How D15 unscreening 5 years in: outcomes, equity, both?

District 15 voted in September 2018 to end academic screening for middle-school admission. From the 2019-20 admissions cycle on, the district uses a lottery with 52% of seats prioritized for students from low-income families, students in temporary housing, English Language Learners, and students with disabilities. District 13, immediately to the north, kept screening at most of its middle schools. The districts are comparable: both Brooklyn, similar borough geography, comparable student populations. District 15 makes a useful treatment group and District 13 a useful comparison.

District 15 has eight currently-open middle schools, all in the OPEN admission category — the lottery system. District 13 has nine, with six in the SCREENED category, two OPEN, and one ZONED. The policy difference is visible in the admissions category labels alone.

From 2018-19 (last full pre-reform year for District 15) to 2024-25, District 15 middle schools' median ELA proficiency rose +2.8 percentage points per school; District 13's rose +6.3. Math: District 15 −0.3, District 13 +5.2. The simple read is that District 13 gained more. The complication: District 13 was starting from a much lower base (median ELA around 29% in 2018-19 versus 62% in District 15), and the citywide post-COVID rebound was largest at the schools furthest behind. A district that started near the citywide median had less room to rise than one that started near the bottom.

If the reform integrated schools, we'd expect between-school scores in District 15 to compress — top schools come down a bit, bottom schools come up. The standard deviation of ELA proficiency across District 15 middle schools fell from 22.1 in 2018-19 to 17.4 in 2024-25. But the same compression happened in District 13: 22.5 → 16.9. The narrowing tracks a citywide pattern — schools that scored at the bottom of the state-test distribution caught up more during the post-COVID period — and isn't a clear District 15 signal.

Critics of screening expected the reform to narrow within-district demographic gaps. The opposite is what we see, in both districts. In District 15, the average school's White-student ELA proficiency went from 71% (2018-19) to 83% (2024-25), while Black-student proficiency went 54% → 42%. In District 13, White went 54% → 83%, Black went 37% → 44%. Both districts saw the White-Black gap widen (from about 17 percentage points in 2018-19 to 41 points in District 15 and 39 in District 13 in 2024-25). That's a citywide pattern through the pandemic, not a District 15 effect. The Black-student n at most District 15 schools is small, so the per-school noise is high, but the direction is unambiguous.

The reform's stated goal was demographic integration — the racial and economic composition of each school becoming closer to the district as a whole. Prior reporting (Chalkbeat, The New York Times) has documented that District 15's middle schools did shift in that direction: schools that were heavily white and Asian moved toward district-wide composition. Our per-school demographic-share metrics aren't loaded yet, so we can't independently confirm or quantify the integration outcome from our data. The test-score picture above is consistent with several stories: that integration happened and test scores held roughly steady at the top while the bottom rose elsewhere; that the reform's effect on tests was small either way; or that selection into the district shifted (families who valued screening left the district). Distinguishing these would require the demographic and within-school subgroup-n data we'd need to load next.