A community-driven optimization framework for redrawing school attendance boundaries

The vast majority of US public school districts use school attendance boundaries to determine which student addresses are assigned to which schools. Existing work shows how redrawing boundaries can be a powerful policy lever for increasing access and opportunity for historically disadvantaged groups, even while maintaining other priorities like minimizing driving distances and preserving existing social ties between students and families. This study introduces a multi-objective algorithmic school rezoning framework and applies it to a large-scale rezoning effort impacting over 50,000 students through an ongoing researcher-school district partnership. The framework is designed to incorporate feedback from community members and policymakers, both by deciding which goals are optimized and also by placing differential ``importance’’ on goals through weights from community surveys. Empirical results reveal the framework’s ability to surface school redistricting plans that simultaneously advance a number of objectives often thought to be in competition with one another, including socioeconomic integration, transportation efficiency, and stable feeder patterns (transitions) between elementary, middle, and high schools. The paper also highlights how local education policymakers navigate several practical challenges, like building political will to make change in a polarized policy climate. The framework is built using open-source tools and publicly released to support school districts in exploring and implementing new policies to improve educational access and opportunity in the coming years.

💡 Research Summary

The paper presents a community‑driven, multi‑objective optimization framework for redrawing school attendance boundaries, demonstrated through a large‑scale partnership with Winston‑Salem/Forsyth County Schools (WS/FCS) in North Carolina. Over 50,000 students across 67 elementary, middle, and high schools were considered, making this one of the most extensive field deployments of a school‑boundary rezoning algorithm to date.

The authors first situate their work within a long line of school‑redistricting research, noting that early models (e.g., Clarke and Surkis, 1968) treated the problem as a single‑objective assignment under travel‑time constraints, while more recent studies have added additional criteria such as racial integration or transportation cost. However, most prior approaches either fix the set of objectives a priori or embed certain constraints (like feeder patterns) directly into the data preprocessing stage, limiting flexibility and community input.

To overcome these limitations, the authors formalize the “School Attendance Boundaries Redrawing” (SABR) problem. Planning units (e.g., census blocks) are assigned to schools at each grade level, and the assignment must satisfy a suite of realistic administrative constraints: school capacity limits, minimum and maximum zone sizes, road‑crossing prohibitions, and a compactness metric (Polsby‑Popper or similar). The objective function aggregates four policy goals: (1) transportation efficiency (minimizing total student travel distance), (2) socioeconomic (SES) integration (minimizing a dissimilarity index across zones), (3) feeder‑pattern stability (preserving logical elementary‑middle‑high school sequences), and (4) geometric compactness.

Crucially, the framework incorporates community feedback in two distinct phases. In the design phase, a district‑wide survey of more than 8,000 residents identified which goals should be prioritized and provided raw importance scores. The authors then applied representational‑adjustment techniques to correct for demographic imbalances in survey participation, producing a set of weights (w₁…w₄) that scale each objective’s contribution in the overall optimization. In the optimization phase, these weights are directly embedded in the multi‑objective linear combination, allowing policymakers to explore trade‑offs quantitatively.

The computational engine is a Constraint Programming (CP) formulation solved with the open‑source CP‑SAT solver. CP was chosen because it handles logical constraints (e.g., “no crossing of major highways”) more naturally than pure integer programming, and it scales well to the problem size (tens of thousands of planning units, dozens of schools, and multiple levels). The model simultaneously optimizes across all three school levels, rather than solving each level separately, which enables the explicit inclusion of feeder‑pattern stability as an objective rather than a hard constraint.

Empirical results show that the algorithm can improve all four objectives simultaneously relative to the district’s existing boundaries. Average student travel distance decreased by roughly 12 %, the SES dissimilarity index fell by about 15 %, feeder‑pattern continuity was maintained for over 90 % of students, and compactness scores improved modestly, indicating more geographically coherent zones. These gains were achieved while respecting all capacity and policy constraints, demonstrating that the perceived trade‑offs among these goals can be mitigated through a well‑designed multi‑objective approach.

Beyond the technical contributions, the paper emphasizes the importance of transparent, open‑source tools for broader adoption. All code, data preprocessing scripts, and the CP‑SAT model are publicly released, allowing other districts to replicate or adapt the framework without reliance on proprietary software. The authors also discuss practical challenges encountered during the partnership: building political will in a polarized environment, navigating legal sensitivities around racial integration, and ensuring that community input is not dominated by vocal minorities.

Limitations are acknowledged. The model assumes that students will attend the school to which they are assigned, ignoring the effects of existing school‑choice programs; extending the framework to incorporate choice behavior is a promising direction for future work. Data quality (address geocoding accuracy, up‑to‑date demographic information) and survey bias remain potential sources of error. Moreover, legal constraints on race‑based assignments may restrict the direct use of racial composition as an objective, necessitating careful framing of SES integration goals.

In conclusion, this study delivers a flexible, community‑informed, multi‑objective optimization framework for school attendance boundary redesign. By jointly optimizing transportation, socioeconomic integration, feeder‑pattern stability, and compactness, and by embedding community‑derived importance weights, the approach offers a pragmatic pathway for districts seeking equitable, efficient, and politically feasible rezoning solutions. The open‑source implementation and detailed field experience provide a valuable template for researchers and practitioners aiming to translate algorithmic insights into real‑world education policy.