Response to "Alabama using new formula to measure dropout rate", AP wire on T&G online, http://www.telegram.com/article/20100225/APN/302259516
Alabama's "on-time graduation rate" is a better standard for schools to measure their success by. But the problem with benchmarks and standards is that they can distort decision-making and have unintended consequences. Measuring schools' performances honestly and working at improving them is good and necessary. The problem is that setting rigid limited goals and then trying to terrorize the teachers into meeting them doesn't really make them better.
Right now under NCLB the schools have a limited set of "capital goals" to meet, and failing to meet them can result in "capital punishment" - firing all the teachers or closing the school. Graduation rate is now one of those goals. The Alabama definition is more honest than the one we are currently using, but because it is stricter and more limited it will result in more distortion, and more decisions that are bad for the children and the schools.
Take another example of this problem. Schools are struggling to meet annual yearly progress on math and English scores, with the survival of the school and the teachers' jobs at stake. They often throw history, languages, music, shop and gym under the bus to get more time and staff focused on math and English. This ends up turning the school experience into a deadly bore for many students. Bored students make trouble, don't pay attention and skip school or class more often, so in the end putting too much time and pressure on math and English can make things worse.
The danger with adopting the Alabama graduation-rate measurement is that when inevitably some students fall a year behind - for whatever reason - then if the school's survival depends only on the "on-time graduation rate", it no longer has an incentive to get the delayed students through anyway, and it will tend to lose interest in working with those students.