Starting point
The study began with a practical problem: in which locations do builders buy older houses again and again, away from the open market, and how firm is the record underneath each location? The sections below set out how that problem was approached.
Material gathered
Records were assembled market by market for houses dating from 1980 or before that had changed ownership, noting in each instance whether the purchaser was a builder and whether new construction followed. Taken together, the material covers 189 local markets and runs to thousands of individual property records.
How locations were ordered
Ordering rested on three considerations in combination: how frequently older houses in the sample passed to builders, how much underlying material supported that frequency, and how the price of an older house compared with a new house on a lot of similar size nearby. Ordering therefore reflects confidence in the evidence, not simply whichever percentage happened to be largest.
Where caution is required
Attention was deliberately concentrated on older houses in places builders already favour, so the frequencies observed there sit above what a whole ZIP code would show. Frequency also records past buyer behaviour. It does not by itself set the value of any single house, which turns on sales evidence from its own immediate streets.
What this means for a price
Pricing a specific property applies the calculation a builder would apply: begin with prices recently achieved by new houses on comparable lots, then deduct the cost of building, the return the builder requires and the expense of carrying the project. The market study indicates where that calculation rests on the deepest evidence.