a better ROI?

In Germany, rent controls are widely successful at stabilizing people's lives while development catches up with demand. Are there examples of rent controls working well without the catch-up development?

Centrally planned housing assignments are not self-evidently the best way to solve this problem. They could be! But other cities in the world have faced these pressures and come out okay. Mostly by letting their built environments grow. Let's learn from them.

The tech workers aren't setting the rents, and they're not the ones taking money from the cooks. They're not the ones setting the politics, either (many of them can't even vote).

> The tech workers aren't setting the rents

They are the demand. The politics is the supply. Both are at fault for making SF unlivable for the people that were there from the start.

Some good points, but "vast majority... complete denial" is a bit much.

Happy to be proven otherwise.