18 percent
was not fine.

A sentence keeps appearing under stories about young buyers: “We paid 18 percent interest and managed.” It is usually offered as a cure for complaint. Sometimes it comes from someone who really did sign a mortgage in that era. The memory deserves respect. The conclusion does not follow.

A rate is the price of borrowing a dollar. It is not the number of dollars a buyer must borrow, or the income available to repay them, or the cost of insuring the thing they bought. Ask a person to remember a single number from a closing forty-five years ago and the interest rate will probably survive. The down payment, the local wage, the houses that never got built, and the neighbor who was turned away may not.

So let us reopen that old closing file. No invented family, no heroic survivor. Just the best national measurements we have, and enough room in the margins to say what they cannot tell us.

The year that was not fine

On October 9, 1981, the average 30-year fixed mortgage rate in Freddie Mac's historical series reached 18.63 percent. It was a peak, not the rate for every 1981 buyer. There is no reason to soften it. Borrowing was brutally expensive.

In the fourth quarter of that year, the Census median price of a new house sold was $70,400. The Census median household income for 1981 was $19,070. Put 20 percent down on that median new house, finance the rest for 30 years at the peak rate, and principal and interest alone come to about $878 a month. That is 55 percent of the median household's pretax monthly income. Taxes, insurance, repairs, and other debt are absent. This is an illustration assembled from different snapshots, not the payment made by a typical 1981 buyer. It is sufficient to puncture the word “fine.”

The housing industry behaved as if it knew. Census historical construction tables count roughly 1.745 million privately owned housing starts in 1979 and 1.084 million in 1981. That fall cannot all be assigned to one weekly mortgage quote, but it is hard to describe the period as an effortless passage into ownership. High rates closed doors then too.

The more interesting question is not which generation suffered more. It is why the old rate can be lower today while the new buyer still feels shut out. The answer begins with the denominator that the internet argument omits.

A lower rate on a larger obligation

The latest Census estimate for a newly sold home was $393,700 in August 2026. Freddie Mac's weekly average for September 24 was 7.03 percent. The newest available median household income was $87,460 for 2025. In nominal dollars, the new-home median went from 3.7 times annual median household income in the 1981 snapshot to about 4.5 times in this newer pairing. The years and groups are not perfectly matched, and new homes are not the whole housing market. The ratio is a lens, not a verdict on the typical buyer.

Here is what the same illustrative loan exercise produces now: a 20 percent down payment of $78,740, a mortgage of $314,960, and principal and interest of about $2,102 a month. That is about 29 percent of the 2025 median household's pretax income. By this narrow test, the 1981 peak was far harsher. If somebody tells you today's financing is worse than that particular week in 1981, these numbers do not support it.

But notice what changed before the first payment. The illustrative down payment was about $14,080 in 1981, or 74 percent of that year's median household income. Now it is $78,740, or 90 percent of the latest median income. A wage cannot be applied to a house purchase all at once. A renter must eat, pay rent, meet other obligations, and accumulate savings while the target moves. The barrier at the front door is not identical to the monthly burden after entry.

The more useful recent comparison may be 2020, not 1981. Harvard's Joint Center for Housing Studies calculates that the monthly cost of a median-priced home, including a low down payment, mortgage insurance, property insurance, and tax under its stated assumptions, rose from about $1,700 in early 2020 to $3,100 in late 2025. The income needed under that method went from about $66,000 to more than $120,000. Existing home prices were up 54 percent from 2020 and stood near five times median household income, rather than the roughly three-to-one ratio of the 1990s. The homes and methodology in this Harvard series differ from the new-home illustration above. They should not be spliced into a single line chart. Together, they describe why the price of admission feels recent and severe.

The interest rate remains important. At a fixed principal, a percentage point can change a lifetime of payments. But the principal is not fixed across generations, and even a dramatic rate cut does not return a house to its former sticker price. You can lower the price of carrying a bag without making the bag lighter.

Two mortgage markets occupy the same street

Imagine two nearly identical houses. In one, the owner bought years ago, refinanced when rates were low, and has no reason to exchange that loan for a new one. In the other, a prospective buyer sees the current asking price and today's financing terms. The houses can have the same roofline. Their monthly budgets live in different decades.

The Federal Reserve's 2025 household survey makes the split visible. Among homeowners with a positive mortgage payment, the national median payment was $1,600. For owners who had moved in 2024 or 2025, it was $2,300. Those are reported payments to mortgage servicers, not a controlled experiment holding home size, region, down payment, or loan type constant. The gap is real in the survey; attributing every dollar of it to interest rates would go beyond the evidence.

That older loan does more than protect a monthly budget. It can keep a house off the market. An Atlanta Fed review of mortgage lock-in research surveys evidence that giving up a cheap fixed-rate mortgage can discourage moving, reducing sales and changing mobility. This is an unusual kind of shelter: the owner is protected from a new rate precisely by staying put. The buyer outside may face both the expensive new loan and a thinner choice of homes.

Some owners would not move regardless of rates. Some sellers have no mortgage left. Builders can offer rate buydowns, and an individual listing may be priced to move. Mortgage lock-in is not a universal explanation for each unsold house or each unaffordable neighborhood. It is a mechanism that changes the amount of housing available to the next person in line.

There is an uncomfortable statistical trick in this divided market. If you average the payments of everyone already inside, the result can look manageable. If you look only at the newly arrived, it can look punishing. Both numbers describe Americans. Neither alone describes the price of joining them. An affordability measure that forgets to ask when did you get your mortgage? is measuring a country with a missing calendar.

Debt is not one household

The other shorthand in the argument says the nation was less indebted in 1981, so high rates then could not have mattered as much. That also deserves a careful audit. Debt has several owners. A family may have a fixed mortgage, a variable credit-card balance, and no student loan. Another may have no house and no mortgage at all. The federal government borrows on another timetable. Put them into one headline number and the mechanism disappears.

The Federal Reserve's Financial Accounts put total liabilities of households and nonprofit organizations at 66.6 percent of disposable personal income in late 1981 and 92.6 percent in the second quarter of 2026. By that consistent stock measure, there is more debt relative to income now. Yet the same series reached 136.8 percent in late 2007. Today's figure is not a new peak. Nor is a stock of outstanding debt the same thing as the payment due next month. Fixed-rate mortgages make that distinction crucial. A higher new mortgage rate bites a new borrower immediately, while an existing fixed-rate borrower may feel no change in that loan's required payment.

The Fed publishes a household debt-service ratio, but its methodology changed beginning in 2024. A neat 1981-to-2026 debt-service line would imply a continuity that the published series does not have. We will not draw it. The absence of a chart can be a form of honesty.

Public debt belongs in a different box. The Congressional Budget Office's 2026 baseline projected about $1 trillion of federal net interest costs for this year, roughly 3.3 percent of GDP. Higher borrowing costs can put pressure on federal budgets as securities mature and are replaced. That does not mean a quarter-point Fed move instantly raises every Treasury coupon, and federal interest expense is not a proxy for whether a renter can buy a house on her block. It is one more reason interest rates matter to the country, not a replacement for the household arithmetic.

The people who do not appear in the mortgage chart

The monthly mortgage comparison has a quiet exclusion: people who never bought. A would-be owner can have no mortgage debt and still be the person most exposed to the new rate, because that rate waits at the gate. It is easy to miss them in a chart of what current homeowners pay.

The National Association of Realtors' survey of buyers put first-time buyers at 21 percent of primary-residence buyers in transactions from July 2024 through June 2025; their median age was 40. The survey observes people who completed a purchase. It cannot directly count all those who wanted to buy and did not. But the shrinking first-time share and older median age are signs worth placing beside the down-payment arithmetic, not beneath a cheerful recollection of 1981.

There is a particularly telling absence in the listings. Harvard's 2026 housing report says the number of homes for sale affordable to households earning $75,000 or less in March 2026 was 60 percent below its March 2019 level, using NAR and Realtor.com data. That is not a statement that all listings fell by 60 percent. It is about homes within reach of a defined income group. A national inventory headline can improve while a particular buyer's search becomes thinner.

Nor does more construction automatically fill that price band. Harvard finds that recent gains in rental supply vary by place. Austin, where construction has been substantial, saw apartment vacancy rise by five percentage points from 2021; Chicago, with more subdued building, saw an increase of only half a point. Over the past decade, the growth in the rental stock has been exclusively in higher-rent units, while the number renting for less than $1,000 fell by seven million through losses or rent increases. The last statistic describes a national change in the stock of low-rent units, not seven million structures demolished. It tells us what kind of home the extra supply did and did not deliver.

Housing shortages and affordability shortages are related, but they are not interchangeable. A vacant luxury apartment cannot necessarily house someone earning $30,000, and a cheaply priced home in a distant market may not let a worker keep the job that pays for it. It matters whether the available unit fits the household's income, location, financing and life. Those qualifications make a map less tidy. They make it more useful.

Renters are not simply waiting in an anteroom, either. They make life choices under the cost of housing now. In the Fed's 2025 survey, 23 percent of renters said they had been behind on rent at some point in the preceding year. Someone whose rent fell due yesterday cannot turn the difference between a 1981 and a 2026 mortgage quote into a down payment tomorrow.

Harvard offers a harder measure of what the word “affordable” leaves after the bill is paid. For 13 million renter households earning less than $30,000, the median amount left for all other necessities after housing was $210 a month in the 2024 data it analyzed, down from $410 in 2019 after inflation adjustment. That residual is not savings waiting to be redirected into a down payment. It is groceries, transport, medicine, and the margin between a normal week and a crisis. The rate comparison has no place for this number, which is exactly why the number belongs here.

Ownership brings costs that a posted mortgage rate conceals. The same Fed survey found 6 percent of homeowners had no homeowners insurance, most of them because of cost. Among insured homeowners, 20 percent wanted more coverage but said they could not afford it. An owner who stretches to pass a lender's payment test may still be exposed to the next insurance renewal, tax bill, or repair. The house is never just the loan.

The dots at the Fed are not a mortgage offer

On September 16, 2026, the Federal Reserve raised its federal funds target range by a quarter point to 3.75 to 4 percent. Its September projections showed a median 4.1 percent federal funds rate at the end of 2026. Read together, that median is consistent with roughly one more quarter-point increase from the current range's midpoint. It is a collection of participants' views of appropriate policy under their assumptions, not a committee promise or a scheduled mortgage-rate increase.

Mortgage rates reflect longer-term bond yields, inflation expectations, credit conditions, and the structure of mortgage securities as well as anticipated Fed policy. They can move before a meeting or in a different direction afterward. “The Fed may raise again” is a reason to examine a range of futures, not to announce what December's 30-year fixed rate will be.