Slow on the uptake again, but here's Class Size Matters' response to last week's NY Times article on charters in Harlem. The response points to some data on the supposed effects on class size of an existing public school when a charter school moves in. (Via Chalkboard.)
I just want to point out that in at least one case, these data are a little suspect. One school, listed as having opened in 2000, is compared to its co-locating school from school year 1999-2000 on. That school, however, did not move into that building until school year 2004-05. Comparing school year 2003-04 (pre-move) to 05-06, in only two grades did average class size go up at the non-charter public school, one of which went up by only half a student. In all other cases class size went down.
The point is, these charts are weird and don't account for all the many reasons class size can go up or down in a school.