The within-district question
When people talk about "achievement gaps" they usually mean state-to-state or race-to-race. The bigger gaps in American education are inside a single school district, where neighboring elementary or middle schools (funded by the same tax base, run by the same superintendent, governed by the same board) post test-score differences of 40 percentage points or more.
What the data reveals
Within-district gaps of 40+ percentage points between the highest- and lowest-performing schools are common in large urban and suburban districts. Districts where this pattern is especially pronounced include those that operate magnet programs alongside neighborhood schools, districts with substantial attendance-zone heterogeneity (mixing wealthy and low-income neighborhoods), and districts whose newest schools serve growing affluent suburbs while older schools serve declining urban cores.
Why within-district gaps matter more
State-level achievement gaps are politically intractable. Changing them requires legislative action and decades of investment. Within-district gaps sit inside the control of a single school board. They reflect choices about attendance zones, magnet admissions, principal assignment, teacher distribution, and capital investment. They are correctable.
For parents, within-district gaps drive the most consequential school-choice decisions. Which elementary attendance zone to live in within the same town. Which middle-school feeder pattern to lobby for. Whether to apply to the magnet lottery.
The data
The full list of districts with the largest within-district gaps, ranked by the percentage-point difference between their highest- and lowest-performing school on the state assessment, is available below. Each district links to its school directory so you can see the per-school breakdown.
Methodology
For each district with at least four schools that have state assessment data, we computed the proficiency rate (percent of students meeting or exceeding standards in math and reading/ELA combined) for each school. The "within-district gap" is the difference between the highest and lowest such rate. Districts with fewer than four scored schools are excluded as the gap measure is unstable at small N.
