At the same time, I understand the logic behind it. The law was enacted in 2002. The proficient targets haven't been set particularly high. Any goal less than 100% would have been tantamount to giving up on some percentage of that year's kindergarten class, to looking the parent of a five year-old in the eye and saying "Twelve years just isn't enough time to teach your child to read and do math at a relatively low level." Settling on a goal lower than 100% also raises the question of how much lower and which children, exactly, would be written off from the get-go. Given the historical relationship between race, class, and educational opportunity in America, it's not hard to guess whose children those would have been. 

This raises the question of what would happen if the federal government didn't mandate a uniform timeline for all states. "Why not let the laboratories of democracy take a shot at it?" one might ask. "They're good people; even if they made different decisions, surely they wouldn't game the system in an utterly fraudulent, cynical way...."

Except we know they would. We know this because, as Anna Habash explains in a new report from the Education Trust, that's exactly what many of them have already done when given flexibility to decide how to hold high schools accountable for graduation rates under NCLB. 

Here's how it works: States are required to hold high schools accountable for graduation. But unlike with achievement scores, where every school has to meet a proficiency target that starts with where the 20th percentile school was in 2002 and ends with 100% in 2014, states can pick whatever end-goal graduation rate target they like and use whatever improvement trajectory they choose. 

That provides two different, exciting ways to render the provision completely meaningless. One is to pick an absurdly low goal. Welcome to the state of Nevada, where all manner of sins are legal and the statewide high school graduation rate goal is 50 percent. In other words, as long as your odds of graduating are better than what you get when you slap down $20 on red at the roulette wheel, you're doing fine. Alaska chose 55.58%, because apparently 56% even was just too heavy a cross to bear. And so on. 

Other states have been a little more clever. Instead of setting the bar at knee height, they adopt a putatively high bar but give schools centuries to get there. Maryland, for example, theoretically has a 90% graduation rate goal. But it will accept any improvement as sufficient progress, even 0.01%. At that rate, the state's African-American students will all be graduating by the year 3117, by which time we'll all be cursing in Mandarin Serenity-style and learning will take place via coaxial cables jammed into the back of your head. 

The point being, this isn't as simple as it seems. To be effective, accountability policies need to challenge schools to achieve more than they may believe they can achieve, but not challenge them so much that the entire process loses legitimacy. The policies also need to balance realism about the potential to improve the system as it stands with the moral urgency of helping students who are enrolled right now and can't wait to grow up while we fiddle with the only schools they have. 

One way way get around the seemingly insoluble 100% problem is to change the back-end responses and incentives as schools approach the final steep climb to the top and pass various points of diminishing returns: the closer you get to 100, the more the consequences of failure become a matter of foregone positive incentives rather than mandatory negative ones. In any case, it's a tricky set of issues for the next Congress and President to address.