I appreciate Jay Mathews' extensive reply, and I think it's best to leave the post-and-riposte at that. People can get a fair idea of the thinking behind the charts, and of the main criticisms, from our respective messages. (Click here for my original piece and here for Mathews' response.) But let me take this chance to mention three things I wish I'd been clearer about the first time around:
- If we are talking about "best" high schools, obviously we should have more in mind than sheer academic attainment. Preparation for a happy and balanced life involves classroom skills but other things too: experience with a broad range of people; opportunities to experiment in art or music or drama; physical fitness; learning about competition and teamwork; handling disappointment; romance; you name it. I went to a very large, American Graffiti-style public high school in Southern California that sent only about half its graduates to college. This school doesn't come close to the current "Top 472" list and wouldn't have done so when I went there, but I have always thought it "best" for me because of the breadth of life it contained. I should have been clearer in saying that even if we consider high schools strictly on academic quality, measuring the numbers of AP tests taken is a weird way to do so.
- About homogeneity: The correlation between a prosperous school district and a high rate of taking AP tests is obviously not perfect. Some comparatively less-privileged districts could decide to specialize in these tests--or, like the No. 1 school on the ranking, could become IB academies. At the other end of the scale, some of the richest districts might have artificially low AP-test rates. Their students, with multiple AP courses under their belts and with Stanford or Princeton in their futures, might decide not to take the full battery of AP tests they were theoretically ready for. Enough tests would let them enter college as sophomores--but they might think: We can afford all four years, why cut it short?
But with these exceptions noted, the general principle still applies: A homogeneous student body will do better on per capita calculations, and a homogeneously rich one will do better than one that's homogeneously poor.
- Now, about "unintended consequences." I know that Jay Mathews' hope for this project is just what he says it is: that it will ultimately lead to better schooling for more children. I disagree with him about whether it will do this--but that's an honest disagreement, and we'll see how it plays out over the years.
But there would be one indisputable effect of judging schools on test-taking, and therefore encouraging more students to take many more tests. It would be a boon to the already-surging test-prep business. The Stanley H. Kaplan test-prep chain is one major player in this field (the other being Princeton Review). Kaplan has recently announced a deal with Vulcan Ventures to provide online preparation for AP tests--the tests Newsweek now equates with high-school quality.
As it happens, the Kaplan chain is owned by the Washington Post Company, which also owns Newsweek. (Click here for information on a joint project by Newsweek and Kaplan.) Jay Mathews is so independent minded that he would never hesitate to criticize these tests, or commercial test prep, if he thought that an emphasis on them was misguided. (Journalism is full of catty, faint-praise type compliments, so I should make clear: I'm not being catty. Mathews' motive is clearly not log-rolling for the company.) Nonetheless, it is a fact that among the beneficiaries of Newsweek's ranking system will be Newsweek's corporate parent.