Chris O'Donnell linked to this essay about homeschooling. I agree with him that this is an excellent article.
One thing, though, that bothers me is an argument that is always used in these types of essays about America's schools and scores vs. international scores. America's scores on these standardized tests are based on practically the whole population of its students. That means the 4-point+ student and the barely squeaking by at 2.0 are all lumped together in our data. Internationally, though, students are separated at an early age to determine whether they will go to secondary or trade school. Our scores are hurt by reporting the scores of all our students vs the creme de la creme of international students.
I'm not bitter though, I just wish people would temper this argument.