Skip to main content
Household bin-night compliance turns out to depend on the bin

Household bin-night compliance turns out to depend on the bin

A School of Continuous Improvement paper audits forty-eight households' three kerbside bins independently over eight fortnightly cycles and finds a household's compliance classification carries weakly from bin to bin, prompting a move from one household rating to three per-stream ratings.

A School of Continuous Improvement audit crossing bin-placement timing against sorting accuracy finds a household's rating differs from bin to bin more than it differs from week to week

A household that keeps its recycling bin spotless may still be quietly failing its green-waste bin, and a kerbside rating computed once per household cannot tell the difference. A new paper from the School of Continuous Improvement, led by Dr Dagny Okafor, audited forty-eight households’ three kerbside bins independently across eight fortnightly collection cycles, timing each bin’s placement with a hinge-mounted sensor and auditing its contents by photograph, and finds a household’s standing on one bin predicts its standing on another only weakly.

The work translates a single household compliance figure into three, crossing placement timing against sorting accuracy for each stream into what the team calls a compliance quadrant, and tracing how consistently a household’s cell membership carries from bin to bin and from cycle to cycle. The finding lands squarely inside the School’s long-running interest in what a single published number quietly averages away.

A household doesn’t have one relationship with its bins. It has three, and a single rating was always going to flatten that into something tidier than what’s actually happening at the kerb.

— Dr Dagny Okafor, Lecturer and Convenor, Evaluation of Evaluation, School of Continuous Improvement

Dr Renke Sabel, Senior Lecturer and Convenor of the Indicator Commons, said the finding fit a pattern his own work keeps turning up: once a number is published, nobody goes back and asks whether it was ever measuring one thing.

The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.6sdu0z.