Let’s recap. SB 828 would empower the California Department of Housing and Community Development (HCD) and COGs to make significant changes to the Regional Housing Needs Allocations (or Assessments) for which local governments must identify adequate sites for new housing development. The RHNA is explained in this tweet thread. In this context, SB 828 states that HCD must:
“(b) Establish a methodology for the comprehensive assessment on unmet need that acknowledges the following:(1) Median rent or home prices that exceed median income will be alleviated by rapidly increasing housing supply, particularly housing supply for moderate and above-moderate income households.(2) Communities with high rates of income growth must also have a high rate of new housing production for households of all income levels to ensure equity and stabilize home prices and communities.”
Part 1 of this analysis concerned (b)(1), and how how “high cost tests” comparing the use of median incomes against median rents and home values would subject nearly all California jurisdictions to large increases in allocations. This posts considers (b)(2). Let’s take a deep dive with some initial thoughts and questions:
- What is a “high” rate of income growth?
- What will be defined as ‘income’?
- Will median or mean income be used?
- Income inequality will, in some contexts, show up more obviously in a statistical mean versus a statistical median. This could mean that using a mean based definition is a good thing… e.g. if you believe that an increase in high-income earners has a disproportionate impact on housing costs versus an equally sized increase in middle or low income earners.
There are many ways this could go. For this analysis I’ll focus on the SF Bay Area for now just because its easier.
Modelling the Question:
To rephrase the language of SB 828 into our research question: Who would be subject to “a high rate of new housing production at all income levels to ensure equity and stabilize home prices and communities” because they are “communities with high rates of income growth” ?
Some Definitions of High Income Growth to test:
- A city has high rates of income growth if its median income grows faster than the regional median income on a dollar basis (as opposed to a percentage basis).
- A city has high rates of income growth if its mean income grows faster than the regional mean income on a dollar basis (as opposed to a percentage basis).
In this case, the region is the Combined Statistical Area (SF-SJ-OAK CSA). The base year is 2011 (from the 2007-11 ACS Waves) and second time point is 2016 (from the 2012-2016 ACS Waves). So who passes the (b)(2) high cost test under these definitions, thus “requiring them to ensure equity and stabilize home prices and communities” ala Senator Wiener’s bill? Maps below:
So as expected, the mean based definition catches more localities as “high rate of income growth” than the median one. Some details:
- Cities that pass the test under both definitions include:
- A Marin Cohort: Corte Madera,Mill Valley, Santa Venetia,
- A Silicon Cohort: Burlingame, Portola Valley, Los Altos Hills, Los Altos, Coupertino, Campbell, Los Gatos (Yesss).
- An Oakland Hills Cohort: Orinda, Alamo, Alhambra Valley & San Miguel.
- Each Cohort adds some folks under just the mean-based definition:
- Marin Cohort pickups: Sausalito, Tiburon, Belvedere, Stinson Beach,
- Silicon Cohort pickups: Sunnyvale, Mountain View, Saratoga, Ladera, Atherton, Half Moon Bay,
- Oakland Hills pickups: Danville, Kingston, North Gate.
So here we see the role that income inequality and economic segregation play in impacting the meanings of our statistics, and why something like selecting the median or the mean would matter in implementation of language like this.
I thought it would be fun to test a definition of high wage growth against inflation (e.g. a scenario where cities would pass the high cost test if median and/or mean incomes grew faster than the rate of inflation). I tried it out and, by those metrics, nearly all jurisdictions in the Bay Area are experiencing high rates of income growth.
Measuring this In Terms of Equity: Comparing Percentiles Within Cities
It might be better to have the requirement test if each city’s 75th percentile or 90th percentile income is rising faster than the median (or 40th percentile) income for that same jurisdiction. After all, the bill states that this second part (b)(2) is about “requiring them [localities] to ensure equity and stabilize home prices and communities.” (emphasis added). lets take out “home prices and” and the sentence reads: “requiring them [localities] to ensure equity and stabilize communities.”
I like that. Under the percentiles comparison approach, cities will be required to zone for housing at all income levels if their higher income earners have incomes growing faster than their lower income earners. Just a thought.
Code and data to make all this stuff is here. If you check it out, let me know if you find an error.