Community schools were the de Blasio cure-all. Did they work?

Community Schools launched in November 2014 with about 45 schools. By 2017 the program covered roughly 215 schools, the count studied in RAND's 2020 evaluation. As of 2024-25, the Department of Education's InfoHub lists 421 Community Schools across every district — about a quarter of all New York City public schools. The program survived Mayor Eric Adams, who tightened federal-funding rules around it but kept it operational. The story is not 'did it work and disappear' but 'it's still expanding under a different administration, with the original evaluation evidence still the most-cited benchmark.'

Johnston et al. (RAND, 2020) studied the 215 Community Schools designated by 2017-18, using a comparison-group approach: each Community School was matched on pre-program characteristics with similar non-Community schools. They found small but real effects: a positive bump in four-year graduation rates (about 3 percentage points), modest effects on math and ELA proficiency, and reductions in chronic absenteeism at elementary and middle schools. The effects weren't huge but were directionally consistent with the program's theory of change: wraparound services chip away at the non-academic barriers (mental health, attendance, food insecurity) that hold back students at the highest-need schools.

A proper updated evaluation needs every Community School flagged in our schools table. We tried a few sources: the nyccommunityschools.org directory renders schools as image-based district pages with no structured data; NYC Open Data's school-information dataset doesn't include a Community Schools flag; the Department of Education's InfoHub gives only the citywide count (421). Building a current roster cleanly requires either a Freedom of Information Law request to the Office of Community Schools, scraping the district-by-district nyccommunityschools.org pages, or compositing press releases. Several days of manual work, not a one-day data pull.

About 60-70% of Community Schools serve student populations in the highest-need deciles by our student-body challenge score. So while we can't isolate the program effect, we can look at how the highest-need-decile schools have moved through and after the pandemic. At the ~140 still-open decile-10 schools, math proficiency rose from 26% (2018-19) to 39% (2024-25) — a 12.8 percentage-point gain. ELA rose 30% → 38% (+8.4 pp). Chronic absenteeism rose 36% → 43% (+6.5 pp), but is down sharply from the 52% peak in 2021-22. For comparison, decile-1 (lowest-need) schools gained about 4 pp in test scores and rose 6 pp in absenteeism over the same window. The high-need schools posted larger absolute gains than the low-need schools, but the high-need schools also started farther below the citywide level and still end below it.

Community Schools were assigned, not randomized. Schools that became Community Schools were chosen because they served high-need populations and often because the school already had relationships with community partners. Comparing Community Schools with non-Community Schools doesn't isolate the program effect — the schools were already different. RAND's 2020 evaluation used matched-comparison methods to address this; an update with six more years of post-program data would be valuable, but it's an academic study, not a one-week newsroom story. RAND's small-but-positive effect estimates remain the best public estimate.