Thread by @qwhelan

A fair number of people mentioned that a recent thread overemphasized the impact of the RH-1 component of the BOS teacher housing ballot measure.

I felt humbled for possibly misleading all of you.

Then I remembered I’m an idiot.

So I checked the data.

Here’s Sup. Fewer’s aide stating I was wrong, and also leaking some useful info (~3,000 units total)

I can now check my math!

Note that a rezoning does /not/ mean an upzoning. More on that later.

https://twitter.com/IanFregosi/status/1153926581351358466

Here’s my calculation on number of parcels touched by the rezoning.

I repeat, this does NOT mean higher density is allowed.

This is an approximate, but I’m in the right ballpark 😀

Replicating Ian’s numbers took me about 5 minutes.

I’m not bragging.

I just want to point how how simplistic the analysis behind the numbers is.

Here’s @MattHaney claiming RH-1 is a “small part”

https://twitter.com/MattHaneySF/status/1153920822446153728

@matthaney Here’s @AaronPeskin’s aide claiming the same:

https://twitter.com/LeeHepner/status/1154067109313306625

@matthaney @AaronPeskin So we’ve got 3 of the 4 votes for this measure represented here.

And as you've probably deduced, they're ... let's go with "not correct"

@matthaney @AaronPeskin So how do you figure out current zoning? With this easy to use, highly legible map!

There’s only 70+ distinct colors to distinguish. I have no idea why small businesses are closing 🙃

@matthaney @AaronPeskin Fortunately, we only need 56 of the 70+ to do this analysis.

Then you look up the lot size minimum in this helpful table. And then find a dozen more tables like it because this one only covers 13 codes

@matthaney @AaronPeskin Or… you ask @deapthoughts nicely and he points you to his github where he compiled this: github.com/vadimg/sfzonin… :D :D :D

Not only is he awesome, but so is his site: sfzoning.deapthoughts.com

@matthaney @AaronPeskin @deapthoughts So that gets us minimum lot size per unit for existing zoning.

What about the new zoning?

We go to the text of the ballot prop and parse this:

@matthaney @AaronPeskin @deapthoughts And we convert that into an effective minimum lot size assuming 500 sqft for a studio/1BR, 750 for a 2BR, and 1000 for a 3BR, plus 200 sqft for common space.

Assume we stack them up to the zoning max (40ft) and we get:

(0.5 * 500 + 0.3 * 750 + 0.2 * 1000 + 200)/4 = 218.75

@matthaney @AaronPeskin @deapthoughts If this is better than current zoning, we use it. If it’s worse, we keep existing zoning.

Here’s what we get.

Ruh roh.

@matthaney @AaronPeskin @deapthoughts Fully 72% of the “thousands” of units have no density changes!

We can already build affordable housing on the vast majority of these parcels.

And we haven’t even gotten to the bad news.

@matthaney @AaronPeskin @deapthoughts We compute the zoning density delta from before and after this proposition, and aggregate by RH-1 and not RH-1.

The claim that RH-1 is a small part of this bill is false. It’s actually the vast majority of new zoned capacity for affordable housing.

@matthaney @AaronPeskin @deapthoughts

@matthaney @AaronPeskin @deapthoughts I was legitimately shocked by this when I first computed it! It’s counter-intuitive at first.

But then you remember SB-827 and SB-50.

For identically sized parcels, upzoning RH-1 gives you far more bang for the buck than upzoning something that already allows AH.

@matthaney @AaronPeskin @deapthoughts You can see this in the ratios of before and after median units per lot.

Not-RH-1 goes up by 2.86x

RH-1 goes up by 14.25x (!!!)

@matthaney @AaronPeskin @deapthoughts Finally, does any of this matter?

After all, the ballot prop does stuff that simply can’t be quantified.

Everyone else can feel free to chime in on the relative significance of that other stuff compared to what I’ve presented.

@matthaney @AaronPeskin @deapthoughts However.

To the extent that the ballot prop claims to upzone *for affordable housing*, the claim that RH-1 is a *small component* is utterly, completely wrong.

It appears to be the vast majority of the upzoning in this bill, and it looks like this:

https://twitter.com/qwhelan/status/1153860358764810240

@matthaney @AaronPeskin @deapthoughts So we're down to 825 not-RH-1 parcels that will actually be upzoned by this proposal.

But wait! There's more!

@MattHaney's aide points out we should haircut that number by another 2/3rds, as most sites are infeasible

https://twitter.com/courtneyannmcd/status/1154416766719627264

@matthaney @AaronPeskin @deapthoughts So the best estimate for number of feasible sites that will be upzoned for affordable teacher housing by the BOS ballot proposition is:

*drumroll*

275 sites.

In the entire city.

@matthaney @AaronPeskin @deapthoughts I know all of you are here for the parcel porn.

To be up front, there's a lot less abject weirdness in this data set (though plenty in the non-upzoned part).

So I'll try to find some of both!

Interactive map here:
qwhelan.github.io/sf_bos_sfh/non…

@matthaney @AaronPeskin @deapthoughts This is 6 of the 825 upzoned lots.

It's better known as the Sunnydale-Velasco Public Housing Project.

It is currently being re-developed into new affordable housing under HOPE SF: hope-sf.org/sunnydale.php

So these 6 are not feasible, as we're already building AH here!

@matthaney @AaronPeskin @deapthoughts Here are 16 lots centered around the Ferry Building.

Any development on the waterfront requires a public vote per Prop B.

@matthaney @AaronPeskin @deapthoughts Here's... too many parcels to count by hand. 50-ish?

All conveniently located by scenic Toxic Waste Site.

Part of the India Basin project.

No new housing here, again these are all upzoned parcels.

@matthaney @AaronPeskin @deapthoughts Here's a big empty parking lot by the Stonestown Mall movie theater.

Perfect!

Except killed by the Shadow Ordinance, as it's surrounded by parks and playgrounds.

@matthaney @AaronPeskin @deapthoughts Here's a bunch of motels and parking garages along Lombard in the Marina that seem plausible.

Unclear if teacher housing's zoning advantage would let it outbid the private developers that would want these parcels.