DI
SCLAIMER: This memo provides a technical overview of the ADU Affordability Survey and its potential implications for affordability designations in Annual Progress Reports (APRs). It does not constitute legal advice.
This memo was published in January 2026.
ADU Affordability Survey Findings and Recommendations
Executive Summary
In partnership with Bay Area jurisdictions, the Association of Bay Area Governments (ABAG) has been administering an affordability survey to learn more about the rents of newly built ADUs. In 2024 and 2025, participating jurisdictions distributed the survey as part of the permitting process. This data is complimented by similar studies in Oakland and San José.
Based on the results of the surveys, the California Department of Housing and Community Development (HCD) is comfortable with ABAG jurisdictions continuing to use the following assumptions in their Annual Progress Reports (APRs) for two more years (APR 2025 and 2026):
• 30% very low income
• 30% low income
• 30% moderate income
• 10% above moderate income
Please read this memo for more information.
History
In 2021, ABAG, in partnership with UC Berkeley, conducted an ADU affordability study to inform affordability assumptions for Housing Element APR reporting. HCD endorsed the findings of this initial study, and most jurisdictions have since used the 30/30/30/10 affordability split in their Housing Elements and APRs.
Building on the 2021 study, ABAG’s Regional Housing Technical Assistance (RHTA) program launched a regional, opt-in ADU affordability survey to collect information from homeowners during the ADU approval process. The survey supports jurisdictions in documenting the affordability of newly constructed ADUs and in Attachment 2
meeting Housing Element monitoring requirements. Developed in consultation with an advisory group of jurisdiction staff and reviewed by HCD, the survey was initially launched in 2024 and revised in 2025. The survey remains active, with additional analysis anticipated as more responses are collected. 2024-2025 Survey and Data Overview
To date, at least 39 jurisdictions have implemented the survey. Some require homeowners to fill it out during the ADU building permit process, while others encourage them to do so.
The survey takes approximately five minutes to complete and asks general questions about the ADU the homeowner intends to build such as type (e.g. detached), square footage, bedroom count, expected inhabitants (e.g. renters, family, friends) and anticipated rent. To encourage accurate responses and protect privacy, the survey is anonymous. Because rent information can be sensitive, the survey collects anticipated rent using ranges rather than exact amounts.
Between July 2024 and September 2025, the survey received 436 responses. To supplement these, two additional datasets were incorporated:
• City of Oakland Survey: 142 responses, (2025) focused on existing ADUs. • City of San José Survey: 79 responses (2024 and 2025), based on the ABAG survey with minor modifications.
This resulted in a combined dataset of 657 total responses, representing 39 jurisdictions in the ABAG region.
Alameda County had the largest share (31%), followed by Santa Clara County (23%) and Contra Costa County (17%).
County Percent of Total
Alameda (including the Oakland survey) 31%
Contra Costa 17%
Marin 3%
Napa 1%
San Francisco 5%
San Mateo 14%
Santa Clara (including the San José survey) 23%
Solano 2%
Sonoma 4%
Attachment 2
2024 and 2025 Initial Results
ABAG discussed survey findings with HCD in December 2025. Based on these discussions and additional analysis, HCD endorsed the continued use of the 30/30/30/10 affordability split for typical ABAG jurisdictions for APR reporting for two more years (APR 2025 and 2026). Jurisdictions may still submit alternative data to HCD if they believe it supports a different affordability distribution. As in prior years, while the 30/30/30/10 assumptions are appropriate for most jurisdictions, HCD may request additional information in select circumstances.
Affordability analysis of survey data collected between July 2024 and September 2025, summarized below, supports the continued use of the 30/30/30/10 affordability assumptions.
Affordability Level Affordability Based on Survey
Responses Recommendation
Extremely Low Income and
Very Low Income 57% 30%
Low Income 30% 30%
Moderate Income 11% 30%
Above Moderate Income 1% 10%
Al
though the survey results suggest deeper affordability than assumed under the 30/30/30/10 framework, ABAG recommends maintaining the current split for the following reasons:
• Less than two full years of survey data are available. • The dataset relies heavily on proposed (rather than occupied) ADUs, which may involve less precise rent estimates.
• While meaningful, the overall sample size remains relatively modest.
Continued jurisdiction participation and active promotion of the ADU affordability survey will strengthen the dataset over time and support more refined affordability assumptions in future APR cycles.
Contact Hannah Diaz, [email protected] with questions or comments about the survey.
Attachment 2
Appendix
Data Filtration
Since the goal of the analysis is to estimate the affordability levels of proposed ADUs (and existing ADUs in the case of the Oakland data) several response types were removed to avoid skewing results:
• Responses indicating the ADU would not be used for housing; • Responses from the original ABAG Survey with “0” listed as expected rent; • Responses selecting “decline to state” for rent amount on the updated ABAG survey; and • Responses designated as “rent not shared” on the Oakland Survey.
After applying these filters, 460 responses remained with adequate rent data for affordability analysis.
Affordability Methodology
The following describes the data standardization and methodology applied to determine the affordability level of the proposed (or existing) ADU. The methodology considers the year the ADU was proposed, the county it is permitted in, and the bedroom count.
Calculating Affordability
In order to determine the affordability of the ADU, a formula was developed based on HCD’s State Income Limits. The formula is based on the year-specific county median income and then adjusted for both occupancy and income category.
The following household size assumptions were used depending on the bedroom count of the ADU:
• Studio: one-person household, 70% of county median income; • One Bedroom: two-person household, 80% of the county median income; • Two Bedroom: three-person household, 90% of the county median income; and • Three Bedroom: four-person household, county median income.
Attachment 2
Once the county median incomes were adjusted for household size, then thresholds for income levels were established as follows:
• Extremely Low Income: Up to 30% of the adjusted county-specific median income; • Very Low Income: between 31% and 50% of the adjusted county-specific median income; • Low Income: between 51% and 80% of adjusted county-specific median income; and • Moderate Income: between 81% and 120% of adjusted county-specific median income.
Rent Thresholds
From the adjusted incomes, a maximum monthly rent threshold and range was established using the assumption that occupants should only spend up to 30% of their income on housing.
Example: The 2024 median income in Alameda County was $155,700. To calculate the maximum rent for a one bedroom, extremely-low-income ADU household, first adjust the median income for expected occupancy (two persons, therefore 80% of the median Income) which is $124,560. Then adjust for the extremely-low- income category to calculate the maximum yearly income to qualify as extremely low income (ELI is up to 30% of the median Income) which is $37,368. Lastly, calculate the maximum monthly rent (rent maximum is 30% of annual income) which is $934 a month. ($37,368 x 0.3 / 12 months) Categorization
The survey responses were categorized first by calendar year since the income limits are updated yearly, and then by location, since each county has its own calculated median income.
Since the second version of the survey asks respondents to choose a range for their intended rent, the analysis uses an even distribution to assign affordability which preserves proportionality.
Example: If low income maximum rent is $15 per month, and 10 people choose the box "between $10 and $20 a month" for rent, five units would be considered low-income units and five units would be moderate income units.
Attachment 2
Updates to the Survey Questions to Improve Rent Data Accuracy An issue identified from the first version of the survey (open from July 2024 to February 2025) was that the rent question was open-ended, asking respondents to write the amount that they intended to charge for monthly rent of the ADU. The question explicitly stated that writing a “0” meant that homeowner did not intend to charge rent.
The findings from the first survey showed that a high percentage of respondents were giving an answer of “0” when asked about the expected rent of their proposed ADU, which led to the concern that respondents were potentially not providing an accurate rent estimate or intended to charge zero rent for their units. This difference has important implications for calculating affordability, since a unit that has zero rent would be considered “extremely low income.”
To address this concern, the rent question was updated in the second version of the survey (open from February 2025 onward) to a multiple-choice response which included a “decline to state” option, a “not planning to charge rent” option, and then rent ranges (for example $1,500 – $2,000).
Attachment 2