This year I read 30 books. The majority (17) were fiction. Only 4 were physical paper books. The rest were audiobooks.

I usually listen to audiobooks when walking or riding the train. But sleeping a baby provided a new time window for listening to books. Unfortunately for my book consumption we now have a toddler and bedtime routine no longer lends itself to book listening.

11 of the books were part of a series, accounting for nearly 2/3 of the fiction I read this year.

Only 2 books were for the book club I am part of. The club meets 10x a year, and so I only read 1/5 of the books and I actually only attended the book club meeting once. Hoping to be a better member next year.

Themes

Trilogies

One theme this year was trilogies. I read three trilogies this year.

In the spring I read the Binti trilogy by Nnedi Okorafor. All the books in the trilogy are fun and also short. Teenage me could have finished all the books in a day. The books were fresh. Fast-paced, full of unexpected developments. Lots of fun. Do recommend.

In the summer I read the Winternight trilogy, by Katherine Arden. The mythology and history of the medieval Rus setting were both thoroughly researched by the author, and it shows. The series mostly stays away from tropes, keeping it interesting. A lot of good imagery in here as well.

In the fall I read the MaddAddam trilogy by Margret Atwood. Atwood is famous for The Handmaid’s Tail. I was curious what her other books were like. By chance I picked up her other post-apocalyptic speculative fiction setting. What I found most interesting about the series is how the writing style changes in each book. The songs that punctuate the chapters in the second book are set to music and sang in the audiobook version. So that was fun and surprising. Of the three MaddAddam books I enjoyed the third the most. Probably for the bits of humor, which was sparse in the first two books.

Favorite Authors

Another theme this year was finding more books by authors I already like. One third of the books were picked this way. These were Ursula Le Guin (4 books), Margret Atwood (3 books, as previously mentioned), James C Scott (2 books), and one by Terry Prachett

I read Le Guin’s “The Left Hand of Darkness” a couple years ago and it became one of my favorite books instantly. Since then I have read a great deal of her books and enjoyed nearly all of them. This year was no exception. I read “Gifts”, “The Dispossessed”, “The Wave in the Mind”, and “The Found and the Lost”. I enjoyed each one, but I enjoyed “The Found and the Lost” the most.

“The Dispossessed” is a future sci-fi in the same setting and taking place before “The Left Hand of Darkness”. Interesting about the book is how it has an anarchic society as just a setting. My experience with sci-fi is that alternative political structures are seen as either flawless, and so the book is preaching utopia; or more often fatally flawed, and so the book is preaching a cautionary tale. This book does not have that preachy feel. The system is presented as flawed, but the book ends with attempts at reform. The book ends with the viability of the political system uncertain. Le Guin is not trying to preach an alternative society, but exploring what real people might be like under a different political structure. So it feels very real and human.

“The Wave in the Mind” is a collection of non-fiction essays and speeches by Le Guin. I should very much like to read it again actually, as I fear most of the content has dropped out of my mind already. Next time I will read with a notebooks, as I remember there were many interesting insights and I cannot now recall a single one. “The Found and the Lost” is a collection of short stories by Le Guin. There is huge breadth in style and subject matter - it showcases the author’s range delightfully. My favorites were “Buffalo Gals, Won’t You Come Out Tonight” and “Paradises Lost”.

“Gifts” is actually the first book in another trilogy. Perhaps I will finish next year.

James C Scott is the author of “Seeing Like a State”. Since I read the book in 2019 I keep returning to it in my mind. So this year I decided to check out if he wrote anything else. I picked “Weapons of the Weak” and “In Praise of Floods” mostly at random. Both books are interesting and I am glad to have read them, but I doubt they will be as impactful on my thinking as “Seeing Like a State”.

“Weapons of the Weak” is a case-study of peasant class resistance in rural Malaysia. It details the effect of mechanization and double cropping (the “green revolution”) on one village. The most interesting insight among many was how mechanization divided society. Before mechanization large landowners relied on tenants and small landowners for labor during planting and harvesting. There was simply no way even a large family could manage alone. After mechanization this was no longer true. Large landowners no longer needed labor from the rest of the village, and it showed. Feasts given by the wealthy in the village to which the whole village was traditionally invited were commonplace before mechanization, but within a few years they became rare. The wealthy essentially no longer needed to participate in village life, and many did not.

“In Praise of Floods” is about rivers, not social systems or governments (though it does touch on both). The best parts of the book were the high-level perspectives on how rivers function, how maps are deceptive, and how long humans have been messing up their river habitats.

Urbanism

I only read two books on urbanism this year, a large shift from the last five years. My focus is more on practical aspects now. I feel I know the theory well enough. Implementation seems to be the sticking point. So I feel it is maybe more productive to read municipal codes and city council agendas over another street design book at this point.

Movement: How to Take Back our Streets and Transform our Lives by Thalia Verkade and Marco te Brömmelstroet was interesting because it was not written for a US audience for once. It is a book by Dutch authors for a Dutch audience, and shows that despite the view from here things are far from perfect in the Netherlands. The main insight I got from the book, obvious in retrospect, is that the focus on moving people large distances quickly is wrong. The way I often encounter transportation questions is in the relative trade-off of different transit modes. For a journey of 5 miles, should it be train, car, bicycle? This book points out could we instead make the city is such a way that people do not need to be travelling 5 miles as often.

Understanding Cairo, by David Sim, I picked up by accident. I was looking to see if David Sims, author of “Soft City”, had any other books. Despite not being the author I was looking for, the book is excellent. I would say a perfect follow-on for “Seeing Like a State”, mentioned above. It is a book on how Cairo succeeded in housing its citizens by the one simple trick of not demolishing their homes. How informal systems can function when left alone, how governments think and act, and how one of the largest cites in the world lives, works and moves around.

Top Picks

If I had to pick a top book this year it would be either “A Room of One’s Own” by Virginia Woolf, or “How to Hide an Empire” by Daniel Immerwahr. But since these are also the last two books I have read: there is a high chance of recency bias. All books mentioned above are recommended.

“A Room of One’s Own” has humor, insight and beautiful sentences. It is also short. Could be read in an afternoon.

“How to Hide an Empire” is a history of the much-overlooked US territories, colonies and overseas territories. It is well researched, fascinating and often infuriating. Sunlight cleanses, but US overseas possessions are rarely in the light. Neither on the map or in the news, the history is one of abuse and neglect.

All Books

I got curious how the sliding scale that the downtown plan used for housing affected the number of total units that can be built in the area.

The Downtown Precise Plan uses a sliding scale to allow more housing density on larger parcels than on smaller parcels. The bulk of the plan area is subject to the table below:

Many of the lots in downtown are small. I was curious how much more housing would be allowed if we got rid of the sliding scale and just allowed the max density for all parcels.

I analyzed downtown nearest the transit center. I left out the parcels facing Castro St, since special rules apply to development on those parcels. The area analyzed 88 parcels with a combined total area of just over 30 acres.

The results for these parcels are as follows:

In the area analyzed the plan currently allows for a max of 1,020 homes. Removing the sliding scale would allow for a max of 1,346, an increase of over 30%.

Mountain View has had several passes at what it calls transit-oriented development in the past, most recently with the East Whisman Precise Plan. The (so far mediocre) results of the plan illustrate that the city does not actually do TOD. What the city does is create zoning that allows for TOD, and then hopes someone else does the actual development. I call this Transit-Oriented Zoning, or TOZ.

There are some pushes at the local and state level to get more housing zoned near transit. Below I analyze the half mile around the Mountain View Caltrain station, looking at how much housing is built, how much housing is allowed to be built, and what changes are currently being considered.

Current Conditions

Current Zoning

The City’s general plan provides general designation of land use. The zoning map subdivides these areas with more specific rules, and when even more fine-grained control is desired the city uses a precise plan

Most of the station area (57%) is designated as exclusively residential, about half low-density and half medium-density. An additional 34% of the area is designated mixed-use, allowing residential in medium to high densities.

Most of the area around the transit center allows some kind of home to be built. The most common residential designation is R3, which allows ~30-40 homes per acre. The second most common is R1, which allows one home on a minimum lot of 6,000 sqft; about 7 homes per acre.

Zoned Capacity

I am not accounting here for state laws such as SB-9, ADUs and JADUs which would increase densities, especially for low-density areas. Nor am I accounting for setbacks, FARs, minimum width, minimum frontage, density bonus laws, and so on that affect densities depending on development type, lot size and other factors. These numbers should be taken as approximate.

Using these simplified calculations I estimate the number of residential units currently built in the station area to be about 4,500. And the number of units allowed to be built under the current zoning to be around 6,500. This superficial analysis gives us around 2k homes that can be built in the station area with no changes to zoning.

The superficial analysis has 2 major flaws:

1. Many areas are already built out at more than their zoned capacity. Using variances, conditional-use permits, and density bonus laws the city council can allow for a development to be built at higher density than zoned for. Additionally many buildings predate the city’s zoning laws and were built to higher density than is currently allowed. For example many single-family homes in old mountain view have a 4-plex or 6-plex in the backyard. Completely illegal under current zoning.

2. Many areas are very unlikely to be redeveloped. Condo or rowhouse developments do not have a single owner and are unlikely to add capacity even if allowed. City-owned property such as parks and city hall are unlikely to be redeveloped into residential housing.

Most of the R1, R2 and R3 areas were downzoned back in the 70s, so it is unsurprising they are mostly at capacity. Available zoned capacity is mostly in precise plan areas.

Downtown

The zoned capacity in downtown is over 1,000 homes. Over half of that capacity is on city owned land, most of it currently used for free parking. The city owns 9 surface lots and 2 parking structures within a half mile of the transit center. The city has plans to build housing on one lot, and build a new parking structure on another.

The areas downtown which currently have housing are built at maximum zoned capacity. The office and commercial areas have a capacity of around 350 units.

The downtown precise plan strongly favors consolidation. The larger the lot, the greater the allowed density. This is why the development on lot 12 can provide 120 homes: it will consolidate 8 parcels into a single development, effectively doubling the zoned capacity.

This is also why most of downtown does not have additional capacity in the map. Theoretically if someone were able to purchase many adjacent parcels they could build more homes than this map indicates.

Most of the multi-family zoning in the city has this idea of higher density on larger lots, but it is most extreme in the downtown precise plan. A 6,000sqft lot has the same zoned capacity as R1: low density. But join three 6,000sqft lots together and the allowed density grows by over 5x.

Evelyn Avenue Corridor

The second largest precise plan in the station area is the Evelyn Avenue Corridor Precise Plan

The plan consists of low, medium, and high density areas, as well as a commercial area.

Most of the area is built out to max density. Remaining un-redeveloped are SFV Portuguese hall, La Fiesta, and two auto repair shops. If these were built to maximum allowed density it would yield around 30 homes.

Middlefield

Between Moffett and HW 85 are three large areas with available capacity of over 800 homes. However two of these are currently occupied by large condominium complexes. So only one, 555 W Middlefield, is likely to be redeveloped.

555 W Middlefield is interesting. It is zoned as a precise plan, but neither the city nor the owners seems to know where the actual plan is. So I used the maximum density allowed under the general plan to figure out it’s “zoned capacity”. There is currently a proposal to build an additional 323 units at the site, which would bring it in line with the general plan maximum. This is at the third such proposal, with the earliest one I could find from 2015. The previous proposals (for the same number of units) were approved by council, but nothing was built.

Moffett

Most of Moffett Blvd is currently zoned CRA, which allows for both commercial and residential. However, there is a minimum lot size for any development in a CRA zone of 20,000 square feet. Most of the lots fronting Moffett are smaller, so cannot be redeveloped under current zoning without consolidation.

The city is working now on a new Moffett Boulevard Precise Plan, which could potentially result in an increase in allowed densities/

Changes

City Plans

Aside from the Moffett Blvd Precise Plan update just mentioned, the city is also working on an update to the Downtown Precise Plan. Both these projects are very early in the process so it is uncertain if an how much the might affect allowed densities in these areas.

Nearer completion is the R3 update. This initiative started out with the possibility of increasing allowed density for all R3 in the city. It seems like it will end with an increase to only some of the R3 areas. And only one such area is within a half mile of the transit center.

This would add over 1,000 units of zoned capacity to the transit center area. 90% of that zoned capacity is in a single apartment complex, which has the same owners as the 555 W Middlefield project mentioned above. So maybe don’t expect anything to get built there anytime soon.

Regional Incentives

The Metropolitan Transportation Commission does not have authority to make any housing policy. However, it can influence policy by use of grant funding via The One Bay Area Grant (OBAG). In order to receive over eight million dollars in cycle 3 the city of Mountain View had to, among other things, comply with state law re: accessory dwelling units and density bonuses.

As part of cycle 4 MTC is creating a Transit Oriented Communities (TOC) Policy. The policy applies to areas within a half-mile of transit centers, the same area we have been analyzing so far.

To be eligible for funding under TOC a city must, among other things, establish a minimum density for residential developments. The transit center area is “Tier 3”, meaning that to qualify under that plan the zones within a half mile of the transit center must have a average minimum density greater than 50 homes per acre, and cannot have an average maximum density less than 75 homes per acre.

“Average” here means some areas can have lower densities as long as other areas have higher densities to make up for it. So if the city rezoned the R3 and Precise Plan areas, which make up over half of the total area, to have a minimum of 100 and maximum of 150 homes per acre the city would then qualify for OBAC 4 TOC grants.

State Law

In the news recently is Senate Bill 79. This bill, if passed, would override city zoning to establish minimum residential densities near transit.

The current version of the bill would define the mountain view transit center as “Tier 1”, and as such would establish a density of up to 120 homes per acre within a quarter mile of the transit center, and up to 100 homes per acre within a half mile.

The bill allows a local jurisdiction to alternatively substitute their own transit-oriented development plan, provided that it allows for at minimum the same number of homes to be built as would be allowed under the state law.

Superficial analysis suggests a plan that complies with SB79 would need to zone for around 25,000 additional homes within the station area.

Density and Intensity

Before concluding I want to go back to the Evelyn Avenue Corridor Precise Plan. This development actually contains the densest housing development in the station area.

On the left, at 61 units/acre, is the densest development in downtown. On the right, taking up the same amount of space (three floors and one city block), is a rowhouse development with a density of 16 units/acre.

Not only is density hard to visualize, but two developments can have wildly different densities and you would not even realize when walking past. Density has little to do with how peasant or unpleasant a development is to live in or to live near.

How we get “livable density” is beyond the scope of this post (which is already very long), but there are many books on the subject. My favorite is Soft City

Conclusion

Many people (myself among them) believe the best place to build dense housing is near transit. (I believe the opposite is also true: build more transit near housing). Mountain View does not build housing, but it does establish restrictions for what housing can be built. And most of the city is already built out. The city’s housing plans have only modest changes envisioned for the transit center area - we are relying on East Whisman and North Bayshore to fill out our housing needs.

Relying on a single developer has bitten us before. In 2006 the city created the Mayfield Precise Plan. This plan added over 1000 units of zoned capacity to the area north of San Antonio Station. To this date none of that housing has been built.

If we want housing to actually get built it makes sense to do broad upzoning. I am not alone. Both the state and MTC have plans that would require an order of magnitude more zoned capacity than currently exists. Allowing housing to be built is only the first step to actually getting housing built. But it is by far the easiest step.

Mountain View is doing well by this metric. More than 80% of the city is within a quarter mile of a park.

There are two large areas that are more than 3/4 of a mile to the nearest park. These areas are in North Bayshore and East Whisman. Both these areas are historically office/industrial. The city has plans to turn both these areas into mixed use developments with more residential and more parks.

These projects are both on hold for the moment, but when the housing does come it will come with many new parks.

The answer depends on the length of the trip. And which trips we optimize for is informed by who we are building the transit system for.

Toy Model

I have written a toy model to illustrate. This model calculates the total trip time using top speed, acceleration, dwell time, and frequency.

The baseline is taken from the VTA orange line. Use the sliders to see the effect of changing top speed or frequency.

The orange line runs in a street median for most of it’s length. So it’s top speed is 35mph. The stops are about a km apart. The schedule has it at around 2 minutes between stops on off hours, and a bit longer during peak. I could not find data on dwell time or acceleration for the VTA light rail. I have assumed 20 seconds dwell and 1.2 meters/second² acceleration, based off of BART numbers.

The model does not take into account the light rail getting stuck at stop lights or the very slow sharp turns it sometimes makes.

But it is good enough to illustrate the effects of changing the speed or frequency. If you are interested in actual speeds you can find those at calitp

For trips of fewer than 12 stops doubling the frequency (8 trains an hour instead of four) has a bigger effect on total trip time than doubling the top speed. We may also want to factor for the fact that people perceive time 1.2x as long when waiting for transit.

Thoughts

In Human Transit Jarrett Walker points out that people in the USA are largely unfamiliar with transit, and so tend to judge transit projects in comparison to the mode of travel they are familiar with: the personal automobile. This creates some wrong intuitions into the trade offs involved.

A second trend is a hyperfocus on commute trips and “reducing congestion”. This manifests in a focus on longer trips – people travel further for work than for shopping, dining, haircuts, etc. On average.

In discussions about the VTA, slow speed is often brought up as a reason for it’s low ridership. Especially through downtown, where the top speed is 20mph. Frequency is rarely mentioned.

This report shows an average ridership response of 0.5. That is, a 50% increase in frequency will on average result in a 25% increase in ridership.

Of course this is an artificial dichotomy. There is no reason transit cannot be both fast and frequent. This is especially true in the way transit is often funded, with operations (running the transit) having a different funding source than capital improvements (building infrastructure and equipment)

There is also more to making a train “fast” than top speed. “Transit Signal Priority” for example can ensure the bus or train always has a green light at intersections. Level boarding can shorten dwell times. Electrification and lighter vehicles improves acceleration. And so on. I present here only an illustration of the relative importance of “top speed”.

Bonus

Here is the model described above, but with all the sliders so you can change parameters and see their effect. Average service levels for Caltrain (local from SJ to SF), BART (orange line) and VTA (orange line) are given as a benchmark.

This overabundance of traffic infrastructure not only takes up a lot of valuable land and emits tire an break particulate matter across the city’s neighborhoods, but it also decreases the functionality of the freeways themselves. Every freeway entrance and exit contributes to congestion

Especially when the ramps are too short, like at Charleston

Gathering Data

I downloaded the Santa Clara County road centerline dataset. Then I filtered out only those whose roadclass was Ramp or Freeway. Finally I removed all ramps that were not joined to a freeway (ramps for expressways, mostly)

I eventually realized that untangling the mess to determine a specific discreet count of freeway entrances and exits per jurisdiction would not work. Too many edge cases. There was no way to automate it, and I wasn’t about to do the calculation by hand.

What I realized I could do was calculate the number of linear feet of freeway and ramp in each city. This would approximate how much real estate was dedicated to freeway infrastructure.

Results

Suspicions confirmed, Mountain View indeed has the most land devoted to freeway of any city in the county.

Measure G will be a new revenue measure on the ballot this November. This is an increase in the transfer tax (assessed when a property changes ownership). It applies only to properties sold for over $6 million.

Measure AA is a replacement of the existing Measure B, a parcel tax that was passed in 2017 and will expire in 2025. The text of Measure AA is not the same as Measure B.

Shoreline Tax. In addition to the two ballot measures there is an ongoing dispute between the city and the school district over the allocation of tax revenue in the Shoreline Park District.

Measure G

Measure G would be a

property transfer tax, imposed on residential/commercial property sales above $6,000,000 only, up to $15.00 per $1,000, generating about $9,500,000 in locally controlled funding annually until ended by voters, with independent audits

In addition, on September 10, the council approved a [resolution] to allocate the funds, should the measure be approved by voters.

There already exists a transfer tax of $3.31 per $1,000. It was enacted in 1973 and is referred to as a “Conveyance Tax” in the city code. This revenue does not have any special allocation.

This new transfer tax would create a progressive tax similar to San Jose Measure E.

Measure AA

Measure AA is meant to replace Measure B. The existing Measure B text describes

a tax of $191 per parcel – an increase of $64 for most homeowners – for 8 years, generating $2,800,000 annually, with independent oversight, exemptions for seniors and funds staying local

In Measure AA the tax is

$0.15/building area square foot capped at $1,750/parcel and $25/unimproved parcel, to be adopted, raising approximately $5,400,000 annually for 8 years, with senior exemptions, citizens’ oversight, and funds staying local

Changing from a flat rate to a per-square-foot tax is an attempt to make the tax more progressive. Under prop 13 even voter approved taxes cannot be based on the actual value of the property. So charging by the square foot has become a common approximation of value.

The cap is probably so that large property owners don’t try to kill the bill.

A home of 1273 square feet would pay the same under the new measure. Everyone else will have a different tax bill. Most would see an increase.

There is an oversight committee that meets 3-4 times a year. Documents and minutes on how Measure B funds were spent can be found here

For the 2022-23 school year these were the high-level numbers

Shoreline

This one is messy. I will sum up as best I can.

The Shoreline Park District is a separate legal entity from the City, funded by property tax increment, with its own budget and financial statements. The Mountain View City Council serves as the Shoreline Park District’s Board of Directors.

The Shoreline Park was created by an act of Congress in 1969. Prior to it’s creation the area was used for a landfill, and had frequent flooding. Shoreline’s creation made it exempt from the normal Prop 13 1% allocation. It is still restricted to 1% in property taxes. Just the distribution is not the same. It is all kept within Shoreline.

In 2005 the city and the district signed an agreement to give the school districts part of the property tax. Over 99% of the tax revenue in Shoreline is from non-residential property.

When the North Bayshore plan was approved last year, the city offered a new agreement, where all new residential tax revenue would be split with the school district at the same rate as the rest of the city. That is, Mountain View Whisman School District would receive 23.75% of tax revenue from all the new residential construction that was theoretically going to happen under the new plan. The city also offered to pay to the school district around 10% of the tax revenue for new non-residential construction, about double what it had been paying before.

The city met with the two school districts (Mountain View Whisman and Mountain View Los Altos High School) last year. The city was able to come to an agreement with the high school district. But not with Whisman.

The Whisman school district has repeatedly made disingenuous statements when arguing for the money they think they are owed. It seems they won’t be satisfied with anything less than the full 23.75% for all property. On the districts web site they repeatedly refer to their “full share” of the revenue. But it is not theirs. It was never theirs.

No other entity (Santa Clara County, Foothill-DeAnza College, Midpeninsula Regional Open Space District, El Camino Hospital, etc) has come to the city demanding their share. And they all get no tax revenue at all from Shoreline.

There is also this presentation I sat through where they repeatedly refer to not getting 23.75% as a “cost to the district”. Office buildings do not generate students. They cost the district nothing. All tax revenue from nonresidential is pure profit for any school district.

Whisman has no actual leverage with the city though. There is no legal obligation for the city to give them anything. So the district has been trying to get the public on their side.

Conclusion

I think Measure G and Measure AA are fine. I would have preferred Measure G without the predetermined allocation. It is unnecessary, as there is already plenty of public input on how the city spends it’s revenue. But it is fine. Good enough.

If Measure B was good, then Measure AA is also good. Costs of things have gone up a bit, and I like the semi-progressive nature of tying it to floor area. Would have preferred it without the cap.

However, the Shoreline fiasco has made me unhappy with the current leadership of the school district. The district has been acting like a child who, after being given a cookie, shouts they want two. And then threatens to throw the first cookie in the garbage unless they get another.

Housing Element

In 2010 city presented their “housing element”. This document contains the city’s 8-year housing plan. The housing element covered the years 2007-2014. To prepare the housing element the city contracted the firm BAE.

RHNA

The state of California creates a needs assessment for each city and county. The state had assigned Mountain View the task of ensuring at least 2,599 homes would be built from 2007-2014. When the housing element came before council 892 homes had already been approved. So the city only needed to plan to get 1,231 more units built.

The housing is additionally divided into income buckets, with very low, low, moderate, and above moderate income divisions.

Predicting Growth

In table 3.9 there are population and employment predictions.

The housing element predicted the population in Mountain View would grow from 73,900 in 2010 to 81,000 in 2020. A growth of 7,100. The 2020 census put our population at 82,376. The prediction was off by 1,376. I call that pretty close.

The housing element predicted the number of jobs in Mountain View would grow from 52,610 in 2010 to 60,690 in 2020. A growth of 8,080. LODES data from 2020 shows 94,529 people worked within the borders of Mountain View. A growth of 41,919. Five times what was predicted.

This is one reason why the jobs-housing imbalance - already a problem in 2010 - has gotten so much worse

Predicting Capacity

Site Inventory

To show that the housing would be built, BAE and city staff identified 32 sites where housing could be built, and estimated a “realistic capacity” for each site. The total “realistic capacity” for all sites was 2,154 homes. This capacity was much more than the remaining state-assigned total of 1,231 homes. So staff concluded that Mountain View could meet it’s obligations, and no rezoning would be necessary.

What Was Built

Today, in 2024, 12 of the 30 sites staff identified have no housing built.¹ Of those with housing, only 10 of the projects were approved before the end of 2014. And only 8 of those 10 were approved under the existing zoning. Those 8 projects, approved under the existing zoning and before the end of 2014, created 567 new homes. About a quarter of the predicted total.²

Below is the site inventory maps from the housing element. Sites which had a housing project approved before the end of 2014 are in green. Sites which have had housing built on them since that time are in yellow. Sites which still have no housing are in red.

The total of all housing projects that have been approved on those sites, as of May 2024, is 1,412 homes. About 2/3 of the predicted realistic capacity.

How Good Were the Predictions

For the most part the predicted capacity was correct. Some had fewer homes, some had more. The total predicted capacity for all sites on which new housing has been built was 1,508 homes. 1,412 homes were built. That is a pretty good prediction of capacity.

The real shortfall is mostly due to the timelines. Many factors need to align in order for new housing to be built, and even a site in a prime location may not see development for decades.³ Even with the enormous housing pressure in the bay area, it seems we should have a site inventory at least 3x our target.

One solution would be “broad upzoning”

Footnotes

Margaret Abe-Koga and Sally Lieber are both running for the Santa Clara County Board of Supervisors District 5.

Since both served on the Mountain View City Council together, it is possible to have an apples-to-apples comparison of their vote record.

I used the legistar API to fetch every vote where they voted differently. The results are below.

Much further below is the code I used to fetch the data, in case you are curious. Many jurisdictions use legistar, but the api is poorly documented.

The Votes

Excluding being absent or recused, Abe-Koga and Lieber differed in votes a total of 29 times in two years.

I have presented the data below exactly as it was retrieved from the api. Some context is lost, however, so curious readers are encouraged to click “details” to find a page with full minutes and video recordings.

In particular, the votes shown below may not be the final votes, as in some cases there are several motions associated with an agenda item. Only those votes where Abe-Koga and Lieber voted differently are documented here.

May 25 2021 City Council Meeting

details

Support for the Principles of the Mayors for a Guaranteed Income Initiative and Support Basic Income, a Cash-Based Pilot Program

June 22 2021 City Council Meeting

details

Charities Housing Notice of Funding Availability Proposal-1265 Montecito Avenue

September 14 2021 City Council Meeting

details

Rowhouse Development at 570 South Rengstorff Avenue

September 28 2021 City Council Meeting

details

Approval of Council Advisory Body Work Plans

October 12 2021 City Council Meeting

details

Public Safety Building, Design, Project 20-49-Various Actions

October 26 2021 City Council Meeting

details

Shoreline Boulevard Interim Bus Lane and Utility Improvements, Design, Project 16-58, and Phase I Construction, Project 18-43-Various Actions

December 14 2021 City Council Meeting

details

Extension of the Castro Street Closure, Extension of the Fee Waiver and Suspension of the Renewal Requirements for Sidewalk Café Licenses, and Amendment of Professional Services Agreement, Project 20-58

February 08 2022 City Council Meeting

details

Residential Development Project at 555 West Middlefield Road

The Council returned to the main motion.

MOTION - M/S - Showalter/Lieber - To:

March 08 2022 City Council Meeting

details

MOTION - M/S - Ramirez/Showalter - At 11:25 p.m., to continue the meeting beyond 10:00 p.m.

May 10 2022 City Council Meeting

details

Plant-Based Eating Resolution and Program

Residential Development at 555 West Middlefield Road

May 24 2022 City Council Meeting

details

Shoreline Boulevard Interim Bus Lane and Utility Improvements, Projects 16-58 and 18-43-Various Actions

June 28 2022 City Council Meeting

details

Mobile Home Rent Stabilization Ordinance-Amendment Related to Accord or Memorandum of Understanding Exemption

The amended substitute motion failed by the following roll call vote:

In response to Councilmember Kamei, Councilmember Lieber indicated support for more robust outreach to mobile home residents.

September 13 2022 City Council Meeting

details

Mobile Home Rent Stabilization Ordinance-Amendment Related to Accord or Memorandum of Understanding Exemption (Second Reading)

September 27 2022 City Council Meeting

details

Appropriation of State Community Development Block Grant Homekey Funding for the LifeMoves Mountain View Project

October 11 2022 City Council Meeting

details

Castro Street Pedestrian Mall-Various Actions

MOTION - M/S - Abe-Koga/Hicks - To:

MOTION - M/S - Abe-Koga/Showalter - To:

The Code

Mountain View uses legistar to store and manage public meeting information. Legistar has an API, though the documentation is sparse. So a lot of what I know about it is from trial and error.

First I get the ids of Abe-Koga and Lieber from the /Persons endpoint

curl 'https://webapi.legistar.com/v1/mountainview/Persons?$filter=PersonLastName%20eq%20%27Abe-Koga%27%20or%20PersonLastName%20eq%20%27Lieber%27'

Abe-Koga’s id is 232, and Lieber’s is 513. Next we get all the votes each has cast

curl https://webapi.legistar.com/v1/mountainview/Persons/513/Votes > Lieber_votes.json
curl https://webapi.legistar.com/v1/mountainview/Persons/232/Votes > MAK_votes.json

I wrote a python script to find items on which both had voted differently

import sys
import json
from collections import defaultdict

combo = defaultdict(list)
json_in = list(sys.argv[1:])

for fn in json_in:
    with open(fn, 'r') as fob:
        for vote in json.load(fob):
            combo[vote['VoteEventItemId']].append(vote)


ignore_votes = ('Absent', 'Recused')


def key_votes(votes):
    return {v['VotePersonName']: v['VoteValueName'] for v in votes}


def differing_votes(votes):
    vote_values = set(v['VoteValueId'] for v in votes
                      if v['VoteValueName'] not in ignore_votes)
    return len(vote_values) > 1


combined = {k: key_votes(v) for k, v in combo.items()
            if len(v) == len(json_in)
            and differing_votes(v)}

json.dump(combined, sys.stdout)

And run it to get a json keyed by itemid.

python3 mergevotes.py MAK_votes.json Lieber_votes.json > combo_votes.json

Unfortunately there is no API for event items. The only way to access details about these votes is on the /Event/(EVENT_ID)/EventItems endpoint. But we do not know what the event ids are. So I did an exhaustive search of all events and looked for matching item ids.

I was able to narrow the search with 2 pieces of information. First, I was only interested in city council events, so I could filter only the events with EventBodyId equal to 138 (the city council body id). Second I knew that Lieber and Abe-Koga were both serving in council from 2020-2022, so I restricted my search to those years. It took a few minutes but it did eventually return results.

import sys
import json
import urllib.parse
import urllib.request


with open(sys.argv[1]) as fob:
    vote_data = json.load(fob)


item_ids = set(map(int, vote_data.keys()))
event_url = ('https://webapi.legistar.com/v1/mountainview/Events?'
             '$select=EventId&$orderby=EventId&'
             '$filter=EventBodyId+eq+138+and+year%28EventDate%29+ge+2020'
             '+and+year%28EventDate%29+le+2022+and+EventId+gt+')
item_url = ('https://webapi.legistar.com/v1/mountainview/Events/'
            '{event_id}/EventItems?')


# odata3 does not have an "in" operator
# I tried to do EventItemId eq x or EventItemId eq y ... etc
# But the api barfed after about 20 items. So in the end I used a range
# and did final filtering server side.
def make_filter(ids):
    return f'$filter=EventItemId+ge+{min(ids)}+and+EventItemId+le+{max(ids)}'


# events has a hard limit of 1000, so I built in pagination
# after putting in the body_id and year filters there are fewer results
# so this is probably unnecessary and overengineered
def fetch_event_ids(last_event_id=0):
    if not item_ids:
        return
    url = event_url + str(last_event_id)
    req = urllib.request.urlopen(url)
    results = json.load(req)
    if results:
        for result in results:
            event_id = result['EventId']
            yield event_id
        yield from fetch_event_ids(event_id)


def find_items(event_id, item_ids):
    if not item_ids:
        return
    url = item_url.format(event_id=event_id)
    fltr = make_filter(item_ids)
    req = urllib.request.urlopen(url + fltr)
    results = json.load(req)
    if results:
        for result in results:
            if result['EventItemId'] in item_ids:
                item_ids.remove(result['EventItemId'])
                yield {**result, **vote_data[str(result['EventItemId'])]}


def fetch_items():
    for event_id in fetch_event_ids():
        for item in find_items(event_id, item_ids):
            yield item


item_data = list(fetch_items())


# now we have all the event items, I want to get the event metadata
event_ids = set(e['EventItemEventId'] for e in item_data)
# InSiteURL is the web link with links to videos and other relevant things
select = ('EventMinutesFile', 'EventInSiteURL', 'EventDate', 'EventComment')


for event_id in event_ids:
    event_url = (
        f'https://webapi.legistar.com/v1/mountainview/Events/{event_id}')
    req = urllib.request.urlopen(event_url)
    result = json.load(req)
    for item in item_data:
        if item['EventItemEventId'] == event_id:
            for k in select:
                item[k] = result.get(k)


# everything is all together now
json.dump(item_data, sys.stdout)

And that is pretty much it. Feel free to follow along at home. This should work on any jurisdiction that uses legistar. A few I know of are San Jose, Sunnyvale, Santa Clara and BART.

The jobs:housing imbalance is at the confluence of the climate crisis and the housing crisis.

The simplified version:

1. Some area has lots of jobs but not enough housing
2. The price of housing in that area goes up
3. People move away where the housing is more affordable and commute
4. Secondary effects of driving (pollution, greenhouse gasses, collisions, etc) all increase.

The bay area has been in a housing crisis for decades. As the price of housing continues to increase people are commuting from further and further away.

What Makes an Imbalance?

The Jobs:Housing ratio is simple to calculate. (number of jobs in an area) / (number of homes in the same area). This ratio is trying to answer the question: would it be possible for everyone who works in this area to also live in this area?

This is a simplification. It does not take into account household size, housing type, housing preference, retired population, etc. A more complex measure that does take these factors into account is the jobs-housing fit

Since the ratio is inexact, we cannot say with confidence what it ought to be exactly. I will be using the range of 0.75-1.5, as according to the epa this is know to reduce VMT (vehicle miles traveled).

I will use the most recent jobs data from the US Census. This is from 2021. But keep in mind that lots of existing office capacity in the bay area was sitting empty in 2021. So this analysis very likely underestimates the actual housing need.

Why Does the Imbalance Exist?

Very briefly, the proximate cause is Euclidean Zoning and downzoning Euclidean Zoning is an interpretation of zoning that originated in the USA and is characterized by large areas zoned exclusively for very specific uses, which keeps everything separate and creates sprawl. Euclidean Zoning started in the 1920s, and by the 1960 had been adopted by nearly every city in the country.

In the 60s and 70s there was a wave of downzoning, where cities throughout the country further restricted zoning by disallowing higher density housing in many areas where it had been previously allowed. This prevented the normal, gradual densificaiton of cities, and pushed growth out to the edges.

Both these changes were motivated by racism

Analyzing Santa Clara County Cities

Ratio

What is the ratio in Santa Clara County cities? Surprisingly, most municipalities are in balance, when we use the 0.75-1.5 range.

By this metric six cities, and unincorporated Santa Clara county, all have a clear housing need. Three cities are imbalanced in the opposite direction. And the county as a whole has a balance of 1.47. Within the tolerance we defined.

Most of unincorporated Santa Clara County is empty. It is Stanford, with over 30,000 jobs, that creates the imbalance. Adding more housing in East San Jose will not fix the imbalance in Stanford. So I will pull out Stanford as if it were it’s own municipality for the rest of the analysis.

According to this chart Los Altos is in balance. But the city is very close to job centers in Mountain View, Palo Alto and Stanford. So building more housing in Los Altos would likely lower VMT for the region as a whole.

Density

Another way to look at it is density, or homes per acre. This might give a better sense of the amount of change needed.

So while Mountain View and Cupertio have similar ratios, Mountain View’s housing need is much higher. And while both Sunnyvale and Mountain View need to build around 30,000 homes to get the ratio below 1.5, it is a much larger change in density in Mountain View. This is because Sunnyvale already has much more housing than Mountain View.

Regional Housing Needs Assessment

The California HCD (Housing and Community Development) department tries to force cities to build housing. The process is called RHNA (Regional Housing Needs Assessment). The process is indirect and has so far been ineffective. But recent legislative changes have made it more difficult for municipalities to ignore state law.

RHNA assigns each municipality an allotment of housing that should be built. But the assessment is not focused on the jobs:housing ratio.

Clearly RNHA has other goals.

Another reason RHNA cannot address the jobs:housing imbalance is that it does not restrict cities adding jobs. Plans like North Bayshore in Mountain View or Moffett Park in Sunnyvale are a big part of how these cities plan to meet their housing goals. But those projects also add new office. So even if they are built the imbalance will remain.

A Regional Map of the Imbalance

HCD and state law operates at the municipal level out of necessity. But a map which only shows the municipal stats would not be of much use. San Jose is “in balance” in the same way the county as a whole is “in balance”.

San Jose is not in balance, because the city is spread out over a huge area, but most of the jobs are concentrated in the center. So just like Mountain View and Palo Alto need to build housing near their job centers, so San Jose should build housing near downtown.

So I will do a similar analysis as above, but without considering municipal boundaries. A more granular approach is needed.

I have been using data from the US Census LEHD Origin-Destination Employment Statistics. 2021 is the most recent data. The data on employee count has a granularity of census block, which can be pretty small.

We can take the centroid of each block, and sum the number of homes and jobs within a certain radius of the center. For large blocks (mostly rural) this will only include the block itself, but for smaller blocks (mostly urban) we can get a pretty good average of the area.

Incidentally the largest spike is because of this one block.

To understand the imbalance, I decided to use a 1 mile radius. For each census block I calculated the jobs:housing ratio for the circle with 1 mile radius centered on the center of the block.

The resulting map shows clearly where the imbalances are.

The imbalance is concentrated in north county. But even Morgan Hill and Gilroy have areas were housing is needed.

There is no area with a spike in housing supply that comes close to the spike in jobs. Several areas would need over 80,000 homes to bring into balance.

Conclusion

The built environment in Santa Clara County guarantees large number of commuters. It is not possible for everyone to live close to work, because the necessary housing does not exist. The existing process is not sufficient to build this housing.

To reduce commuting, over 100,000 housing units are needed. The existing housing element and RHNA framework will not build these units. Something different is needed.