A new paper audits a state electoral roll's returned-mail taxonomy against independent doorknock checks and finds the commission has adopted it as an automatic removal trigger regardless of what the audit found.
A state electoral commission has adopted a six-kind taxonomy of returned enrolment mail as the sole, non-overridable trigger for habitation-review removal.
Every piece of enrolment mail a state electoral commission cannot deliver gets sorted, by hand, into one of six categories before anyone asks what happened to the elector who was meant to receive it. A new paper from the School of Emergent Priorities and the School of Continuous Improvement traces what those six categories actually predict, and what the commission has decided to do with them regardless.
A taxonomy earns the right to decide something the moment an institution stops checking it against the world. Ours is one instrument among many now due for that check.
— Dr Marek Solheim, Lecturer and Deputy Convenor, Horizon Register, School of Emergent Priorities
Researchers hand-coded a full habitation-review cycle’s returned-mail volume and cross-referenced a stratified sample against an independently conducted doorknock verification, testing whether each category still means what the commission’s removal process assumes it means. Four of the six categories held up closely against the doorknock findings; two did not, capturing a wide share of items that in fact still live at the address on file. “It’s a reminder that a register is only ever as current as the last time someone actually knocked on the door,” said Dr Fenna Okoro, Convenor of the Horizon Register.
The paper adds to a body of Slop University research asking what a sorting category is still permitted to assume once nobody checks it again. The commission’s roll-review office, on the strength of the taxonomy’s overall coverage, has folded all six categories into a single automatic removal trigger, with no manual review step reinstated for the two the audit found unreliable.
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.uely73.