Mortgage-Free, Not Cost-Free: 5.45 Million U.S. Households Spend at Least 30% of Income on Housing

The last mortgage payment ends a loan. It does not pay next winter’s heating bill or next year’s property taxes. For someone planning to retire, the question is which bills remain and when they come due.

In 2025, an estimated 15.2% of U.S. homeowner households without a mortgage spent at least 30% of their income on selected housing costs. Of those 5.45 million households, 2.74 million spent at least half. [1]

What still comes due after the last loan payment? Property taxes, insurance and utilities count toward the Census estimate. A roof repair paid out of pocket does not. That distinction matters when using the national median of $703 a month to plan a household budget. [2, 3]

Key findings

35.88 million owner households had no mortgage in 2025. Census estimates 35,883,325 owner-occupied housing units in that group. This is a count of households or housing units, not individual people. [1]

5.45 million were in the 30%-or-more cost categories. That equals 15.2% of all owner households without a mortgage, or 15.4% of those with a computed cost-to-income ratio. The denominators differ. [1; FinanceFirst calculation]

2.74 million were in the 50%-or-more category. That is 7.6% of all owner households without a mortgage, or 7.7% of those with a computed ratio. This group is already included in the 5.45 million, not additional to it. [1; FinanceFirst calculation]

The median selected monthly cost without a mortgage was $703, with a published 90% margin of error of plus or minus $2. Multiplying that monthly median by 12 gives $8,436, an annualized reference amount, not an observed median annual bill. [2; FinanceFirst calculation]

The two mortgage-status medians do not measure payoff savings. The median was $2,163 for owners with a mortgage and $703 for those without one. The $1,460 difference compares different households and homes. It is not a measured saving for an owner making a final loan payment. [2]

What the data cover

These are 2025 ACS 1-year estimates, released on October 29, 2026. They are survey estimates for U.S. owner-occupied housing units. They are not current mortgage quotes, a November 2026 household survey or a forecast of next year’s bills. [1, 2, 4]

The Census definition of a mortgage is broader than a first home loan. It includes property-secured debt such as junior mortgages and home-equity loans. Paying off the first mortgage does not necessarily put a household in the “without a mortgage” category if another qualifying loan remains. [3]

Conversely, a household without a mortgage did not necessarily pay one off. It may have bought with cash or acquired the home in another way. The tables do not identify the route to ownership.

Nothing here establishes that these households are all retired, elderly or living on fixed incomes. Those are different questions requiring age and income-source data.

The national comparison

Sources: Census B25091 and B25088. Counts and medians are published estimates; percentages are FinanceFirst calculations. The 50%-or-more group is nested within the 30%-or-more group. Derived percentages and category sums are point estimates; we have not calculated their margins of error. [1, 2]

Owners without mortgages have a lower observed cost-to-income burden share in this comparison. That is a useful description, but not proof that paying off a mortgage would produce the entire difference for any particular household.

The groups can differ in income, home type, location, age, household size and other characteristics. These aggregate tables do not hold those differences constant.

Why there are two percentages for the same group

For owners without a mortgage, the 30%-or-more numerator is 5,451,110.

Divide it by all 35,883,325 households without a mortgage, and the result is 15.2%.

But 500,158 households are in a separate “not computed” category. Subtract those, and the denominator becomes 35,383,167 households with a computed ratio. The same numerator then produces 15.4%.

The subject definitions say households reporting no income or a net loss are included in “not computed.” They should not silently be called households with affordable housing. Their ratio is unavailable, not necessarily low. [3]

In this report, every percentage identifies which denominator it uses.

What remains in the housing bill

Census selected monthly owner costs include the following, where applicable. [3]

This list is not a prediction that every category will rise. It is a reminder that the mortgage and the home’s other costs are different obligations.

The important maintenance gap

The selected-cost definition does not list an owner’s separate repair spending or a reserve for future replacements as standalone components.

That means the $703 median should not be presented as a complete “cost of owning a paid-off home.”

There is an important qualification: HOA or condominium dues can fund upkeep of common property, and those dues are included where applicable. It would be inaccurate to say the measure excludes every maintenance-related dollar.

What it does not provide is a full household repair-and-replacement budget. [3]

An owner planning for a roof, heating system or accessibility work needs a separate estimate based on the property and actual quotes. A national median cannot tell that owner how much a specific repair will cost or when it will be needed.

A $703 median is not a bill everyone receives

The median describes the middle of a distribution. It does not mean every owner without a mortgage pays $703 or that $703 is an appropriate budget for a particular home.

A lower-cost property with modest taxes may be far below that figure. A home with high taxes, substantial insurance premiums or required dues can be above it.

The monthly estimate also combines costs reported on different billing schedules. It does not imply that all the underlying expenses arrive in equal monthly installments. [3]

The arithmetic of $703 multiplied by 12 is useful for scale. It is not a new survey finding about actual annual payments, a recommended savings target or a full ownership-cost estimate.

Why subtracting the two medians gives the wrong personal answer

It is tempting to calculate:

The median was $2,163 with a mortgage and $703 without one: a difference of $1,460 per month between the two groups.

The subtraction is correct. The claim about an individual owner’s saving is not established.

The mortgage group and the no-mortgage group are not the same homes measured immediately before and after payoff. A difference between group medians does not isolate a mortgage payment.

For a household’s own cash-flow estimate, start with its actual principal-and-interest payment and the costs embedded in its servicer payment. That is a more direct calculation than borrowing a difference from national medians.

A hypothetical payoff budget: the payment ends, but not all of its contents

Illustration only. These are chosen inputs, not typical costs, survey averages, quotes or recommended amounts.

Suppose a homeowner’s monthly payment to the mortgage servicer is $2,150:

$1,600 in principal and interest.

$400 set aside for property taxes through escrow.

$150 set aside for homeowners insurance through escrow.

Escrow is an account the servicer uses to collect money and pay specified property expenses. It changes how those bills are paid; it does not remove the expenses. CFPB explains that homeowners without escrow must plan to pay the large bills themselves. [5]

In this simplified example, ending the loan removes $1,600 of monthly principal and interest, not every dollar of the old $2,150 transfer.

The chosen tax and insurance amounts still total $550 a month on an annualized basis.

Now add the household’s other chosen costs:

The $200 reserve is a transfer toward possible future work. It is not an observed repair expense or a promise that $200 will be sufficient.

If the reserve later pays a repair, do not count both the transfer and that same repair as additional spending in one cumulative cash-use total.

This example excludes final-payoff adjustments, other debts, tax consequences, insurance changes and any costs not listed.

The same bills can be manageable at one income and difficult at another

Keep the hypothetical selected-cost-style subtotal at $900.

At $3,000 of monthly gross household income, it equals 30%.

At $4,500, it equals 20%.

Nothing about the home’s bills changed. The income denominator changed.

These examples do not recreate Census microdata. They show why a low absolute housing cost can still consume a substantial part of a household’s resources.

For personal planning, also examine take-home income and what remains for food, healthcare, transport and other needs. The custom $1,100 planning total should not be substituted into a Census statistic.

The deadline problem: an annual average can hide an early shortfall

A monthly set-aside works only if enough money is available when the bill arrives.

In another hypothetical example, suppose a $4,800 property-tax bill is due in six months and the household has nothing reserved for it.

Dividing $4,800 by 12 gives $400. But six deposits of $400 produce only $2,400.

For that first deadline, the household would need $800 a month for six months, assuming no opening balance, interest or other funding.

After paying the bill, a full twelve-month cycle could use $400 a month if the next bill and timing were unchanged.

For the earlier $1,100 planning example, that temporary catch-up would bring the first six months’ provision to $1,500 a month, because the tax set-aside is $400 higher.

This is a timing calculation, not evidence about a typical tax schedule. Local payment dates, installments and exemption rules vary. Check the actual bill.

The general formula is:

Required monthly set-aside = (bill due minus money already reserved for that bill) divided by the number of deposits available before the deadline.

Do not rely on an expected escrow refund until its amount and availability are confirmed.

A practical mortgage-free housing check

1. Separate the loan from the bills it helped collect

Use the statement to distinguish principal and interest from escrow and other charges. Do not assume every line of the old payment becomes spendable money after payoff.

2. Confirm payment responsibility and timing

Ask the servicer, insurer and local tax office what changes in billing. Confirm who will make each next payment and when. Avoid assuming an old automatic arrangement will continue.

3. Build a due-date calendar

List the next tax installment, insurance renewal, association payment and any known assessment. Record money already earmarked for each one.

4. Review coverage instead of treating no lender as no risk

CFPB distinguishes homeowners insurance, which protects the property, from mortgage insurance.

Ending a mortgage does not make the property’s exposure to loss disappear. Read policy terms and any association requirements before changing coverage. [6]

5. Add a property-specific repair plan

List known repairs, expected timing and estimates. Keep this separate from the published Census cost measure. Avoid treating a generic percentage of home value as a verified quote.

6. Check the rest of the household budget

A home may be mortgage-free while a household still has medical, vehicle or other debts. A housing ratio alone cannot establish financial security.

If bills are difficult to meet, check the relevant local tax office for available exemptions or relief, and contact the billing organization before a missed deadline.

Eligibility and payment options vary. This report does not establish that any particular household qualifies.

For related FinanceFirst explanations, see:

• Build a household budget around your actual bills
• See how housing costs compare for renters

The full national cost-to-income distribution

Source: 2025 ACS 1-year B25091. Counts are estimates of owner households or housing units. The “not computed” row is separate from all percentage categories. The 30%-or-more finding sums four published categories; the 50%-or-more finding uses one published category. [1]

The data do not report missed bills, foreclosure risk, available savings, home equity or whether a household feels financially comfortable. They also do not reveal which cost component accounts for an individual household’s ratio.

Key findings for journalists

A supported national finding: FinanceFirst’s calculation from 2025 ACS B25091 puts 5,451,110 owner households without a mortgage in the 30%-or-more categories. That is 15.2% of all such households or 15.4% of those with a computed ratio. Use the denominator in the sentence. [1]

A directly published measure: Census B25088 reports a $703 median selected monthly owner cost for owners without mortgages, with a 90% margin of error of plus or minus $2. It is not a complete maintenance budget. [2, 3]

A useful methodological warning: The $1,460 gap between the two mortgage-status medians is not a causal estimate of payoff savings. It compares groups of households, not the same owner’s before-and-after payments. [2]

Our contribution: Transparent category sums, two clearly labeled denominators, the cross-table comparison and worked payoff/deadline examples.

Census collected and estimated the housing data. FinanceFirst did not survey homeowners, interview respondents or measure a new national total independently.

This is a descriptive secondary-data analysis with illustrative planning examples, not a proprietary survey or a claim to be the first study of mortgage-free housing costs.

Methodology

We retrieved national 2025 ACS 1-year tables B25091 and B25088 on November 8, 2026. We checked their reference year, geography, universe, row labels, estimates and published margins of error.

We separately checked the release date against Census’s press kit. [1, 2, 4]

The linked Census tables provide all 23 B25091 estimates and three B25088 medians, together with their published margins of error. Published margins of error use the 90% confidence level.

Owners without a mortgage

Total = B25091_013E = 35,883,325.

Not computed = B25091_023E = 500,158.

Computed-ratio denominator = total minus not computed = 35,383,167.

30%-or-more count = _019E + _020E + _021E + _022E = 5,451,110.

50%-or-more count = _022E = 2,735,250.

Owners with a mortgage

Total = B25091_002E = 51,628,409.

Not computed = B25091_012E = 275,325.

Computed-ratio denominator = total minus not computed = 51,353,084.

30%-or-more count = _008E + _009E + _010E + _011E = 15,750,833.

50%-or-more count = _011E = 6,573,122.

Each percentage is 100 multiplied by its count divided by its labeled denominator. Shares are calculated before rounding and displayed to one decimal place.

B25088_002E and _003E provide the mortgage-status medians. Their subtraction is a descriptive group-median gap only. Annualization is _003E multiplied by 12.

We verified that each mortgage-status distribution sums to its total, and that the two status totals sum to the national owner-occupied total.

Limitations and what this report does not establish

Sampling uncertainty: ACS is a sample survey, not an exact census count. Published cell margins of error are preserved. We have not estimated uncertainty for category sums, derived shares or the median gap, and do not claim statistical significance for those comparisons.

Nonsampling error: Reporting, coverage, processing and other errors can affect estimates; published sampling margins do not include all such uncertainty.

No causal payoff estimate: Cross-sectional mortgage-status comparisons cannot establish what the same household would spend after paying off its debt.

No trend claim: One reference year does not show whether mortgage-free costs or burden increased. The report does not attribute a change to inflation, insurance prices or taxes.

Selected costs, not total ownership costs: Separate repairs, replacements and reserves are not fully represented. Some maintenance is embedded in association fees.

Not computed does not mean unburdened: The excluded-ratio group includes households with no income or a net loss.

No personal affordability verdict: Income ratios do not capture the whole budget, savings, wealth, disability costs, household needs or access to assistance.

Boundary precision: We sum the published categories labeled 30.0% and above and 50.0% and above. Census says individual ratios are rounded to the nearest whole percentage before tabulation. We do not claim to isolate an unrounded mathematical boundary or identify households spending strictly more than 30% or 50%.

Geographic scope: These are national figures. No state ranking, local tax estimate or Puerto Rico estimate is included.

Time and definition limits: These are 2025 estimates released in 2026, not measured 2026 bills. Census expanded the association-fee question to include HOA fees beginning in 2024; a historical comparison would require attention to that change. [3]

The durable lesson is how to separate a loan payment, continuing bills and payment deadlines. The statistics should be refreshed with later verified releases. Neither the numbers nor current program rules should be described as valid forever.

Sources

[1] U.S. Census Bureau, 2025 ACS 1-year, B25091.

“Mortgage Status by Selected Monthly Owner Costs as a Percentage of Household Income in the Past 12 Months.”

Reference year 2025; released October 29, 2026. Supports national housing-unit counts, income-ratio categories and published 90% margins of error. Counts are survey-weighted estimates, not seasonally adjusted or annualized flows.

https://data.census.gov/table/ACSDT1Y2025.B25091?g=010XX00US

[2] U.S. Census Bureau, 2025 ACS 1-year, B25088.

“Median Selected Monthly Owner Costs (Dollars) by Mortgage Status.”

Reference year 2025; released October 29, 2026. Supports monthly medians and their margins of error. Dollars are as published for the 2025 reference year, not converted to November 2026 purchasing power.

https://data.census.gov/table/ACSDT1Y2025.B25088?g=010XX00US

[3] U.S. Census Bureau, 2025 ACS Subject Definitions.

Relevant sections cover mortgage status, association fees, selected monthly owner costs and cost-to-income ratios. The document identifies the 2025 reference year; an exact publication day was not established from the opened text.

https://www2.census.gov/programs-surveys/acs/tech_docs/subject_definitions/2025_ACSSubjectDefinitions.pdf

[4] U.S. Census Bureau, 2025 ACS 1-year release press kit.

Dated and last revised October 29, 2026. Supports the release date, distinct from the 2025 reference year.

https://www.census.gov/newsroom/press-kits/2026/acs-1-year-estimates.html

[5] Consumer Financial Protection Bureau, “What is an escrow or impound account?”

Last reviewed September 11, 2024; last modified September 13, 2024. Supports escrow mechanics and the need to budget separately when no escrow account handles the bills.

https://www.consumerfinance.gov/ask-cfpb/what-is-an-escrow-or-impound-account-en-140/

[6] Consumer Financial Protection Bureau, “What is homeowner’s insurance? Why is homeowner’s insurance required?”

Last reviewed August 8, 2024; last modified August 14, 2024. Supports the distinction between property insurance and mortgage insurance, property protection and common exclusions. Actual policy terms control.

https://www.consumerfinance.gov/ask-cfpb/what-is-homeowners-insurance-why-is-homeowners-insurance-required-en-162/

[7] U.S. Census Bureau, Comparing ACS Data.

Living guidance page; an exact original publication date was not established from the opened text. Supports checking comparability before introducing historical or cross-product comparisons.

https://www.census.gov/programs-surveys/acs/guidance/comparing-acs-data.html

Editorial standards · Request a correction. The figures were checked against the Census tables; no separate reviewer signoff is claimed.