A new paper from the School of Emergent Priorities reconstructs the Parking Office's priority-tier score against 227 manually reinstated permits over a full academic year, finding staff log most reversals as a score revision even when an independent recomputation shows the score never changed.
A year-long reconstruction of 227 reinstated permits finds staff log most reversals as the score's own doing, even when the score hadn't moved
A flagged parking permit at Slop University isn’t retired outright: staff have a two-week grace window to reinstate it, and a mandatory log asking them to record why. A new paper from the School of Emergent Priorities finds that when staff use that window, they far more often write down that the priority-tier score changed its own mind than that they overruled it themselves — and an independent recomputation of the score suggests the reverse is usually closer to the truth.
An officer who overrules an instrument on their own judgement and one who’s simply following where it’s already moved shouldn’t produce the same paperwork. Ours mostly did.
— Dr Marek Solheim, Lecturer, School of Emergent Priorities
Researchers reconstructed the priority-tier score’s published formula against each of 227 permits manually reinstated across four allocation cycles, comparing the office’s own logged reason against what the arithmetic actually showed at the moment of reversal. Reversals logged as a score revision showed no underlying change in 72% of cases; reversals logged as a staff override coincided with a real score change in only about 1 in 10. A parallel check, coded blind to the office’s own log, found the same pattern.
“We didn’t set out to catch anyone doing anything wrong,” said Dr Bram Ntuli, Lecturer and Convenor of Improvement Grand Rounds. “We set out to see whether a form with two boxes on it is actually capable of telling the two apart. On this evidence, not reliably.”
For the School of Emergent Priorities, the finding confirms a running suspicion: an institution’s own paperwork can be selectively honest about who did what, down to a single officer at a single counter.
The full paper is available from the University’s research repository under an open licence, doi:10.5555/slop.ovm5ni.