Private and charter schools appear to have significant but modest effects on test scores but much larger effects on educational attainment and even on long-run earnings. A new working paper from Booker, Sass, Gill and Zimmer and associated brief from Mathematica Policy Research finds that charter schools raise high school graduation, college enrollment and college persistence rates by ~7 to 13%. Moreover, the income of former charter school students when measured at 23-25 years old is 12.7% higher than similar students. Similar in this context is measured by students who were in charter schools in grade 8 but who then switched to a traditional high school–in many ways this is a conservative comparison group since any non-random switchers would presumably switch to a better school (other controls are also included).
The effect of charters on graduation rates is consistent with a larger literature finding that Catholic schools increase graduation rates (e.g. here and here). I am also not surprised that charters increase earnings but the earnings gain is surprisingly large; especially so when we consider that the gain appears just as large among charter and non-charter students both of whom attended college (i.e. the gain is not just through the college attendance effect).
I wouldn’t bet on the size of the earnings effect just yet but what we are learning from this and related research, such as Chetty et al. on teachers, is that better schools and better teachers appear to have a significant and beneficial long-run impact that is not fully captured by higher test scores.
As I said in Launching, one of the factors that makes me optimistic about education in the United States is that it remains relatively decentralized and open to experimentation and evolution.