The Affordability Problem

Method

Income needed to qualify

20% down, a 28% front-end ratio, principal and interest only, on a 30-year fixed at the rate quoted for that date. The 28% front-end ratio is the lender's conventional test: housing payment against gross pay.

“The same house”

One house, not the mix of houses that happened to sell. The 2019 median sale price is carried forward by Case-Shiller, which is a repeat-sales index — so what you're seeing is one asset repriced, not a shift toward bigger or smaller homes. The raw median sale price today is lower than the figure the film uses; that gap is mix, not error.

Rates

Mortgage rates are weekly survey averages. The 7.79% peak is the week of 26 October 2023. Where a rate is quoted as “now,” it is the most recent weekly print at the time of publication. The Fed's own rate is the effective rate, monthly average — where the film says “five and a half,” that is the top of the target range, and 5.33% is what the rate actually did.

Why price and rate multiply

A 58.1% price rise and a 39.6% rate effect do not add to 97.8%. They compound: 1.581 × 1.396 = 2.208, or +120.8%. The film shows the addition struck through, because the mistake is the point.

Who qualifies

The worker figures are full-time workers, from the weekly earnings distribution — not the population, and not households. The household pair is dated 2019 against 2024, because household income data stops at 2024; it is the honest apples-to-apples comparison, and it counts everyone, including retirees and people not trying to buy.

There are two rent numbers on this site, and they disagree by design

The film uses CUSR0000SEHA, the BLS index for rent of primary residence, which is up 34.1% since October 2019. The front page uses ZORI, Zillow’s observed asking rent, which is up 39.5% over the same window. Both are measured through July 2026, both are correct, and they are describing different people.

The BLS series averages everyone currently paying rent, most of whom are part-way through a lease signed at an older price. Zillow’s index is what it costs to sign a lease now. So the official number describes the population and the asking number describes the decision — and someone looking for an apartment faces the second one.

It also arrives damped. Asking rents peaked at +15.4% year-on-year in February 2022. The BLS series peaked at +8.8% — thirteen months later, at a little over half the height. The official series never reported the size of the move, only its echo.

This is why the front page and the film quote different figures. Neither is wrong; the film is describing what happened to renters as a group, and the front page is describing what it costs to move.

“The bottom of the freeze” is not the bottom of the series

If you pull ACTLISCOUUS you will find a lower reading than the one the film uses: 346,514 in February 2022, against the 562,453 quoted. That is not an error, and the wording is the reason. February 2022 is a month before the Fed's first hike — it is the bottom of the pandemic buying frenzy, when everything for sale sold. The freeze starts with the rate rise in March 2022, and its own bottom is March 2023. That is what “the bottom of the freeze” means, and it is the comparison the film makes. Measured from February 2022 instead, listings are up 3.25× rather than 2.00× — the film takes the smaller number, because it is the one that answers the question being asked.

Revisions

Some of these series restate recent months after the fact. Case-Shiller has already revised May 2026 from the 335.104 the film used to 335.430, a change of +0.10%, which moves the qualifying income by about $110. Figures are quoted as published, and every one is dated. Where a revision is large enough to change a claim, it goes in Corrections below, with both values and the date.

Inflation

Nothing on this site or in the film is adjusted for inflation. Where inflation appears it is reported as what it did, from the published index — never used to deflate someone's income.