The Brooklyn New School lost 52 points of math proficiency in six years
The question
Did Brooklyn New School really collapse academically, or is its published 52-point drop a testing-opt-out artifact?
Brooklyn New School (P.S. 146 in Carroll Gardens) shows a 52-percentage-point drop in its fifth-grade math proficiency rate: 70% in 2018-19, 18% in 2024-25. The school tests only about 11 of its roughly 100 fifth-graders each year — Carroll Gardens has a long-running parent-led testing opt-out movement, and with 11 students tested, a single child's score moves the published rate by 5 to 10 points. The school's demographics have been stable, teacher survey scores rank among the highest in the city, and chronic absenteeism has been falling. The drop in the published rate doesn't reflect a drop in instruction at the school.
Data analysis
Investigation run 2026-05-30.
TL;DR
The 52-point published 'collapse' is largely a testing-opt-out artifact. Brooklyn New School tested only 11 of its roughly 100 fifth-graders in 2024-25. Two of those 11 scored proficient, which produced the 18.2% figure. The school's teachers reported 100% recommend and 100% principal trust on the survey, and chronic absenteeism has been falling (40% in 2021-22 to 22% in 2024-25). The metric reflects who chose to test, not how well the school is doing.
Year-by-year proficiency at P.S. 146
Fifth-grade math proficient rate at P.S. 146 — 70% (2018-19, 10 students tested), 42% (2021-22, 12 tested), 63% (2022-23, 8 tested), 56% (2023-24, 18 tested), 18% (2024-25, 11 tested). The number of students tested matters here: Brooklyn New School has about 100 fifth-graders enrolled, but tests fewer than 20 of them. A single student's score swings the published proficient rate by 5 to 10 points.
Third-grade math: 60% in 2023-24, 31% in 2024-25. Same pattern — small number tested, big swings.
The third-grade reading score (English Language Arts) moved the opposite way over the same window — 17% to 47%, with 12 and 15 students tested — which is consistent with the swings reflecting sampling noise on top of an opt-out culture, rather than real changes in what students know.
How this compares with the rest of the city
Citywide, the fifth-grade math proficient rate moved from 45% in 2018-19 to 54% in 2024-25 — a gain of about 10 percentage points over the period, most of it from the New York State change in how it scores the test. Brooklyn New School's published numbers moved 52 points in the opposite direction, which is well outside what a real classroom-level decline would produce.
The typical New York City school tested about 58 fifth-graders in 2024-25. Brooklyn New School tested 11. With about 640 students across kindergarten through grade 5, the school's grade-5 cohort isn't unusually small — the small number tested reflects how many families opted in to the test.
What teachers and parents are saying
On the 2023-24 New York City School Survey, every Brooklyn New School teacher who responded said they would recommend the school to a parent, trusted the principal, and would recommend the school to another teacher as a place to work. Teacher trust in colleagues registered at 95%. Those numbers rank among the highest in the city.
The school's enrollment, share of students from low-income households (around 38%), and racial composition (around 45% white) have been stable across the period.
What's behind the published numbers
Carroll Gardens has had a parent community organized against the New York State standardized tests since the early 2010s. Brooklyn New School positions itself around portfolio-based assessment and project-based learning rather than test scores.
The published math proficiency rate is the share of tested students who score 3 or 4 on the state test. When most students opt out of testing, the rate reflects only the small group whose families chose to test — a self-selected sample that isn't representative of the school.
The 18% figure for 2024-25, then, doesn't say much about how Brooklyn New School is doing. It says the state test isn't measuring the school the way it measures schools where most students take the test. Whether state testing should be the primary school-quality measure is a separate debate.
Caveats
We don't have student-level data, so we can't directly compute opt-out rates. The case here is inferred from three things: the small number of tested students relative to the citywide median, the stability of enrollment and demographics, and teacher survey scores that don't fit a school in academic decline. A reporter could confirm directly by filing a Freedom of Information Law request for test-registration data per school per year.
What's next
Worth a feature on the broader New York City opt-out picture. Brooklyn New School isn't alone — Manhattan's Earth School, P.S. 87 in some years, and parts of the small-progressive-schools network show similar tested-cohort patterns. Story angle: some of New York City's lowest-published-scoring elementary schools are also the schools where families have most deliberately stepped away from the test.
Methodology & replication recipe — the data sources, queries, and steps behind this analysis.
More on specific schools
- Six NYC schools where 100% of teachers say students bully each other
- Bronx Aerospace High School cut chronic absenteeism by 34 points. How?
- The Anderson School scores at the ceiling — and proves nothing
- Two MS within a half-mile: one screens, one zones. Same neighborhood, different futures.
- When the principal leaves, scores fall — but not always