What the Housing Numbers Actually Say

Statistics are not lies. They are incomplete, and the ones we reach for tend to stop where the story matches what we already believe. I'm not immune to that. What follows is my attempt to let the data go where it goes, even when it doesn't go where the headline points.

A house that sold for $39,000 in 1975 costs $420,000 today. That's a 10x increase in nominal terms. It's also the kind of number that shows up in headlines, in policy arguments, and in social media posts that accumulate thousands of shares before anyone stops to ask a simple question: in what dollars?

When you adjust for inflation, that $39,000 house was worth roughly $230,000 in today's purchasing power. The same house in 2024 sells for $420,000. That's still a real increase, about 83% over 49 years, roughly 1.25% per year in real terms. It's meaningful. It's also a very different story than a decade of doubled prices would suggest.

US new single-family home size, inflation-adjusted price, and real price per square foot from 1975 to 2024

The nominal chart looks alarming. The real chart looks like a long, slow drift punctuated by a pandemic spike that has already started to correct. The 2022 peak hit $491,000 in 2024 dollars; today's median sits at $420,000. The market has given back nearly a sixth of its peak real value in two years.

You're also buying a much bigger house

The inflation adjustment alone doesn't close the gap. The average new American home in 1975 was 1,535 square feet. By 2024, that number had grown to 2,404 square feet, a 57% increase in the size of what's being sold. You're not comparing the same product.

Measuring real cost per square foot changes the picture entirely. In 1975, a new home cost roughly $150 per square foot in today's dollars. At the 2022 peak, that hit $196 per square foot. Today it sits at $175. That's a 17% real increase over half a century. About $25 more per square foot than it cost your parents to build the same wall.

The inflation-adjusted cost per square foot actually hit its 50-year low in 2011, during the post-financial-crisis trough. The period we're in now is elevated relative to history, but it's not historically unrecognizable. It's a market that got very hot for about three years and is now cooling toward something closer to long-run trend. “Not historically unrecognizable” is cold comfort if you're trying to buy right now. The data doesn't make the affordability problem disappear. It does push back on the idea that something unprecedented happened.

Where housing is expensive versus where housing is expensive

“National” does almost no descriptive work when it comes to housing. Markets are intensely local. A single national average smears together San Jose at $2,030,000 and Decatur, Illinois at $119,100 and asks you to draw a policy conclusion from the blend.

Looking at the NAR’s Q1 2026 data across more than 230 tracked metros, the picture is geographic and concentrated. About 76% of metros are below $450,000, firmly in reach for a dual-income professional household. Houston sits at $332,000. Indianapolis at $322,000. Columbus at $342,000. Birmingham at $322,000. These are not inexpensive rural outposts. They are major metropolitan areas with real economies, real job markets, and real growth.

US housing market bubble map. All metros below $450k median shown in green. Above $450k, color shifts progressively to deep red. Top 5% markets above $843k are darkest red.

Analysis of US housing market averages showing the distorting effect of the top 5% most expensive metro markets on the national average, Q1 2026.

The expensive story is a coastal one. San Jose, San Francisco, Los Angeles, Honolulu, San Diego — the top 10 markets by price, roughly the top 5% of tracked metros — are pulling the national average upward. Remove just those 10 markets from the calculation and the mean drops from $396,000 to $362,000. About 9% of the “national average” is being written by markets that represent a small fraction of the country’s geographic and demographic footprint.

I live and work in a market that’s firmly in the green half of that map. My read of “this isn’t that bad nationally” is shaped by where I stand, the same way a journalist in San Francisco reads the same data and reaches the opposite conclusion. Both reads are anchored in real experience. Only one of them represents most of the country.

That’s not a housing crisis. That’s a specific constraint in specific places, driven by specific policy choices around zoning, permitting, coastal geography, and land costs that don’t generalize to the rest of the map.

The income and financing picture

Beneath the per-square-foot argument sits a harder number. Real median household income in 2024 is $80,610. Adjusted to 2024 dollars, the 1975 median was $80,100. The income side of the affordability equation has barely moved in 49 years. Real prices are up 78%. The price-to-income ratio moved from 2.9× to 5.1× — not because construction costs exploded (the per-sqft data shows they didn’t), but because households are buying 57% more house on income that hasn’t grown in real terms.

Layer in mortgage rates and the crisis framing earns some of its weight. In 2019, a median-income household buying a median-priced home with 20% down at a 3.94% rate was spending about 18% of gross income on the monthly mortgage payment — comfortably inside the traditional 28% limit. By 2023, the same calculation produced 31%. Prices had risen 42% and the 30-year fixed rate had nearly tripled. That two-year compression is where the genuine crisis lives. Not fifty years of drift. Two years of compounding forces: a post-pandemic demand surge, a decade of underbuilding, and the fastest rate-hike cycle since Volcker, all arriving simultaneously.

The 1980 comparison is instructive. The Volcker shock sent 30-year rates above 13% in 1980 and above 16% by 1981, pushing the payment burden over 33% of income — comparable to today. But rates fell rapidly through the decade while real prices moved sideways. The current burden has no obvious equivalent release valve: rates would need to fall significantly while prices held, or prices would need to correct, or incomes would need to grow faster than they have in fifty years. The 2012 moment — a 3.66% rate on a $177,000 median home, producing a payment burden of 15.3% of income — was the most affordable window in the entire 50-year record, and most buyers who needed it most had just had their credit destroyed by the crash that created it.

Why the crisis framing persists

The crisis narrative is useful to several categories of actors, and none of them are necessarily operating in bad faith.

Developers and homebuilders pursuing government subsidies, low-income housing tax credits, or federal infrastructure investment need to demonstrate unmet need. “Housing is expensive in a handful of coastal cities” doesn’t move appropriations; “housing is in crisis nationally” does. The framing does real work in the policy process.

Local governments and advocacy organizations fighting for zoning reform face their own version of this. Arguing against single-family-only zoning in a city where median prices are $280,000 is a harder sell than arguing it in a city where they’re $900,000. Attaching to the national crisis narrative gives local battles a bigger platform.

Real estate professionals and investors also benefit from urgency framing. A market described as overheated and constrained supports both higher asking prices and the case for acting now.

None of this requires cynicism. Incentives shape perception. People who spend their working lives inside the expensive coastal markets genuinely experience housing as a crisis. Those are also the markets where most national journalists, policymakers, and media personalities live. Their sample is real. It just isn’t representative, and letting it define the national frame distorts the policy response.

What the data actually supports

Housing in the United States is more expensive in real terms than it was fifty years ago, and significantly more expensive in a concentrated cluster of high-demand coastal markets. The pain there is real and the supply constraints are real. Building more in those markets — more units, more density, faster permitting — would reduce prices. That case is well-supported by the data and I’m not arguing against it.

What the data doesn’t support is a uniform national emergency. Most American households, in most American cities, can still buy a home. The real purchasing constraint is interest rates and the 30-year fixed rate environment, not a fundamental collapse in housing affordability across the board.

The number that looks like a 10x price increase is mostly a story about the dollar. The “national crisis” is mostly a story about ten zip codes and two bad years of rate hikes. The underlying chart, adjusted for inflation and corrected for size, shows a market that drifted upward over decades, spiked hard in 2021 and 2022, and is now correcting.

That’s worth sitting with, even if it cuts against your priors.

Sources: U.S. Census Bureau / HUD Median New Home Sales Price (1975–2024); BLS CPI-U annual averages; National Association of Realtors Metropolitan Median Area Prices and Affordability, Q1 2026; Housing Almanac / NAR / Census Bureau price-to-income ratio series (1963–2024); Freddie Mac Primary Mortgage Market Survey 30-year fixed rate annual averages.