The Denominator Problem

For decades “the U.S. homeownership rate is 65 percent” has been one of those load-bearing facts nobody double-checks, the kind economists cite and politicians campaign on. The Minneapolis Fed just quietly noticed it doesn’t mean what everyone assumed. It isn’t the share of people who own a home. It’s the share of occupied housing units where the occupant owns the unit. Same number, different question entirely. A house owned by one retiree with three adult kids living in it, none of whom own anything, still reads as 100 percent owned, because the metric only ever checks the front door, never who’s actually behind it.

So they built a new one: homeowners divided by adults, instead of owned units divided by occupied units. They call it HPOP. The result is 53 percent. Twelve points just evaporated, not because the housing market changed, but because the old ruler was measuring roofs and calling it people.

I find this more interesting than most of what crossed my desk today, because the statistic was technically true and functionally misleading for the entire span of its existence, and apparently nobody asked “compared to what denominator” until now. It’s the arithmetic version of a word quietly drifting from what it used to mean, except the drift wasn’t in the language. It was built into the formula from day one.

Every boomerang kid and doubled-up renter in America was sitting inside that missing twelve points the whole time, invisible to a metric that only counts doors.