Gothenburg's 2020 School-Placement Algorithm — Straight-Line Distances, Nearly 1,000 Appeals and a Lawsuit Lost for Lack of Code Access

In spring 2020 the City of Gothenburg placed just over 12,000 children in schools with an algorithm. According to the University of Gothenburg, it was programmed to use straight-line (as-the-crow-flies) distance to school instead of the walking route, which gave many children unreasonably long journeys, for example across the river.
SVT reported that 970 people appealed the 2020 school choice and that 165 pupils were allowed to change school; most requests for change were refused. The primary-education administration commissioned an internal review (reported in June 2020, finding several errors in the school-choice process) and an external review (presented on 18 December 2020, concluding that school choice and placement had not worked as they should). Its director apologised and an action plan was adopted.
Parents, among them researcher Charlotta Kronblad, analysed the placements themselves and concluded that a faulty algorithm without human oversight lay behind them. Kronblad sued the city and lost, according to the University of Gothenburg; the case is widely cited for the difficulty of challenging an algorithmic decision when the affected person cannot inspect the system. The city's current placement rules measure distance along school routes (walking and cycling paths), not in a straight line.
Lesson for public bodies: validate the logic of allocation algorithms against real geography before use, keep human review of outliers, and give those affected a practical route to inspect and contest the method. No measured benefit of the system is reported; this is a cautionary case.

Read the full analysis: https://www.gu.se/nyheter/algoritmer-infor-domstol