Density Penalties for Affordable Housing Proximity to S.F. Homes

For an upcoming lecture I’m talking about exclusionary zoning.  As always, I am surrounded by ample examples of whatever I am teaching here in California.

Davis is about to approve Creekside, a small affordable housing complex for seniors.  Roughly 40% of the 70 units will be set aside for formerly homeless individuals and those with special needs. Amazingly, this site is also across the street from a typical California tract, ticky-tack suburban subdivision (one I’ve rented in recently).


Highlighted site across the street from single family tract homes (the red roofs)

This is really remarkable.  Because I’m still in the area, I’m privy to all the ugly vitriol against the proposal on  Much of whats in the proposal for the site is designed to minimize the ‘asthetic impact’ of the site.  This is reflected in the zoning.  The figure below, lifted from a report by city staff to the Davis Planning Commission, compares the previous zoning (left) to the proposal of the site.

What it illustrates to me is that zoned densities do not matter, even in towns where local NIMBYs treat the General Plan like some kind of holy document.  I want to advance the hypothesis that developers take density penalties–voluntary reductions in density below zoned limits–to assuage NIMBYISM.  Presumably, the ‘Density Penalty’ is correlated with proximity to owner occupied housing across space.

Here’s the density penalty for Creekside:


The General Plan density for the site (left) and the project’s attributes (right)

It reminds me of things I’ve seen repeatedly in debates over projects, both market rate and affordable.  Developers appear to sacrifice height and floor space (via expanding front setbacks).  If the site abuts a highway, then developers will absorb all the rear setback and build a sound wall.  The goal appears to be increasing front setbacks–minimizing the ‘aesthetic pain’ of drivers-by.

In a different Davis planning commission meeting, a commissioner strongly criticized a proposed project for its aesthetic damage, saying, among other things, “the people who have to drive by this monstrosity deserve better” and comparing it to Soviet Bloc housing.  See it below:


“Monstrosity… Soviet-style housing… We deserve better! The drivers who pass by deserve better!” –How an insane Davis Planning Commissioner described  this image. 

In both cases, developers emphasized reduced heights and increased setbacks.  To win anything near single family residences, we have to mask density… we have to pay that premium.  Actual densities do not matter.  Populations served often do not appear to matter… maintaining the suburban illusion is apparently something we all must pay for through reductions away from highest and best use!

What the UC & NIMBYs Have In Common

note: this is more of a late night venting than an intentional blog. 

Episodically, during my 10 years as a UC brat, I’ve heard it basically said that “students graduate in four, six years… but the faculty & administration will spend the rest of their lives here… so we must prioritize labs, monuments & offices… but not the demands of students… like on campus affordable housing.”  It has a certain logic to it.
Off Campus, at least here in Davis, you will hear actually hear landlords say–quite bluntly– “students only live here for 4 years and then move on. Home owners spend the rest of their lives here. So we should only listen to home owners.”  It also has a certain logic to it.

Two sides of the same coin.

Here’s a comment by a Davis landlord on the competing views of students and home owners over a development that recently passed:


Implicit in this statement is the idea that the value of your opinion in “citizen planning” (something Davis pats itself on the back for having) is proportional to how long you will be staying in said community.  We have a vacancy  crisis (0.3% vacancy rate).  We have students sleeping in cars.  And apparently this should be discounted because this is a “short term” need… (it’s not, UC Davis will be around another few years probably, but the individuals in crisis will only be our problem for a few years so their needs should be ignored, or so the commenter thinks…)

The logic unravels when you find out most of the NIMBY homeowners in Davis are only in Davis because they were UC Davis students or staff at one point.  But let’s look beyond college town dynamics…

Here are my questions for someone defending this kind of thinking anywhere in California (or anywhere period, for that matter):

  1. Does this mean we should value a farm-worker’s voice at 1/2 a home owner’s, since farm-workers only live here seasonally?  Same logic, applied to a different population…
  2. Since out of state students may plan to move back home, should California ban them from voting in our state? Same logic, same population…
  3. The chronically homeless are impermanent everywhere–should they be ignored everywhere? Same logic, overlapping populations…

And then I realize that this landlord gets what she wants most of the time in terms of power dynamic… and she’s complaining because this time the home owners didn’t get everything the wanted (admittedly, the home owners most impacted  for this particular projects are lower income themselves… which also tells you a lot about Davis!)  And come to think of it, its a retirement community.  So this landlord gets what she wants in a sense:  If “the rest of your life” in the community is the weight of the value of your voice, then of course the seniors get stuck with the “undesirables!”  The grotesqueness of it all comes full circle. I’m exhausted now.

Thanks for listening.  This was a rant. Not a thoughtful blog.  But I had to put this on paper because the logic people use to rationalize their exclusionary politics is quite frustrating.


3 Crazy Housing/Planning Ideas

Prepping for a move to Australia, doing side jobs to pay for said moves, and doing other things… haven’t been able to blog.  But want to throw some out-there ideas to stimulate thinking:

  1.  San Diego, San Francisco, Los Angeles.  What do they all have in common? Lots of homeless folks and lots of harbour. So… Let’s re-purpose old cruise ships into homeless shelters.  That’s got to be cheaper on a per-bed basis, and the lifespan may not be significantly shorter if the ships stay in the harbours.

You see a ship. I see housing!

2.  If California is going to pour billions of dollars into our transit systems, then we need to ensure they will work.  The state should take direct control of land use planning within 3/4ths of a mile of each transit station of funded systems.  Our transit and affordable housing dollars will go further if they are all used in concert by one actor. “coordination” isn’t working.


Use it Or Lose It

3.  Let’s found a new Bay Area growth pole… a few square miles (8-10) near an edge BART station or station under construction (hi South/East Bay!)…. with some minimum density requirements (no maximums!) and see what happens.



The McKinsey Metric Applied to Cali Cities: It’s Not Pretty

The McKinsey Global Institute produced a fantastic study on how to solve California’s housing crisis.  Embedded in the study is this fantastic and illustrative figure:


I decided to explore what it would look like to apply the final metric on the far right hand side to our cities in California.  It’s the ratio of new housing units relative to new population, per 1,000 people.  Let’s call this “the McKinsey ratio” for now.

Because American Community Survey estimates are unreliable for 2005 at a local scale, I get population data for the two years (2005 & 2014) from the Department of Finance’s Population Estimates tables for jurisdictions, and new housing data from Census permitting records.  They’re not perfect, but they work for this exercise.

First, I want to show you the distribution of “the McKinsey ratio” for the 157 jurisdictions that had adequate information and a population over 50,000 in 2005, with some of the state estimates in the figure above drawn over in dashed lines.  Communities of less than 50,000 are left out due to extreme results.


The first thing you notice is that the vast majority of California cities are on the smaller side of the state’s average as estimated by McKinsey.  A small but not insignificant amount of this is due to the data differences, I’m sure.  But regardless, just 24 of 157 jurisdictions score above the statewide average.  The first thing a Geographer notices is this variable might not be appropriate for jurisdictions, which can very dramatically in size (there are some with ratios above 1000 and below -1000 cut off for visual reasons).

But look at where other state’s stand.  Some folks are really pumping up Texas as the “Yes” state.  It’s ratio is actually somewhat close to California’s according to McKinsey.

But look at New York.  There are 14 sizable jurisdictions in our state producing as much as the New York average or more, and they are the first 14 ranked in the table below.  With only a couple of exceptions, they are all in southern California.  The table ranks jurisdictions in my dataset with over 50,000 residents by their “McKinsey Ratio” score.  Note that in areas that lost population, the ratios don’t really make sense.  But they are worth including to know where they are.  See anything interesting?

McKinsey Ratio Rank Jurisdiction McKinsey Ratio DOF Estimated Pop 2005
1 South Gate 2731 97,461
2 Rosemead 2570 54,677
3 Paramount 1150 55,606
4 Vacaville 906 93,954
5 Pasadena 764 137,501
6 Ontario 730 164,504
7 Arcadia 730 55,521
8 Pomona 696 152,106
9 Burbank 671 103,122
10 Gardena 642 59,277
11 Long Beach 622 470,781
12 Santa Ana 621 332,878
13 Torrance 605 143,738
14 Chino Hills 591 75,414
15 Fountain Valley 484 55,193
16 Encinitas 456 59,929
17 Lodi 439 61,431
18 Hawthorne 436 85,030
19 Downey 434 111,416
20 Lynwood 389 70,733
21 Newport Beach 331 81,678
22 Santa Monica 321 88,692
23 Los Angeles 317 3,769,131
24 Santa Barbara 312 88,854
25 Upland 308 72,216
26 Oceanside 289 166,958
27 Costa Mesa 289 109,030
28 Redding 281 87,152
29 Glendora 281 50,490
30 Garden Grove 276 168,219
31 Whittier 270 85,433
32 Laguna Niguel 257 63,310
33 Colton 257 51,522
34 San Buenaventura 245 103,374
35 La Habra 241 59,828
36 Orange 226 133,542
37 Vista 223 92,110
38 Anaheim 222 331,458
39 Thousand Oaks 219 124,169
40 Westminster 219 89,268
41 Highland 210 50,901
42 Walnut Creek 208 64,705
43 Inglewood 202 112,417
44 Redondo Beach 198 65,931
45 Oakland 198 389,937
46 Davis 196 63,889
47 San Diego 195 1,261,035
48 Buena Park 194 78,619
49 Modesto 190 201,980
50 Concord 189 122,373
51 Escondido 172 139,585
52 Fullerton 171 132,913
53 Oxnard 169 185,994
54 Mountain View 169 70,629
55 San Bernardino 169 201,295
56 San Clemente 167 62,286
57 San Leandro 162 81,802
58 El Cajon 161 97,364
59 San Francisco 159 780,187
60 Rialto 158 98,224
61 National City 155 55,948
62 Napa 154 74,499
63 Sacramento 153 442,662
64 Simi Valley 152 118,961
65 Livermore 147 78,019
66 Salinas 138 147,387
67 Petaluma 134 55,973
68 Pleasanton 132 66,890
69 Yorba Linda 131 62,574
70 Richmond 127 101,098
71 Woodland 124 52,474
72 Cupertino 124 53,632
73 Compton 123 96,133
74 Lake Forest 121 76,635
75 Stockton 120 277,485
76 Palo Alto 119 60,723
77 Daly City 119 100,379
78 Corona 116 144,719
79 Turlock 115 65,301
80 Santa Clara 112 107,058
81 San Jose 111 901,159
82 Union City 110 67,544
83 Santa Rosa 109 157,175
84 La Mesa 107 55,354
85 San Mateo 106 93,396
86 Rancho Cucamonga 105 156,854
87 Milpitas 105 62,177
88 South San Francisco 105 60,172
89 Riverside 104 284,715
90 Chino 99 74,463
91 Fremont 99 206,712
92 Fresno 96 457,786
93 Berkeley 95 105,880
94 Redwood City 95 74,621
95 Santee 95 52,110
96 Fairfield 89 102,553
97 Tustin 89 70,116
98 Folsom 85 66,362
99 Pittsburg 84 61,120
100 San Rafael 83 56,247
101 Camarillo 83 61,515
102 Santa Cruz 79 56,387
103 Chula Vista 73 219,939
104 Hayward 72 140,530
105 Merced 70 72,402
106 Santa Maria 69 91,313
107 Alameda 68 71,727
108 Sunnyvale 66 131,853
109 Lancaster 64 137,268
110 Carlsbad 57 94,161
111 Antioch 57 99,713
112 Yuba City 56 57,975
113 Manteca 56 60,598
114 Clovis 55 84,552
115 Tracy 54 78,228
116 San Marcos 51 72,564
117 Hemet 51 68,943
118 Roseville 49 104,105
119 Palmdale 49 135,179
120 Chico 49 72,459
121 Visalia 48 106,054
122 Rocklin 46 51,206
123 Madera 46 51,735
124 Bakersfield 44 299,363
125 Moreno Valley 44 167,262
126 Hesperia 36 76,548
127 Santa Clarita 35 165,431
128 Irvine 35 179,975
129 Temecula 35 78,808
130 Murrieta 32 85,769
131 Elk Grove 32 125,703
132 Rancho Cordova 31 55,476
133 Fontana 31 161,728
134 Victorville 23 87,813
135 Perris 23 50,650
136 Indio 21 62,024
137 San Ramon 10 53,923
138 Cerritos -62 51,674
139 Lakewood -85 81,040
140 Huntington Park -132 61,318
141 Carson -178 94,236
142 El Monte -210 118,295
143 Baldwin Park -223 77,383
144 Norwalk -282 106,921
145 Mission Viejo -518 95,427
146 Alhambra -548 86,541
147 West Covina -642 107,955
148 Pico Rivera -652 64,635
149 Vallejo -757 117,993
150 Monterey Park -884 61,647
151 Diamond Bar -925 56,703
152 Redlands -1030 68,471
153 Huntington Beach -1430 192,581
154 Montebello -2669 63,359
155 Citrus Heights -2890 85,153
156 Bellflower -3802 76,306
157 Glendale -5403 197,042



Unpacking The Real Impact of By-Right

I’m hearing a lot of things from advocates on all sides of housing issues in California, with folks operating from sets of assumptions that are contradictory in nature.  I’d like to unpack them for a bit.

First, two key points are emerging from the YIMBY narrative:

  • Restrictive zoning makes a lot of naturally affordable housing impossible to build.
  • SB 35 (Wiener-D San Francisco) will make housing great again by imposing by-right on cities that fail to meet their RHNA.

I’m not convinced both of these statements can be true at the same time to the extent they are being sold by supply-siders.  SB 35 brings “by-right” to any city that isn’t meeting its Regional Housing Needs Allocation, yes.  But a project can only be by-right if it is within existing zoning for a parcel (sans the density bonus, of course).  If the YIMBY narrative about restrictive zoning is true (it is, see figure below), then it’s probably very easy to overstate the impact of SB 35.


Zoning Capacity Changes in Los Angeles (Morrow 2013, pg. 3)

In theory, by-right will help the red line in the figure above intersect with the black (for every jurisdiction, not just L.A.). By-Right will help reduce the ability of local community members to block new housing at existing zoning capacities + density bonus allowance.  That’s mostly it. It will not increase zoned capacity.  And I have my doubts about that really being enough to put a dent in the supposed need. I say supposed because 85% of any new housing produced through by-right will be market rate, and the state is producing a significant, albeit insufficient, amount of market rate housing.  Furthermore, the public draft of the Statewide Housing Assessment suggests this market is impacting low and very low income residents the hardest, with most above moderate income households doing ok.

But let me use the 2015 ACS one year wave for California to explain this to you in a different way.


California Households’ Area Median Income (AMI) Designation, Sorted By The AMI Designation Of Their Rental Units. Data: 2015 ACS PUMS 1-Year Wave

There simply aren’t many above moderate income households paying too much for their housing.  In contrast, quite a few low income households pay too much when they rent at these above moderate levels.  All the households colored a shade of blue in the two furthest right columns are paying at or near market rate rents they simply cannot afford.  This encompasses nearly a million households, and doesn’t include the almost one million households who manage to pay less than this yet still cannot afford their rents.

In contrast, the folks in green and yellow in the middle column (Low Income units) have it pretty good.  What makes anybody think that if new Above Moderate units are built, those folks will leave their great situation and pay higher rents (shifting from the middle to furthest right column, paying more rent for what is presumably a higher quality unit)?

Ask yourself: what single person in San Francisco making $7,000 a month will “trade up” from paying $1,500 for a room in an old Victorian in a neighborhood they love, to paying $3,000 in a 20th floor studio where it takes 10 minutes to get to street level and there’s nothing fun nearby because it’s in the FiDi (the CBD)?

I really like the idea of holding cities accountable for their failure to build, but I think we are overstating the impact this legislation is going to have on the affordability aspect of the vaguely termed housing crisis.  This is why, at the very least, SB 35 needs to be paired with a permanent source of funding for programs that produce housing for those low and very low income families the market cannot help.

ACS Based FMRs Do Not Work Anymore

Here’s the 101:

FMRs: maximum payouts for Section 8 housing vouchers, set at a regional scale.

ACS: The Census data utilized to set these FMRs (generally on a 3 year lag, and the ACS itself is a compilation of five years of census gathering efforts).

Since some of the data used to set these thresholds is literally 8 years older than the year in which the thresholds are applied, many believe we need a new system for calculating FMRs.

If you’re not convinced… here’s all the evidence you need, from an enormous set of for-rent scrapings on San Francisco proper:


When Boeing and Waddell said there was only 1% affordability in the Craiglist’s postings in Manhattan (and similar places), I thought there must be something off about Craigslist.  When Deb Niemeier and I  looked at voucher access in 2012-13 it wasn’t that bad.  But the proprietary data for 2015 shows the exact same thing folks at Berkeley found via Craiglist:  Catastrophic loss.

-> and here I have to note that a small but meaningful segment of voucher affordable sites are probably never listed online or in newspapers, so there is bias yet in our ‘big data’ tools as well.

But the point remains: maybe we shouldn’t set policy using lagged data… (that itself is also a five year wave and is thus also lagged….!!!!!)


Someone Check The ACS Please

Image result for american community survey data

The 2011-2015 5 year “wave” estimates for the American Community Survey are out.  These are exciting because we now have enough “waves” of the ACS to look at changes in Census tracts over time without having to utilize those painful crosswalks…

My hope is that someone will take a look at the 2006-2010 ACS estimates for tracts against the 2011-2015 estimates and let us know a few things:

  1.  Are the samples robust enough to give us significant differences where we know there are significant differences?
  2. What can they tell us about urban change, if anything?
  3. How has demographic change itself impacted the stability of the estimates in certain places?

Many people use 5 year wave data like the 2011-2015 dataset and, in their manuscripts, refer to it only as the “2015 ACS.”  This is ok if you have acknowledged towards the beginning of the paper that you are referencing what is a five year rolling sample… but many practitioners, media folks and the public at large sees “2015 ACS” and thinks that is what you are talking about.

Failing to appreciate this difference means ignoring what the data is really telling you…

California’s Legislative Analysts’ Office (LAO) twisted the ACS data used by Karen Chapple and Miriam Zuk at Berkeley to try to say, essentially, the opposite of what Zuk and Chapple argued in their own work… to the point where they had to respond.

The biggest critique I have in all of this is much more basic: what does a median rent in a Census Tract really tell you?  When you tell me that adding affordable housing to a census tract over a period of time reduces its median rent change in that period, all you are telling me is that adding numbers to the bottom of a distribution of numbers shifts the distribution’s median downwards.  How much can we really say with the blunt instrument that is the ACS 5 year tract estimates?

Last point on that Berkeley-LAO debate: they relied on the 2009-2013 5 year wave data, but kept constantly referring to their findings as measuring change from the 2000 Census and 2013 ACS.  The 2013 wave includes 2009, when we were in a recession. It also includes 2013, which was undoubtedly a ‘boom’ year for housing in the Bay Area…  I’m wondering if someone ran the Berkeley analysis with the latest ACS wave, what would they find? It’s the first wave we’ve got tract-scale estimates for that is clear of the recession.  Someone please dig in!