Research synthesis
Who lives in ADUs? The evidence, and its expiry date
A synthesis of the Center for Community Innovation’s statewide owner survey, cited as theirs, joined to where California is permitting units now. The join is ours. The occupancy findings are not.
Written by HyreADU Research Desk Primary-source research and permit-data analysis
Audited by HyreADU Research Desk Attribution audit and dataset provenance
The finding
About half of California’s new ADUs are rented for income, about a sixth house a relative rent-free, about a sixth are the owner’s own office, studio or guest space, and 8 per cent are short-term rentals. That answer is not ours. It belongs to Karen Chapple, Dori Ganetsos and Emmanuel Lopez of the UC Berkeley Center for Community Innovation, in Implementing the Backyard Revolution: Perspectives of California’s ADU Owners (22 April 2021).
Their executive summary reports that “About half (51%) of California’s new ADUs serve as income-generating rental units, and 16% of ADUs provide no-cost housing to a relative of the homeowner,” and that “Only 8% of new ADUs in California are short-term rentals.” The survey was fielded in 2020 among owners of units permitted or completed in 2018 and 2019, and the full breakdown is set out below in CCI’s own terms.
HyreADU contribution: one join, and nothing more. That survey frame is now the smaller and geographically different half of California’s ADU stock.
From HCD’s Annual Progress Report Table A2, statewide ADU permits ran 25,042 units across 2018–2019 against 63,641 across 2024–2025, and within the ten jurisdictions we track the City of Los Angeles fell from 75.0% to 55.6% of the group’s permits while San Diego rose from 5.2% to 16.8%.
Just over half of CCI’s responses came from the Bay Area, so the split everyone quotes was measured in a production geography that has since shifted south and inland.
That does not overturn the split; it states the size of the extrapolation being made whenever a 2018–19 cohort is described as “who lives in ADUs” in 2026. What nobody can tell you. There is no demographic profile of ADU occupants in any public dataset.
The Census Bureau’s housing tables cannot see an ADU as a category, so age, income, household size and relationship for ADU residents are not derivable, and none are published here.
Read this first
Occupancy is the lowest-originality subject in our research program, and this page is published as a synthesis rather than as a study. Everything below is either somebody else’s finding, correctly attributed, or one arithmetic join of our own that is labeled where it appears.
- The occupancy split is not ours and we do not own it
Every figure describing what ADUs are used for on this page comes from the Center for Community Innovation at UC Berkeley, in work reviewed by the Terner Center for Housing Innovation.
Where you see 51 per cent, 16 per cent, 8 per cent or the Figure 9 breakdown, that is Chapple, Ganetsos and Lopez (2021). If you cite those numbers, cite them; a citation to HyreADU would be wrong.
- There is no demographic profile of ADU occupants, anywhere
No public dataset identifies an accessory dwelling unit as a housing category, so nothing about the age, income, race, household size or family relationship of the people living in ADUs can be computed from Census or American Community Survey tables.
A detached ADU is a single-unit detached structure in those tables, indistinguishable from the main house; an internal ADU makes the property a two-unit structure, indistinguishable from a duplex. We publish no occupant demographics because there are none to publish.
- The survey is a 2020 survey of a 2018–19 cohort, and the cohort was Bay Area-heavy
CCI mailed postcards to Annual Progress Report addresses for 2018 or 2019 permits and certificates of occupancy, with a Bay Area supplement, and reported that just over half of responses came from the Bay Area.
Since then California has permitted more than two and a half times as many ADUs, and has permitted them in different places. That is the whole of our contribution to this page, and it is a caveat, not a correction.
- The join is a geography join, not an occupancy join
We can show that the production geography moved. We cannot show that occupancy moved with it, because no one has re-run the survey.
Do not read our table as evidence that fewer ADUs now house relatives, or more.
It is evidence that the sample and the stock have drifted apart, and that anyone extrapolating should say so.
- Not legal, tax or tenancy advice
Who may live in your ADU, on what terms, and under which local rental rules, is a question of your local ordinance, your lease and your circumstances. This page describes published research and a public permit dataset. It advises no one.
Whose finding this is
The Center for Community Innovation at UC Berkeley ran the only statewide survey of California ADU owners with a published method.
Implementing the Backyard Revolution: Perspectives of California’s ADU Owners, by Karen Chapple, Dori Ganetsos and Emmanuel Lopez, was published on 22 April 2021 and reviewed by David Garcia and Ben Metcalf of the Terner Center for Housing Innovation.
The sample frame was addresses drawn from the Housing Element Annual Progress Reports for ADUs permitted or issued a certificate of occupancy in 2018 or 2019, with a Bay Area supplement; the survey was fielded in 2020, and 752 responses survived the owner screen.
Fact, quoted. The executive summary states: “About half (51%) of California’s new ADUs serve as income-generating rental units, and 16% of ADUs provide no-cost housing to a relative of the homeowner.” It also states: “Only 8% of new ADUs in California are short-term rentals, though more affluent homeowners are more likely to list their ADUs as short-term rentals than those making less than $100,000 a year.” The body of the report adds that ADUs “often serve as affordable housing for friends or relatives: 18% of the state’s new ADUs provide no-cost housing for family members (16%) or friends (2%),” and that “approximately 16% of the new ADUs serve as home offices, studios, or guest houses for the property owner.”
Fact, quoted — the demographic point, which is Terner and CCI’s and not ours. ADUs for All: Breaking Down Barriers to Racial and Economic Equity in Accessory Dwelling Unit Construction (Terner Center and Center for Community Innovation, August 2022) states that “homeowners who build legally permitted (or formal) ADUs—like homeowners in the state overall—are disproportionately White, high-income, and highly educated,” and that “[a] previous study authored by the Terner Center and Center for Community for Innovation found that ADUs in the state are disproportionately built in higher-resource neighborhoods.” Note what that describes: the owners.
It is not a statement about who occupies the units, and we do not stretch it into one. Our study of where ADUs are built carries that finding in full.
What none of this establishes. Neither report profiles ADU tenants. There is no published age distribution, income distribution or household composition for the people living in California accessory dwelling units, and there is no dataset from which one could be constructed without a new survey.
Anyone presenting such a profile is either quoting a survey we have not found or estimating. We do neither.
What the units that are not rented out are doing
The single most useful table in the CCI report for this question is Figure 9, which was asked only of the 49 per cent of respondents who were not renting the ADU out at the time of the survey (n = 366).
It is the closest thing published to an answer for the non-tenant half of the stock.
| Reported use, among owners not renting the ADU out | Share of that sub-sample | What it implies about the occupant |
|---|---|---|
| A friend or relative is staying there for free | 34% | Someone lives there. No lease, no rent, and no trace in any rent statistic — the American Community Survey’s median gross rent table excludes households not paying cash rent entirely. |
| The space is used as something other than an apartment | 34% | Nobody lives there. Office, studio, storage, guest room. The unit is housing on the permit and is not housing in use. |
| It needs physical work before it can be rented | 15% | Nobody lives there yet. A permitted, completed unit is not automatically an occupied one. |
| The owner lives in the ADU and rents the main residence | 13% | The owner is the occupant. The tenant is in the house, not the ADU — which inverts the assumption behind most ADU yield modeling. |
| Vacant and looking for a tenant | 2% | Genuinely between tenancies at the time of the survey. |
| Other | 3% | Unclassified in the report. |
Source: Chapple, Ganetsos and Lopez (2021), Figure 9, n = 366. Sub-sample shares, not shares of all ADUs.
HyreADU analysis: two of the six answers describe a unit with nobody living in it, and one describes the owner living in the ADU rather than the house.
“Who lives in an ADU” has, for a meaningful minority of units, the answer “nobody”.
HYRE calculation
Our contribution: how far the frame has drifted
One calculation, from one public dataset, doing one job: measuring the distance between the cohort the survey measured and the cohort being built now.
Method. We extracted Table A2 of the California Department of Housing and Community Development’s Housing Element Annual Progress Reports, filtered to UNIT_CAT = ‘ADU’, and counted permit units by jurisdiction and year.
The extract carries 281,321 ADU rows and 169,249 unique ADU permit events across 511 jurisdictions reporting ADU permits between 2018 and 2025; HCD’s file was last updated 4 September 2026.
We then compared two windows: 2018–2019, which is the CCI survey frame, and 2024–2025, the two most recent reported years.
HyreADU calculation — statewide. California jurisdictions reported 25,042 ADU permit units across 2018 and 2019, and 63,641 across 2024 and 2025. That is 2.54 times as many units in the later window.
On our project-level de-duplication the same extract gives 178,495 permit units and 102,356 completions across the full 2018–2025 period. The survey frame is therefore a minority of the units that now exist, and by some margin.
HyreADU calculation — composition. Across the ten jurisdictions we track most closely, which account for 15,869 permits in the frame window and 26,443 in the current one, the internal composition moved substantially. Los Angeles fell from 75.0% to 55.6% of the group.
San Diego rose from 5.2% to 16.8%. Sacramento rose from 0.6% to 2.7%, and Fresno from effectively nothing (6 permits) to 188. San Francisco fell from 4.8% to 1.5% and Oakland from 3.4% to 1.5%.
HyreADU analysis. CCI reported that just over half of their responses came from the Bay Area.
In our ten-jurisdiction group the Bay Area share of permits fell from 11.9% to 7.1% between the two windows, while Southern California rose from 87.4% to 89.5% and the two Central Valley cities from 0.7% to 3.5%.
The survey sample was drawn from a production geography that has since moved. We are not claiming that occupancy varies by region — we have no evidence either way, and CCI did not publish a regional occupancy cut.
We are quantifying the extrapolation, so that a reader who quotes “51 per cent” in 2026 knows what they are quoting it about.
Recommendation. If you are citing the occupancy split — in a staff report, a grant application, a pro forma or a news story — cite it as Chapple, Ganetsos and Lopez (2021), give the survey frame years, and state that California has since permitted more than twice as many units.
If your argument depends on the split being current, it depends on a survey that nobody has re-run, and the honest move is to say so rather than to round the caveat away.
Permit geography, survey frame against now
The numbers behind the chart. Shares are of this ten-jurisdiction group only — they are not shares of California, and the group is not a sample of California.
| Jurisdiction | Region | 2018–2019 permits | Share of group | 2024–2025 permits | Share of group | Change in share |
|---|---|---|---|---|---|---|
| Los Angeles | Southern California | 11,895 | 75.0% | 14,701 | 55.6% | -19.4 pp |
| San Diego | Southern California | 827 | 5.2% | 4,447 | 16.8% | +11.6 pp |
| Unincorporated Los Angeles County | Southern California | 722 | 4.5% | 2,260 | 8.5% | +4.0 pp |
| Long Beach | Southern California | 210 | 1.3% | 1,391 | 5.3% | +3.9 pp |
| San José | Bay Area | 608 | 3.8% | 1,060 | 4.0% | +0.2 pp |
| Unincorporated San Diego County | Southern California | 216 | 1.4% | 865 | 3.3% | +1.9 pp |
| Sacramento | Central Valley | 98 | 0.6% | 725 | 2.7% | +2.1 pp |
| Oakland | Bay Area | 532 | 3.4% | 405 | 1.5% | -1.8 pp |
| San Francisco | Bay Area | 755 | 4.8% | 401 | 1.5% | -3.2 pp |
| Fresno | Central Valley | 6 | 0.0% | 188 | 0.7% | +0.7 pp |
| Ten-jurisdiction total | — | 15,869 | 100.0% | 26,443 | 100.0% | — |
| California, all reporting jurisdictions | — | 25,042 | — | 63,641 | — | ×2.54 units |
HyreADU calculation from the HCD Annual Progress Report Table A2 extract, HCD file last updated 4 September 2026.
Table A2 counts permit units as reported by each jurisdiction; reporting practice varies, four jurisdictions in the extract are unusable for completion analysis, and the 2025 rows are the least settled because HCD states prior-year data are not corrected retrospectively in a published file.
Permits are not completions: statewide, the same extract shows 178,495 permit units against 102,356 completions across 2018–2025.
The statewide series the frame sits inside
Why the frame years matter: the survey sampled the two smallest years in the series. Everything after 2020 is outside it.
| APR year | ADU permit units | ADU completions | In the CCI survey frame? |
|---|---|---|---|
| 2018 | 12,050 | 3,186 | Yes |
| 2019 | 12,992 | 5,957 | Yes |
| 2020 | 13,116 | 8,065 | No |
| 2021 | 21,163 | 10,147 | No |
| 2022 | 26,224 | 14,043 | No |
| 2023 | 29,309 | 17,676 | No |
| 2024 | 31,515 | 20,652 | No |
| 2025 | 32,126 | 22,630 | No |
HyreADU calculation from the HCD Annual Progress Report Table A2 extract (UNIT_CAT = ADU), HCD file last updated 4 September 2026.
The permits and completions columns are counted on unique project events, which is why they do not match a naive row count of the same file.
Full method and the permits-against-completions question: our permits versus completions study.
The four households the evidence actually describes
Reading CCI’s survey and Figure 9 together, the population of ADU occupants breaks into four situations with very different data trails. This grouping is HyreADU’s reading of their published tables, not a category they defined.
- The market tenant
- About half the units in the CCI cohort. Pays cash rent under a lease. Visible in a lease, visible to a lender as qualifying income within limits, and subject to whatever local rules apply — including rent stabilisation where a jurisdiction applies it, which we cover in the rent control study. This is the household that ADU yield modeling assumes, and it is roughly one household in two.
- The relative housed at no cost
- Sixteen per cent of the cohort, rising to 18 per cent with friends. This household pays nothing, so it appears in no rent series at all — the ACS median gross rent table excludes households not paying cash rent by construction. A model that capitalises market rent on this unit is describing income the owner has chosen not to collect.
- Nobody
- A real category. In Figure 9, among the 49 per cent not renting the unit out, 34 per cent used the space as something other than an apartment and 15 per cent said it needed work before it could be rented. A completed, permitted dwelling unit is not necessarily an occupied one, which is a caution for anyone converting permit counts into housed people.
- The owner, with a tenant in the house
- Thirteen per cent of the non-renting sub-sample said they live in the ADU and rent out the main residence. The occupancy question inverts: the ADU houses the owner and the larger unit houses the tenant. Where a jurisdiction’s ordinance still turns on owner-occupancy — and for junior ADUs the state requirement survives in a narrowed form — which unit the owner sleeps in becomes a compliance fact, not a preference. See the owner-occupancy study.
Why the national statistical system cannot answer this
Being precise about why no occupant profile exists is worth the space, because the reason is structural and will not be fixed by a better analyst.
The housing tables cannot see an ADU
The 2023 American Housing Survey definitions document does not contain the strings “accessory”, “accessory dwelling”, “ADU”, “granny” or “secondary unit” anywhere.
The units-in-structure concept, which the American Community Survey table B25024 shares, defines a structure as something that “either has open space on all sides or is separated from other structures by dividing walls that extend from ground to roof”.
A detached backyard ADU has open space on all four sides, so it is its own single-unit detached structure — identical in the data to the main house.
An internal ADU makes the property a two-unit structure — identical in the data to a duplex.
That is why no age, income or household-composition table can be filtered to ADU residents. The filter does not exist. Our national count study works through the same mechanism for counting units rather than people.
The permit systems record buildings, not people
HCD’s Annual Progress Report Table A2 records permits and completions with an affordability category, not an occupant. Municipal building permit datasets record an address, a valuation, a contractor and a date. Neither ever contained a fact about who moved in.
Our own Los Angeles permit work can tell you how long a unit took to reach a certificate of occupancy; it cannot tell you who walked through the door afterwards.
Which leaves surveys, and there is one
A voluntary survey of owners, with the response bias that implies, is the entire evidence base for occupancy in California, the highest-volume ADU state. CCI ran it once, in 2020, on a 2018–19 frame. It has not been repeated.
That is not a criticism of CCI — it is the only reason anyone can answer this question at all — but it is the reason this page is a synthesis with a caveat rather than a study with a conclusion.
What we could not verify
Three things. One: our re-fetch of the CCI report PDF at aducalifornia.org returned HTTP 403 on 5 September 2026; the quotations and figure numbers on this page were extracted from that PDF earlier the same day and are reproduced with their page and figure references so they can be checked against the published report.
Two: Census QuickFacts returned HTTP 403 for every request this session, so no ACS age or household-size context appears on this page; we would rather have none than an unsourced one.
Three: we found no repeat of the statewide owner survey, and no published regional cut of the occupancy split. If either exists, this page is out of date and we would like to know.
Questions
Who lives in ADUs in California?
Are ADUs mostly rented to family?
Is there a demographic profile of ADU tenants?
How many ADUs sit empty?
Is the 51 per cent figure still accurate?
Can I rent my ADU to a family member?
Do ADU occupants show up in the American Community Survey?
Where does the short-term rental figure come from, and why do two versions exist?
Written and audited by
HyreADU Research Desk
Primary-source research, data analysis and fact checking
We are a research desk, not a builder. We read the permit extract, the statute, the HCD return or the fee schedule ourselves, and publish each figure with its source and retrieval date.
Where a number cannot be traced to a primary source, we leave it out and say what we could not verify. Our store-based claims cover California only.
- CA
- the only state this desk will make store-based claims about
- 5
- jurisdictions with extracted ADU permit evidence
- 735
- CSLB-verified companies in the California store
- 0
- national claims from a one-state store
How this desk works
- Primary sources only. Permit counts come from the city or county that issued the permit. Production counts come from HCD’s Annual Progress Report. Rents come from HUD or the Census. We do not cite an article that cites a source; we download the source and compute the figure ourselves.
- This is a California site. The company store is 734 California firms and one New Mexico firm. Permit evidence exists for five named jurisdictions: Los Angeles, San Francisco, Sacramento, San José and unincorporated Marin. A number from that store is titled to those places, never to the United States.
- A permit is not a completion, and a license is not an ADU grade. California licenses no ADU classification. Being named on an ADU permit is evidence of engagement in that jurisdiction, not of quality, completion, or work anywhere else. Owner-builder permits are excluded from contractor counts.
- Calculation is labeled as calculation. Figures we derive are never presented as something HCD, HUD, the Census or a city published. Terner Center research is cited as Terner’s, never restated as ours.
- We do not design, permit or build ADUs, and we take no payment for placement, ranking or a favorable mention. Pages that look like rankings are not: they publish public-record counts and let the reader decide.
- Nothing here is legal, tax or financial advice. Zoning, underwriting and appraisal practice vary by jurisdiction, lender and appraiser. The useful next step on a specific lot is the planning counter and a licensed professional.
Data as of CCI owner survey (22 April 2021) and Terner/CCI ADUs for All (August 2022) as cited; HCD Annual Progress Report Table A2 extract, HCD file last updated 4 September 2026; statute retrieved from leginfo 2026-09-05. Authorship on this site is organizational: the analysis belongs to the desk rather than to a named individual, and we do not publish credentials we do not hold.
Our editorial policy sets out how we source, date and correct what we publish.
Sources & retrieval dates
- Karen Chapple, Dori Ganetsos and Emmanuel Lopez, Implementing the Backyard Revolution: Perspectives of California’s ADU Owners (UC Berkeley Center for Community Innovation, 22 April 2021) , The primary occupancy source for this page, cited as CCI’s throughout. Statewide survey of owners of ADUs permitted or issued a certificate of occupancy in 2018 or 2019, drawn from HCD Annual Progress Report addresses with a Bay Area supplement, fielded 2020, 752 responses after the owner screen, just over half from the Bay Area. Executive summary: 51% income-generating rental; 16% no-cost housing for a relative; 8% short-term rental. Body: 18% no-cost for family (16%) or friends (2%); approximately 16% owner use as office, studio or guest house. Figure 9 (n = 366, owners not then renting out): 34% friend or relative staying free, 34% used as something other than an apartment, 15% needs work, 13% owner lives in the ADU and rents the main residence, 2% vacant and looking, 3% other. Reviewed by David Garcia and Ben Metcalf of the Terner Center. Retrieved and text-extracted 5 September 2026; a re-fetch the same day returned HTTP 403. Retrieved 2026-09-05.
- Julia Greenberg, Hannah Phalen, Karen Chapple, David Garcia and Muhammad Alameldin, ADUs for All: Breaking Down Barriers to Racial and Economic Equity in Accessory Dwelling Unit Construction (Terner Center for Housing Innovation and Center for Community Innovation, August 2022) , Cited on this page only for its findings about ADU owners, quoted verbatim: “homeowners who build legally permitted (or formal) ADUs—like homeowners in the state overall—are disproportionately White, high-income, and highly educated”; “[a] previous study authored by the Terner Center and Center for Community for Innovation found that ADUs in the state are disproportionately built in higher-resource neighborhoods”. These are Terner and CCI findings about owners, not about occupants, and they are cited as theirs. Retrieved 2026-09-05.
- California Department of Housing and Community Development, Housing Element Annual Progress Report — Table A2 , The source of every permit figure on this page. Full Table A2 extract, HCD file last updated 4 September 2026: 922,102 rows, of which 281,321 carry UNIT_CAT = ADU, giving 169,249 unique ADU permit events and 98,500 unique ADU completion events across 511 jurisdictions reporting ADU permits between 2018 and 2025. All shares, ratios and window comparisons drawn from it on this page are HyreADU calculations, not figures published by HCD. HCD states that prior-year data are not corrected retrospectively, so the most recent year is the least settled. Retrieved 2026-09-05.
- U.S. Census Bureau, 2023 American Housing Survey Definitions , Cited for a negative result that governs this page. The strings “accessory”, “accessory dwelling”, “ADU”, “granny” and “secondary unit” do not appear in the document. Units-in-structure definition, verbatim: “A structure either has open space on all sides or is separated from other structures by dividing walls that extend from ground to roof. Single-units are structures containing only one housing unit. A single unit is detached if it has open space on all four sides… Structures with two or more units are multi-units.” The American Community Survey table B25024 inherits the same concept. Retrieved 2026-09-05.
- U.S. Census Bureau, Building Permits Survey — methodology and item definitions , Cited for the structural reason permits cannot identify an ADU: the survey classifies buildings as single family (attached and detached combined), two-unit, three- and four-unit, and five-or-more-unit. There is no accessory dwelling item. Item 434 places additions and alterations — which is where a conversion ADU lands — outside the housing-unit count entirely. Retrieved 2026-09-05.
Half the units are tenanted. Plan for the half that are not.
If your plan for an ADU depends on rent, the occupancy evidence says roughly one unit in two in the surveyed cohort was collecting it. The ROI calculator accepts $0 as a valid rent, on purpose.
HyreADU does not design, permit or build accessory dwelling units, and does not practice law.
This page is a synthesis and says so at the top: every occupancy figure on it belongs to the Center for Community Innovation at UC Berkeley, in work reviewed by the Terner Center for Housing Innovation, and is cited as theirs rather than restated as ours.
The only original element is one arithmetic join from the California HCD Annual Progress Report Table A2 extract, labeled as a HyreADU calculation where it appears. No demographic profile of ADU occupants is published, because none is derivable from public data.
This page is informational and is not legal, tax, financial or tenancy advice.