K-8 schools weathered COVID better than separate ES + MS — why?

Standalone elementaries gained 4 more points than K-8 schools over the same window. This is the opposite of what the original thesis predicted. Caveat: much of the citywide gain is the 2024-25 test recalibration; if K-8 schools had less ceiling room (or were more aggressive about opt-out), they'd capture less of the recalibration bump. Both standalone elementaries and K-8 started at nearly identical baselines (~45%), so that's not the explanation.

HS had the best attendance recovery — older students with more autonomy actually came back more reliably than the younger kids whose parents control attendance. K-8 schools have the WORST attendance recovery — exactly opposite to the original hypothesis. Standalone elementaries did better. This is consistent with K-8s having larger, more disrupted student bodies that include the middle-school years (where attendance norms collapse during the pandemic-affected cohort).

The original thesis was: "K-8 schools keep kids together longer, building stronger attachment, which buffers against pandemic disruption." That predicts K-8 should have smaller dips and faster recovery.

The data shows the opposite. Two plausible explanations:

1. Larger school = more bureaucracy. K-8 schools are bigger (because they serve more grades). Bigger schools mean weaker personal teacher-student bonds, less individualized re-engagement after disruption. Standalone elementaries are typically smaller and more nimble.

2. Middle-grade dynamics dominate. K-8 schools include grades 6-8 (puberty, peer-network shifts, the standard "middle school cliff"). Those grades had the worst pandemic disruption. Including them in the K-8 cohort drags down the school's average recovery.

The original framing was wrong because it implicitly assumed K-8 schools are 'elementary plus.' Operationally they're 'elementary plus middle school,' and middle school is the harder part to operate.

One nuance: within K-8 schools, grade-5 cohort outcomes might still be better than what those kids would have gotten if they'd transitioned to a standalone middle schools in grade 6 and then taken grade-7 tests at the new school. This is the "transitions are bad" version of the hypothesis.

To test it: compare grade-7 math at K-8 schools vs standalone middle schools schools (where kids enrolled in grade 6 from elsewhere). We don't have student-level data, but a school-level comparison would work as a first cut.

The data above measures grade-5 outcomes — those kids didn't transition. A real K-8-vs-transition story would measure grade-7 in standalone middle schools vs grade-7 in K-8.