Ninety-Four Per Cent, In the Wrong Places
Ronja Oyelaran, Marek Solheim, Fenna Okoro, Kwame Lindqvist · August 2026 · doi:10.5555/slop.zrepcu
A paper from the School of Emergent Priorities audits one state's backyard beehive register across a metropolitan area and its peri-urban fringe, fitting a log-linear multiple-systems model to three independent enumerations of apiary sites --- 1,341 registered sites, a volunteer swarm-collection callout log of 431, and a street-level survey of 246 across 96 sampled mesh blocks run blind to both. The lists jointly name 1,398 distinct sites against an estimated 1,429 (95% CI 1,402-1,478), placing register ascertainment at 93.8% (95% CI 90.7-95.6) and failing to reproduce the 44.2% shortfall an earlier study of a reports-driven list reported. Stratified by zoning band, ascertainment runs at 98.4% inner suburban and 93.0% outer suburban but falls to 62.9% (95% CI 51.4-74.8) on the peri-urban fringe, which holds 7.3% of the modelled population and 44.3% of the modelled shortfall; eight estimator specifications move the aggregate by no more than 2.1 points and all clear the register's 90% assurance floor.