Inputs and Outputs
http://www.tpmcafe.com/story/2006/1/29/12463/4986
By Matthew Yglesias | bio

Here's a study (full study (pdf), NYT summary, relevant blog post) showing pretty comprehensively what a lot of earlier research has indicated -- private school students do better than public school students in America. They do better, that is, until you introduce some fairly basic demographic controls at which point they often do worse and certainly don't show any systemic superiority. The anti-vouchers point here is sufficiently obvious that I won't belabor it.

There's a broader lesson/problem here for almost all efforts at educational reform, namely that study after study of education at almost all levels tends to point in this direction -- educational outcomes vary widely, but almost all of this is accounted for by the differences in inputs. Socioeconomic variables account for a huge proportion of the variance in outcomes, not only between whole schools (i.e., schools full of poor minority students do worse than schools full of middle class white ones) but also within schools (i.e., poor minority students tend to do worse even if they attend a school mostly full of middle class white kids).

Roughly speaking, as a society we've given our schools the task of trying to ameliorate America's huge levels of inequality, but education simply seems to be inadquate to the task. Nobody knows a reliable method of making schools that are capable of overcoming disadvantages that kids from certain families face ex ante. One thing that does seem to work pretty well is having the government offer bribes to "at risk" kids to do better in school. This tends to make people queasy for a variety of reasons, but the empirical evidence suggests it works and I think we should try it.

True confession time: I "bribed" my middle class son by telling him that his "job" was to do well in school and by paying him X amount for an "A" and X amount for a "B". He was not paid, or deductions were made, for the few "Cs" he ever received. It worked like a charm.
(I intensely disliked doing it but it was the only way to get this bright child to study.) The result:
he graduated from the International Baccalaureate Program with a GPA high enough to get him full
scholarship money to any state university. He was also automatically placed within the Honors Program at the university he chose. He is now mature enough to realize that education is of value and his first semester grades in college were straight "A's". It works, in my experience, to
"bribe" kids.

I'd much rather have parents doing this than have kids working after school. Unfortunately, we live in a 2+ income society now, and teenagers are an important part of the local workforce.