After eight years of making education systems use a one-size-fits-all, simplistically linear model for judging a school's performance, known as Adequate Yearly Progress (AYP), the federal government is searching for a new form of accountability. The Obama administration recently said it wants to ditch the AYP provision in the No Child Left Behind law, which judges annually whether a school is succeeding or failing based mostly on achieving pre-set targets for student test scores. AYP has been the target of endless complaint and opposition. According to a New York Times report, the administration concluded the AYP system "fails to differentiate among chaotic schools in chronic failure, schools that are helping low-scoring students improve and high-performing suburban schools that nonetheless appear to be neglecting some low-scoring students. Instead, under the administration’s proposals, a new accountability system would divide schools into more categories, offering recognition to those that are succeeding and providing large new amounts of money to help improve or close failing schools."
That's good, because AYP was grossly inadequate. But solving the problems of performance accountability in education won't be easy. Imagine that you are a charter school authorizer with a portfolio of, say, 50 schools, or the superintendent of a school district with a similar number of schools.
- To what performance standards should you hold schools accountable?
- Should all schools be held to the same standards?
- What should happen when a school fails to meet the standards?
- And since even a failing school has some students, teachers, classes that are meeting the standards--"islands of success"--what should happen to them?
Let's tackle just the first two of these questions to get a sense of the complexities involved in designing a new performance measurement system for education.
Which performance standards? States are beginning to converge on a set of college- and career-readiness standards that mostly will involve test scores, perhaps on the ACT sequence of three aligned tests beginning in 8th grade. So let's use those. With trepidation--since right now perhaps only a third of high school graduates actually achieve those standards.
Add other standards if you agree with people who argue that test scores in subject matters simply doesn't measure enough of a student's college readiness. Add even more if you think (I do) that rates of high school graduation, enrollment in post-secondary programs, and college completion also are part of a school's performance picture.
What about elementary and middle schools--how will you measure their performance? If a middle school is supposed to prepare students for success in high school, how is high school readiness measured? If an elementary school is part of a three-school feeder system, can it be held accountable for whether its students do or don't go to college seven years after they leave fifth grade?
Finally, recall that schools increasingly exist in a competitive "market" environment. Unsatisfied parents can move their kids to other schools--or home school with a link for a cyber school--at the drop of a textbook. So it seems important to measure parent and student satisfaction with the school--and this won't necessarily hinge on the school's academic performance. Parents and students care about many other characteristics of a school.
Universal or variable standards? Should an urban school system with 70% of its minority, immigrant students living in low-income households be held to the same performance standards as a suburban school system with 90% white students living in middle- and upper-middle class households? Consider the track records of these two schools. The city schools graduate 40-50% of their students, maybe 10% are college ready, and maybe 20-25% enroll in post-secondary programs. The suburban schools graduate 80-90%, 30-40% are college ready, and 60-70% enroll in post-secondary programs.
With this starting point for measuring performance, should we say that a 90% graduation rate is the standard for both systems? If not, what is the lower standard for the low-income school? And what's the explanation--e.g., "poor kids can't learn"--for holding a lower standard?
But if all schools have the same performance standards it seems obvious that even if the low-performing school improves its performance substantially, it will take a number of years to achieve the same performance that the suburban school already achieves. How many years should it take? How can you tell if it is making enough progress fast enough? How can you even tell if the school has a good plan and enough capability to make progress?
| 
