The homeownership rate is the share of occupied housing units that are owner-occupied; interpretation requires the correct denominator, survey, and margin of error.
The homeownership rate is the percentage of occupied housing units that are occupied by their owner or co-owner. It is calculated by dividing owner-occupied units by all occupied units. Vacant housing units are excluded from the denominator.
In the United States, the Census Bureau’s Current Population Survey/Housing Vacancy Survey (CPS/HVS) publishes quarterly and annual homeownership estimates. The rate measures housing tenure, not the percentage of people who own property, the percentage of homes owned free and clear, or the share of buyers using mortgages.
Because total occupied units equal owner-occupied units plus renter-occupied units, the same formula can be written as:
where O is owner-occupied units and R is renter-occupied units.
Suppose an area has:
The number of occupied units is 100, not 110:
Dividing 65 by all 110 housing units would produce 59.1%, but that is not the homeownership-rate definition because the 10 vacant units do not have resident owner or renter households.
A housing unit is owner-occupied when an owner or co-owner lives in it. The classification applies whether the unit is:
The rate therefore does not distinguish a household with substantial Home Equity from one with a high loan balance. Both are owner households if an owner lives in the unit.
All other occupied units are classified as renter-occupied under the Census tenure definition. This includes units rented for cash and units occupied without payment of cash rent.
As a result, the renter-occupied share of occupied units is the mathematical complement of the homeownership rate:
This complement is not the rental vacancy rate. Rental vacancy measures vacant units available for rent relative to the rental housing inventory and uses a different denominator.
The rate classifies the household’s occupied residence. A person who owns an investment property but rents the home in which they live belongs to a renter-occupied household for this measure. A tenant does not become an owner-occupant because their landlord owns the building.
Likewise, ownership of a vacation home or other property does not put that property in the numerator unless it is occupied by its owner as the relevant residence under the survey rules.
Rates are commonly compared in percentage points. If the homeownership rate rises from 65% to 66%:
1Percentage-point change = 66% - 65% = 1 percentage point
2Relative percent change = (66 / 65 - 1) x 100 = 1.54%
Saying the rate “rose 1%” is ambiguous. The clearest statement is that it rose 1 percentage point, equivalent to a relative increase of about 1.54% from the earlier rate.
A small percentage-point movement can also be statistically uncertain. The margin of error must be checked before describing the difference as a real change in the broader population.
The Census Bureau administers the CPS/HVS using a probability-selected sample of occupied and vacant housing units across all states and the District of Columbia. Its current methodology describes a sample of about 72,000 housing units.
Sample households follow a rotating pattern: they are included for four consecutive months, leave the sample for eight months, and return for four more months. Census field representatives collect responses through personal and telephone interviews.
The HVS is weighted to housing units rather than to the population. This is appropriate for estimating owner-occupied, renter-occupied, and vacant units, but it also explains why household or housing-inventory estimates from other Census products may not match.
Available tables include homeownership rates by:
Availability, frequency, and reliability vary by table and geography. State and metropolitan estimates generally require more caution than the national rate because their effective samples are smaller.
When a table reports homeownership by age, race, ethnicity, or another householder characteristic, the category refers to the householder used for tabulation. It does not mean every person living in the unit shares that characteristic.
Group differences are descriptive. They do not establish that age, family status, race, ethnicity, or income directly caused the tenure outcome. Income, wealth, location, household structure, credit access, historical conditions, and many other factors may overlap.
CPS/HVS estimates come from a sample and can differ from the result that a complete census would produce using the same definitions and procedures. Census publishes margins of error for national, regional, and selected subgroup rates.
The HVS release uses 90% confidence intervals. If an estimated rate is 65.0% with a margin of error of 0.5 percentage point, the reported interval is:
165.0% - 0.5% = 64.5%
265.0% + 0.5% = 65.5%
Suppose the next estimate is 65.3%. The point estimate is 0.3 percentage point higher, but that alone does not prove the underlying rate increased. The source’s statistical test, standard errors, and margins of error determine whether the difference is statistically significant.
Do not use the overlap or non-overlap of two confidence intervals as a universal significance test. Use the Census comparison statement or an appropriate test based on the published standard errors.
Margins of error describe sampling uncertainty, not every possible error. Survey estimates can also be affected by:
Large-looking subgroup differences can be unreliable when margins of error are also large.
Census provides seasonally adjusted and not-seasonally-adjusted homeownership series for some national tables. Seasonal adjustment attempts to remove recurring within-year patterns so quarter-to-quarter changes are easier to compare.
Before combining or comparing observations, confirm:
Seasonal adjustment cannot remove sampling error, structural changes, or unusual events. It also does not turn a descriptive rate into a forecast.
The HVS provides both revised and non-revised historical tables for some housing-inventory series. Census explains that:
Updated controls can incorporate newer estimates of housing units, building permits, housing losses, and other administrative information. The latest revised series may therefore differ from the values available at an earlier date.
For historical research or model backtesting, record the series name and vintage. Use as-published data when the question is what an analyst knew at the time, and revised data when the objective is the latest consistent historical estimate.
Several Census programs publish homeownership estimates, including CPS/HVS, ACS, and AHS. Their results should not be assumed to match.
| Survey | Primary analytical role | Typical strength | Comparison caution |
|---|---|---|---|
| CPS/HVS | Timely vacancy and homeownership measurement | Quarterly national and regional monitoring | Sample-based estimates and margins of error matter |
| ACS | Broad annual social, economic, demographic, and housing estimates | Detailed geographic coverage | Different collection period, weighting, residence rules, and sample design |
| AHS | Detailed housing-stock, occupant, cost, and quality analysis | Rich housing characteristics | Different purpose, frequency, and survey design |
Choose the source that fits the question. CPS/HVS is usually the relevant source for the current quarterly national homeownership headline. ACS is often better for detailed geographic or demographic analysis. AHS can support deeper analysis of housing characteristics and costs.
Do not splice levels from one survey into the time series of another without documenting a methodology break.
Because the rate is a ratio, it can change through movements in either owner or renter households.
| Change | Possible rate effect, all else equal |
|---|---|
| Owner households increase faster than renter households | Rate rises |
| Renter households increase faster than owner households | Rate falls |
| Owner households decline while renter households are stable | Rate falls |
| Renter households decline while owner households are stable | Rate rises |
| Owner and renter households grow at the same percentage rate | Rate remains unchanged |
This arithmetic does not identify why the counts changed. Relevant factors can include:
No single factor has a guaranteed effect. For example, lower mortgage rates can improve payment affordability for some borrowers while stronger demand raises prices or limited inventory constrains purchases.
Assume an area has the following estimated occupied units:
| Period | Owner-occupied | Renter-occupied | Total occupied | Homeownership rate |
|---|---|---|---|---|
| Period 1 | 650,000 | 350,000 | 1,000,000 | 65.0% |
| Period 2 | 653,000 | 342,000 | 995,000 | 65.6% |
The rate rises by 0.6 percentage point:
Owner occupancy increased by 3,000 units, but renter occupancy fell by 8,000. The higher rate therefore does not mean 0.6% of the population bought homes or that all of the increase came from purchases.
Before interpreting the change, an analyst should check:
The example shows why the numerator and denominator matter as much as the headline rate.
| Measure | What it answers | Why it differs |
|---|---|---|
| Homeownership rate | What share of occupied units are owner-occupied? | Tenure ratio; excludes vacant units |
| Renter-occupied share | What share of occupied units are renter-occupied? | Complement of homeownership among occupied units |
| Homeowner vacancy rate | What share of homeowner housing inventory is vacant and for sale? | Includes a specific vacancy category and different denominator |
| Rental vacancy rate | What share of rental inventory is vacant and available for rent? | Not the renter share of occupied units |
| Existing Home Sales | How many covered resale transactions completed? | Transaction flow, not tenure stock |
| New Home Sales | How many qualifying new-house contracts or deposits occurred? | New single-family contract activity |
| Home equity | What is the owner’s value net of secured debt? | Balance-sheet measure, not occupancy classification |
| House Price Index | How did covered home prices change? | Price-change measure, not ownership share |
These measures can diverge. The homeownership rate can remain stable while sales fall, prices rise, or mortgage applications change because the stock of occupied households adjusts more slowly than transaction flows.
The rate provides broad context on the share of occupied housing held in owner tenure. It does not measure how many owners have mortgages, their loan balances, loan types, delinquency status, or refinancing behavior.
Mortgage demand depends more directly on home purchases, refinancing, household credit, rates, and loan balances. Use the homeownership rate as a structural context variable, not a substitute for origination data.
Owner occupancy can expose households to home-price changes, maintenance costs, taxes, insurance, and mortgage obligations. It can also support equity accumulation when property value exceeds secured debt.
The rate alone does not quantify any of those outcomes. A highly leveraged owner and a mortgage-free owner receive the same tenure classification, while property value and costs vary widely.
Homeownership rates can help describe tenure patterns across regions and groups. Policy analysis should pair them with affordability, income, rent burden, housing quality, supply, credit access, and mobility data.
A policy that raises the rate is not automatically beneficial if households assume unaffordable debt or if the comparison ignores risk and opportunity cost. A lower rate is not automatically harmful when renting better fits household mobility, finances, or preferences.
Aggregate tenure can inform market context, but it does not establish a particular borrower’s ability to repay or a property’s collateral value. Underwriting still depends on verified income, assets, debts, credit, appraisal, loan terms, and applicable lending requirements.
The prior page suggested that a stable rate around 60% to 70% is generally desirable. There is no authoritative universal threshold supporting that conclusion.
Homeownership and renting involve different tradeoffs:
The appropriate question is not whether a rate is good in isolation. It is what the rate, its components, and related evidence show about tenure, access, risk, and household outcomes in the specific context.
Before using a homeownership rate, verify:
Homeownership statistics are educational market evidence, not personalized mortgage, investment, appraisal, legal, or tax advice. Verify the current Census table and methodology before using an estimate in a financial decision.