PhiladelphiaDistrict vs chartermedium
Philly schools with 1000+ students average 42% PSSA ELA — 17 points above 200-500 schools
Thesis
Philly schools binned by enrollment: 201-500 students (n=58) average 25.4% PSSA ELA, 501-1000 (n=176) average 25.8%, and 1000+ (n=106) average 42.4%. The biggest-school bucket outperforms smaller schools by 17 percentage points. Counterintuitive given the usual 'smaller schools work better' assumption — but here the largest schools include the comprehensive HS that draw the city's best-prepared students, and the larger charter networks.
Supporting findings
- ≤200 students: bucket empty in our PSSA data — these are the smallest specialty schools.
- 201-500: n=58, avg PSSA ELA 25.4%.
- 501-1000: n=176, avg 25.8%. Essentially identical to the 201-500 bucket.
- 1000+: n=106, avg 42.4%. +17pp jump.
- Effect is driven by selection — large schools include selective HS + scaled charter networks, both of which serve different student bodies.
Reporting directions
- Same query but split by district vs charter. Does the size effect persist within sector?
- Within 1000+ bucket: list the top 10 by size, show their PSSA values. Are they selective?
- Compare to NYC's stronger 'smaller schools work better' finding (MDRC's NYC small-schools-of-choice). Different city, different result.
Methodology & replication recipe — the queries + sources behind the findings above.