Sierra Leone's GIS-Supported Teacher Deployment Algorithm

Sierra Leone has struggled to place qualified teachers in disadvantaged rural/remote districts, facing high turnover, subject-specialist shortages in maths/science, and under-representation of female teachers. The Teaching Service Commission (TSC), with the Ministry of Basic and Senior Secondary Education, EdTech Hub, the Learning Generation Initiative and analytics firm Fab Inc., built on earlier UK Aid-funded GIS analysis of pupil-teacher-ratio disparities.

To replace ad hoc, discretion-heavy teacher placement with a data-driven, rules-based deployment mechanism that better matches teachers to schools while directing more new positions to the 10 most disadvantaged of Sierra Leone's 16 districts (identified via national exam and learning-assessment results).

For the 2024/25 deployment cycle, the TSC introduced an open-source GIS-supported preference-matching algorithm adapted from the deferred-acceptance ("stable matching") design used in Nobel Prize-winning work matching doctors to hospitals and patients to kidney donors. Schools list required teacher attributes (subject specialism, experience, gender balance, willingness to stay); teachers list preferred locations including language and family ties; the algorithm produces placements respecting both sides' preferences.

For the 2024/25 academic year, 56% of the 2,000 new Ministry of Finance-funded teaching positions and 53% of 1,000 new early-childhood-development positions were assigned to the 10 disadvantaged districts, meeting a results indicator tied to $4.5 million of GPE financing. A companion budget-execution target rose to 92% in 2023/24, from a 2019 baseline of 78%.

This is a genuine step toward data-driven deployment, with criteria and results externally verified through GPE's results-based financing mechanism. There is no independent evaluation yet of downstream effects on teacher retention, pupil-teacher ratios or learning outcomes, and disparities may be larger within districts than between them, so district-level targeting may not fully close intra-district gaps.