AboutInsightsToolsSign in
All insightsHousing affordability
Oct 10, 202618 min

Can Housing Really Become Affordable Again by 2031?

The payment can become affordable before the buyer can afford to buy.

Begin the story

The monthly payment

The money to get in

The quote fits. The bank balance does not.

Put yourself five years ahead, in October 2031. This is an experiment, not a prediction. The household is invented; the arithmetic is not.

The home cost $450,000 when the experiment began. With prices rising 2.1% a year, it now costs about $499,277. The household's gross income has grown from $100,000 to about $124,618. A new mortgage is available at 6%, down from the experiment's initial 7.5%.

The monthly housing bill is about $3,070. That includes principal, interest, property tax and homeowners insurance. It takes 29.6% of gross income. On a screen with a 30% housing threshold, the payment finally fits.

Then the buyer opens the savings account.

There is about $95,293 in it. A 20% down payment, 3% closing costs and a $15,000 cushion left untouched would require about $129,834. The shortfall is roughly $34,541.

The payment has become easier to carry. The purchase has not become possible under these assumptions.

It is a small distinction with a large consequence. A forecast can announce that affordability is returning while someone trying to buy is still saving toward a target that moves every month. The household does not need one clock to strike the right hour. It needs two.

We will come back to that account. First, consider what a promising headline is actually measuring.

Our opening household / before purchase / USD

A payment is monthly.
Entry happens all at once.

Down payment$90,000

20% of the $450,000 home

Closing costs$13,500

3%, an illustrative estimate

Reserve kept$15,000

Not paid at closing

Cash needed before purchase$118,500

Initial savings$35,000

Still to assemble$83,500

The reserve remains in the account. Deposit plus closing costs total $103,500. This household and these costs are illustrative, not a reported buyer or lending requirement.

What would have to happen by 2031?

On October 8, 2026, Redfin published two hypothetical routes back to its August 2018 payment-to-income benchmark: about five years with a 6% mortgage rate and 2.1% annual price growth, or about six years with a 7.5% rate and flat prices. The study is not a prediction.

There is something useful in asking the question this way. A housing market does not need prices, wages and interest rates to return to their old levels for the relationship between them to improve. The relationship can heal while the individual numbers remain unfamiliar.

A home can grow more expensive in dollars and less expensive relative to income. A mortgage rate can stay uncomfortable while years of income growth make its payment more manageable. Conversely, a lower rate need not rescue a household if the price of the home rises fast enough to absorb the benefit.

These are different routes, with different people helped along the way. Someone with ample savings but a tight monthly budget might gain most from a cheaper loan. Someone with a manageable income and little cash might need a lower purchase price, more time to save, or an eligible lower-down-payment product. An existing owner with a paid-off home may be watching a different problem altogether.

That is why the interesting question is not whether the year 2031 will be good for housing. It is which obstacle would have moved by then, for whom, and what would still be standing in the doorway.

Normal is a benchmark.
Not a promise.

6% rate / prices grow 2.1%

About five years

7.5% rate / prices stay flat

About six years
Report date+ 7 years
Redfin, October 8, 2026. Scenarios, not predictions. Approximate endpoints on a shared scale, not an invented monthly series.

Normal is a benchmark. Not a promise.

A national statistic is a useful place to begin. It is a poor place to stop.

Think of the word normal as a ruler placed beside a changing market. The ruler can tell us how far a relationship has moved. It cannot tell us whether the old relationship was fair, comfortable or available to everyone. Returning to a previous condition is not the same achievement as solving the condition.

There is another problem with a single ratio: the numerator and denominator can each conceal a different life. A median-priced home is not necessarily the home a median-income household can find in the right school district, near a job, with manageable repairs. Neither median identifies a particular buyer.

Even a perfectly calculated market ratio cannot promise that your salary will follow the assumed income path. A change in the median can reflect changes in who lives in a place, as well as changes in what existing residents earn. A personal plan needs a personal income path.

Nor does a ratio know what happens after the paycheck arrives. Taxes, childcare, other debts, food and medical costs make claims on it. The CFPB defines debt-to-income using all monthly debt payments and gross monthly income; a housing-only screen is not that full calculation. A 30% line on a chart is neither a lender's approval nor a verdict that a household can live comfortably.

The benchmark still has value. It makes a moving problem legible. The mistake is asking it to answer a question it does not contain.

The money paid once can keep you out for years.

At the start of our experiment, a 20% deposit on the $450,000 home is $90,000. Closing costs at our chosen 3% add $13,500. The $15,000 cushion does not go to the seller or lender. It stays in the account, available for life after the keys.

Together, those amounts make $118,500 of cash needed before purchase. Of that, $103,500 would be spent on the deposit and modeled closing costs. Calling all $118,500 cash-to-close would be wrong: the reserve is money deliberately not spent.

The CFPB's homebuying guidance puts typical closing costs, excluding the down payment, at 2% to 5% of the purchase price. It also advises leaving an emergency cushion. Our 3% and $15,000 are editable choices, not a quote for a particular loan or a universal reserve requirement.

Now compare the household's two ways of making progress. Its income grows by 4.5% annually in the experiment. Its monthly saving starts at $900 and grows at the same rate. Those are separate assumptions: a raise does not automatically become a deposit in a savings account.

The distinction matters because a household receives its income repeatedly but can retain only a fraction of it. The price of the target home is applied to a much larger base. Two percentages that look close can produce very different dollar movements.

In the first year, 2.1% price growth adds $9,450 to the home's price. At 20% down plus 3% closing costs, that adds $2,174 to the entrance cash target, before any change to reserves. Meanwhile, twelve $900 deposits would add $10,800 to savings before deposit growth. Saving wins that particular annual race, but the account begins $83,500 behind.

Closing a gap is not the same as having a favorable growth rate. The starting distance matters.

This is where advice to wait can become strangely incomplete. Wait for what? A lower payment? A larger account? A slower-moving price? The same year can bring progress on one measure and disappointment on another.

Twenty percent is a choice, not the only door.

Our opening household chose a 20% down payment. It did not establish that every buyer needs one.

Fannie Mae's HomeReady program, for eligible borrowers, can allow 3% down. Eligibility and loan terms matter. A household in this experiment is not automatically eligible for that program simply because a slider reaches 3%.

Reduce our illustrative down payment to 5% and the initial cash requirement falls from $118,500 to $51,000. That is a very different savings journey.

But the money not supplied at the beginning must generally be financed. The loan rises from $360,000 to $427,500. At the same 7.5% rate, more principal produces a larger monthly payment. Add our illustrative mortgage-insurance assumption and the housing bill rises further.

The CFPB explains that borrowers putting less than 20% down typically need mortgage insurance. Its cost and structure vary with the product and borrower. Our study uses a simple annual PMI rate on the new loan below 20% down. It does not quote a lender or model FHA's distinct premium structure.

There is no contradiction here. A smaller down payment can make entry possible sooner and ownership more expensive every month. A larger down payment can make the bill easier and the entrance harder.

The right comparison therefore has two columns, not a moral ranking of good and bad deposits. One asks what cash remains after purchase. The other asks what income remains after paying for the home. A buyer who empties the account to improve one column can damage the other kind of safety.

Try the three deposit settings below with the same home, income and rate. Then use the longer study to see how the trade-off changes over time. The cheapest monthly bill is not necessarily the earliest attainable purchase.

Same home / same rate / three illustrative deposits

Lower the cash hurdle.
Raise the monthly bill.

5%
Cash needed$51,000
Monthly housing$3,803/mo
10%
Cash needed$73,500
Monthly housing$3,634/mo
20%
Cash needed$118,500
Monthly housing$3,117/mo
$450,000 home, 7.5%, 30 years, tax and homeowners insurance included. PMI modeled at 0.6% annually on the loan below 20% down. Cash includes 3% closing costs and $15,000 retained reserve. No product eligibility or lender quote is implied.
Try a different deposit

A fast return can still lead to an expensive place.

The local comparison below separates three things that are easy to confuse: a reported change in prices, an assumed income path, and a date when a previous relationship returns. None, by itself, tells a particular household whether it can buy.

A quick improvement in an expensive place can be real without making it inexpensive. The direction of travel and the height of the hill are different facts.

This is an easy mistake to make on a map. Color the places by the year a benchmark returns, and a quick-return metro can look welcoming. Color them by the actual payment faced by a household, and the message might change. Add the cash needed to begin, and it might change again.

The same applies within a metro. A citywide result is not an offer on the next available home. A household may face a smaller property with an association fee, an older home with repairs, or a longer commute that changes the monthly budget. The median cannot choose among those compromises.

For a buyer, a useful local dashboard would show the total monthly bill, entrance cash, income required and available homes that meet the household's needs. A date alone is not enough.

For a researcher, the lesson is to keep relative recovery separate from absolute affordability. Otherwise, an improvement chart can become a misleading shopping guide.

Redfin / reported prices and scenario income assumptions

Direction is not
the height of the hill.

San Jose

Reported price growth / YoY-3.2%

Projected annual income growth+6.5%

Return to its local benchmarkOctober 2027At a 7.5% mortgage rate

Chicago

Reported price growth / YoY+5.5%

Projected annual income growth+3.9%

Return to its local benchmarkMore than ten yearsAcross the report's modeled rates
October 8, 2026 report. Each place returns to its own 2018 relationship. This comparison does not rank how inexpensive either metro is.

The people who bought are not everyone who tried.

There is a second kind of selection hiding in the evidence. We often learn about buyers by studying people who managed to buy.

In NAR's 2025 Profile of Home Buyers and Sellers, first-time buyers had a median down payment of 10%, versus 23% among repeat buyers. First-time buyers represented 21% of buyers. These are completed primary-residence purchases from July 2024 through June 2025, not today's share of all households hoping to own.

The difference in deposits is worth looking at. So is the limit of what it can tell us.

A successful buyer's balance sheet may contain savings, a gift, proceeds from a previous sale or some combination. Someone still renting may have none of those advantages, or may have chosen not to buy. The survey does not let us assign a single explanation to everyone absent from the closing table.

It does remind us that the doorway is not equally wide. A repeat buyer can arrive with wealth accumulated through a previous home. A first-time buyer must assemble an entrance payment without selling a home first. Those are different starting positions even when both households receive the same mortgage quote.

Suppose rates fall. The existing owner might be able to sell and carry equity into another purchase. The first-time buyer might gain borrowing capacity without gaining a dollar of cash. If competition lifts the price of the target home, some of that financing improvement may be used to bid rather than to reduce the eventual bill. That is a possible mechanism, not an empirical forecast of how much prices will rise.

The point is not that better financing is useless. It is that a change in credit terms and a change in accumulated resources do not reach the same people in the same way.

An affordability recovery should be judged partly by whether new households can enter, not only by the ratios faced by households already able to appear in a transaction dataset.

NAR 2025 profile / completed purchases / older context

The deposit tells us
who reached the door.

First-time buyers10%

Repeat buyers23%

Median down payment / shared 0% to 25% scale

NAR, November 4, 2025. Primary-residence transactions July 2024 through June 2025. The sample describes completed purchases, not everyone who tried.

Two clocks. One purchase.

The study below follows one household and one hypothetical home's price for ten years, beginning in October 2026. It does not reproduce Redfin's model. It lets you examine the mechanisms a national scenario cannot resolve for an individual household.

At each month, the model asks what a fresh 30-year mortgage would cost if the household bought then. It does not pretend the household has already purchased the home, and it does not amortize an existing loan while the household waits.

The first plot divides the prospective housing bill by gross monthly income. The second compares savings with the down payment, modeled closing costs and retained reserve. The outputs record the first month the payment test passes, the first month the cash test passes, and the first month both pass together.

That last phrase is important. If a line crosses a threshold and later crosses back, simply taking the later of two first-crossing dates can be wrong. Both conditions must hold in the same month. Even then, the model has passed two screens, not obtained a lender's approval or located a suitable home.

Start with the default household. Then lower the down payment. The savings hurdle gets smaller, while the payment grows. Raise future insurance growth without changing the mortgage rate. The cost of occupying the home can become less forgiving even as financing improves. Change income growth and saving growth separately. A household can earn more without being able to put the entire increase away.

Finally, try prices growing faster than income and set monthly saving to zero. The study will not manufacture a reassuring year. If a condition never holds within the ten-year window, it says so.

Do not use a favorable output as permission to purchase. Use the unfavorable cases to identify the assumption on which the favorable one depends.

Two clocks.
One purchase.

The payment can become affordable before the buyer can afford to buy.

Illustrative household. Not a forecast or Redfin model replication.

Payment test

Housing cost / gross income (%)

01020304020262028203020322034203630% screenBill /income

Monthly housing$3,070

Gross-income share29.6%

Cash test

Savings and cash needed / USD

$0$50k$100k$150k$200k202620282030203220342036SavedNeeded

Savings$95,293

Required cash$129,834

First payment-test passJul 2031
First cash-test passSep 2034
First month both pass togetherSep 2034

Cash shortfall$34,541

At this datePayment passes. Cash does not.

Change assumptions Price, income, savings, down payment, rates and growth
Costs, reserves and additional assumptions

Valid changes update both charts. An empty or out-of-range entry keeps the last valid value and restores it when you leave the field.

Download CSV

Assumptions and method

Fresh 30-year mortgage at every possible purchase date, plus property tax, homeowners insurance and illustrative PMI below 20% down. Cash needed includes a retained reserve, not only funds paid at closing. No association fee, maintenance, other debts or household expenses are included in the housing screen. Savings deposits occur at month end. Markets do not respond to one another in this model. A pass is not loan approval.

The offered rate moves linearly to your future rate over the chosen period; it is not a prediction. Both-pass date is the first simultaneous pass, not necessarily the later first crossing. Conditions can reverse later. Full method and sources.

Exact annual observations
Selected October observations. Full precision monthly series and assumptions are in CSV.
YearMonthly housingIncome shareSavingsCash neededBoth pass
2026$3,11737.4%$35,000$118,500No
2027$3,10935.7%$46,021$120,673No
2028$3,10134.1%$57,538$122,893No
2029$3,09132.5%$69,573$125,158No
2030$3,08131.0%$82,150$127,472No
2031$3,07029.6%$95,293$129,834No
2032$3,13628.9%$109,027$132,245No
2033$3,20428.3%$123,379$134,707No
2034$3,27327.6%$138,377$137,221Yes
2035$3,34427.0%$154,050$139,788Yes
2036$3,41726.4%$170,428$142,408Yes

A ratio can improve while the bill rises.

Return to the five-year view. The payment is roughly $3,070, only a little below its starting $3,117. Gross income, by contrast, has risen by nearly a quarter. Much of the apparent improvement came from the denominator.

Continue to October 2036 under the same assumptions. The monthly bill has risen to about $3,417. It now takes only 26.4% of the modeled household's gross income. The payment is higher; the ratio is lower. Both statements are true.

This is why a falling burden line should not be drawn as if it were a falling bill. It is also why nominal income growth cannot be treated as a complete account of rising living standards. If other necessities become more expensive, the household may not feel the full relief suggested by the gross-income ratio.

The default saving path assumes the household can keep making deposits while waiting. Rent, childcare, repairs to a car, a job interruption or a medical bill could change that path. We have not predicted those events or smuggled in an estimate of their probability. Instead, saving and saving growth remain visible inputs.

There is a deeper limitation. The three market paths are chosen independently: rates, prices and income are not generated by an economic model that makes them respond to one another. You can test a combination that is mechanically consistent but economically difficult to sustain. That is useful for sensitivity analysis, not evidence that the combination is likely.

A good scenario tells you what follows if the inputs hold. A forecast must also defend why the inputs are likely to hold. This page offers the former.

For a real household, the next step would be to replace the simple assumptions with income after tax, actual insurance and tax quotes, other debts, realistic expenses, product-specific financing and a reserve appropriate to that household. The aim would not be to get the earliest date possible. It would be to find a date that survives a less generous set of assumptions.

Our default experiment / nominal bill and gross-income ratio

The bill can rise.
The burden can fall.

2026$3,117/mo

37.4% of gross income

2031$3,070/mo

29.6% of gross income

2036$3,417/mo

26.4% of gross income

One illustrative household. In 2036 the nominal housing bill is higher than in 2026, but the assumed income path makes the ratio lower. Not a forecast of wages or living standards.

What would count as a recovery?

There is a temptation to define recovery as the moment a line reaches an old mark. It creates a clean ending. Housing rarely provides one.

A meaningful recovery would have to work at several levels. Monthly ownership costs would need to fit more household budgets. Entrance cash would need to be attainable without stripping away the money needed to absorb trouble. Suitable homes would need to exist where those households can use them. Financing would need to be available on terms they can sustain.

These conditions suggest different tests for different interventions. A rate reduction should be evaluated for its effect on the full payment and on the purchase price a buyer ultimately faces. Deposit assistance should be evaluated for the cash it supplies and the obligations that come with it. A smaller down payment should be evaluated alongside the larger loan. Additional homes should be evaluated for location, total cost and suitability, not only the count of units.

None of those observations establishes a universal policy winner. They provide a better way to ask whether a change reaches the constraint actually blocking entry.

Consider two households with identical incomes. One has $140,000 in savings; the other has $35,000. A mortgage quote that becomes manageable can open a door for the first while leaving the second waiting. Calling the market affordable for both because their incomes match would discard the very fact that separates their decisions.

Now consider two households with the same savings but different income security. A fixed monthly bill may fit both today and be a very different commitment for each tomorrow. Affordability is not simply capacity to get through closing. It is capacity to remain housed afterward.

That does not make measurement hopeless. It makes the choice of measure important. Track the payment, the entrance cash, the buffer after purchase and the homes actually available. Show who each improvement reaches. Keep the test visible rather than allowing a word like normal to perform all the work.

The two answers meet

A household
cannot buy a ratio.

The account we left open.

Back in October 2031, the buyer's quote still fits the chosen screen. The savings account still falls short.

In the default experiment, both tests finally hold in September 2034. Change the assumptions and that date changes, or disappears. There is no prophecy hidden inside it.

What has changed is the question the household would ask of the headline. Not simply: will housing become affordable in five years? But: would this home fit our monthly budget then, and would we have enough money to enter without leaving ourselves exposed?

The first question is about a bill. The second is about an account. The home waits where those two answers meet.

There is room for cautious hope in a market that can improve without an abrupt collapse in prices. There is also a reason to insist that improvement be described carefully. A household cannot buy a ratio, and a national return date cannot make a down payment.

The mortgage quote may be the first page of the story. The balance left after closing is part of the ending.

Evidence and model notes

Primary sources

Our household experiment

All dollars are nominal USD. The default home is $450,000; gross household income $100,000; initial savings $35,000; initial monthly saving $900; down payment 20%; closing costs 3%; retained reserve $15,000. Prices grow 2.1% annually, income and monthly saving 4.5%, insurance 3%. These are illustrative inputs, not national estimates. Homeowners insurance begins at $2,250 annually; property tax is 1.1% of the future modeled price each year. Savings yield defaults to zero.

The offered mortgage rate declines linearly from 7.5% to 6% over five years and stays there. This path is an assumption, not Redfin's method and not a forecast. Every month represents a new possible purchase with a new 360-month loan. Principal-and-interest payment is Lr/[1-(1+r)^(-360)], where L is loan principal and r the monthly rate. At zero interest, payment is L/360.

Price and income compound by annual growth raised to months/12. Insurance compounds from its starting dollar premium, independently of price growth. Property tax is recomputed from the modeled price; actual assessments and tax rules may differ. Below 20% down, illustrative monthly PMI is loan principal times the annual PMI rate divided by twelve. There is no cancellation path: each observation prices a new loan, not an already amortizing one.

Savings grow by the chosen annual yield converted to an effective monthly rate, with a deposit at each month end. The first deposit equals the initial monthly saving; subsequent deposits grow at the separate saving-growth input. Yield is interpreted as after-tax and after-fee. No withdrawals, house-price feedback, job shocks or investment volatility are modeled. The retained reserve stays fixed in nominal dollars.

Required cash is down payment plus modeled closing costs plus retained reserve. The reserve is not paid at closing. The payment screen is housing costs divided by gross monthly income; it excludes other debts, utilities, maintenance, association fees, taxes on income and nonhousing expenses. The default threshold is 30% and can be changed. It is not a full affordability budget or underwriting test.

The window has 121 monthly observations, October 2026 through October 2036 inclusive. First crossings are reported separately. The combined date is the first observation at which both tests hold simultaneously; it does not guarantee they continue to hold. CSV downloads include assumptions and the full monthly series. Screen values are rounded; calculations are not. This is an educational sensitivity study, not personal financial advice, an interest-rate offer or a guarantee of a purchase date.

Read as Markdown Explore the mortgage calculator
More insights