Everyone's been passing around a survey where boomers answered and millennials didn't. Turns out, that makes the data a little boomer-heavy. Who could have predicted this.

The Briefcase Team | April 21, 2026

You have seen the chart. You have seen it so many times you probably have an opinion about it.

It's the National Association of Realtors survey that claims the median U.S. homebuyer is now 59 years old, up from 31 in the early 1980s. First-time buyers, per NAR, are now 40.

The chart has gone truly viral. Fortune covered it. Apollo's chief economist Torsten Slok used it in a client note. The New York Times ran it. It's been shared a million times by TikTok finance guys, bitter millennials, and every politician trying to score housing-policy points.

There's one tiny problem: the number is wildly, provably wrong.

Not "slightly off." Not "subject to methodological debate." Wrong by about 15 to 20 years. The actual median homebuyer in America is somewhere around 42. And in a rare moment of plot twist, your bitter millennial cousin on Facebook has been passing around a statistic that is worse than the truth.

The truth is already bad enough. We do not need the fake version.

Connor O'Brien, a researcher at the Institute for Progress, pulled the receipts last November. He went to two data sources that dwarf NAR's survey in rigor: the American Housing Survey and the American Community Survey, both run by the U.S. Census Bureau and considered the gold standard for American housing data.

His finding: the median homebuyer in 2023 was 42, per AHS data. The American Community Survey put it at 41, unchanged for well over a decade.

Meanwhile, NAR's survey says the median buyer in 2023 was 49, and by 2025 had somehow aged to 59. That's ten years of aging in two calendar years. This would require either the invention of a time machine or a survey with a methodological problem. One of those is more likely.

Other organizations piled on. The American Enterprise Institute and Cato Institute, using mortgage data from the New York Federal Reserve's Consumer Credit Panel (a 5% random sample of all credit reports tied to Social Security numbers), found the median first-time buyer was about 33, steady from 2001 to 2024. The Mortgage Bankers Association found a modest rise from 30 to 33 over the past decade, then a dip to 32 in 2025. Redfin's own analysis, shared with Business Insider, pegged the median first-time buyer at 35 last year, slightly younger than the year before.

And the kicker: Zillow ran its own national homebuyer survey this summer and got a median age of 42. Which is 17 years younger than what NAR reported with a straight face.

Six independent data sources. One outlier. Guess which one went viral.

This is the part that should make every data-literate person in real estate laugh out loud. NAR's Profile of Home Buyers and Sellers is a mail-and-text survey sent to recent buyers. This year they got 6,103 responses.

Out of how many? About 175,000 surveys sent. That's a response rate of roughly 3.5%.

Now, a 3.5% response rate isn't automatically disqualifying. What is disqualifying is who actually answers mail surveys in 2026. Hint: it's not 28-year-olds. AEI researchers crunched the respondent pool and found buyers under 35 were under-represented by 17 percentage points. Buyers aged 45 to 74 were over-represented by 18 percentage points.

The chart didn't measure the median homebuyer. It measured the median homebuyer who still answers paper mail. Those are dramatically different populations. One of them has an AOL email address and a rotary phone in the kitchen.

To NAR's credit, they've been responsive about the debate. Their economists publicly defended their methodology and offered a possible theory: young buyers priced out of their home markets are buying investment properties instead of primary residences, and NAR's survey only captures primary residences. It's an interesting hypothesis. There's just no evidence it's happening at anywhere near the scale needed to turn a 42-year-old median into a 59-year-old one.

If you stopped caring three paragraphs ago, come back. This is not a pedant's dispute. Bad data drives bad policy.

The fake 59-year-old chart has been cited in Congressional testimony, think tank reports, and at least three op-eds arguing for massive federal first-time buyer subsidies. Marketplace ran it. Fortune treated it as fact. Politicians used it to argue for down-payment-assistance programs that, in a supply-constrained market, would do exactly what every first-year economics student predicts: push prices up and benefit sellers, not buyers.

The real story, the 42-year-old story, points to a completely different policy response: build more houses.

Because here's what the Census data actually shows. At age 32, 41.3% of millennials own homes. At age 32, 54.7% of boomers owned homes. That's a 13-point gap. It's real. It's bad. It's the actual statistic that should drive housing policy.

But it points to supply problems, zoning problems, and construction-cost problems. Not to a "late-starting buyer" problem solvable with a $15,000 federal credit. The fake chart implies the cure is cash. The real chart implies the cure is houses.

We believe the housing affordability crisis is real. Prices are up 82% since 2000 while incomes grew 12%. First-time buyer market share is at an all-time low of 21%. America is 10 million homes short. These are facts, documented across multiple independent sources.

What we don't believe in is inflating those facts with bad data to make a point that the real data already makes. When a 3.5%-response-rate paper survey produces a number that contradicts every other survey, mortgage dataset, and Census analysis, you don't use that number. You use the six that agree.

The fake chart hurts the housing-reform cause. It turns every serious supply-side argument into an easy target for skeptics who can just say "the data is wrong." It makes policymakers chase fake problems with fake solutions. And it makes the real problem - that a 32-year-old millennial is 13 percentage points less likely to own a home than a 32-year-old boomer was - easier to ignore because it sounds less dramatic.

Good housing policy deserves good housing data. A 3.5% response rate from people who still check their mail isn't it.

The median American homebuyer is not 59. They're 42. The median first-time buyer is not 40. They're somewhere between 32 and 35, depending on whose dataset you trust. These numbers are not flattering for the housing market, and they should absolutely drive a conversation about supply, zoning, and construction costs.

But the viral version of the chart - the one your uncle keeps forwarding, the one cable news keeps running, the one politicians keep citing in press conferences - is based on a paper survey with a 3.5% response rate in which boomers over-answered by 18 percentage points and buyers under 35 under-answered by 17. That is not a dataset. That is a group text. The truth is already bad enough. We don't need to make it 17 years worse to care about fixing it.