In dire circumstances — a battlefield, a natural disaster or an overcrowded emergency room — we accept the rationing of scarce resources as a necessary, if regrettable, choice. We triage. We divide patients into three groups: the safe cases, those suitable for treatment and the hopeless. And we ration resources in an effort to do the most good for the largest number.

But there are areas where we have rejected the idea of triage. Public education, an institution charged with disbursing equality of opportunity for all children, is one of them. In our loftiest moments, we see public education as one place where we dispense with the blunt, utilitarian logic of triage and seek equal treatment for all. But try as we might, deep inequalities persist and belie our rhetoric.

It's ironic that the No Child Left Behind Act, intended to right the injustices suffered by poor and minority children, has caused more rationing of education. Five years ago, this law made accountability our nation's educational blueprint. Schools must increase passing rates on annual tests so that all students are proficient in reading and math by 2014. The idea is that what gets measured gets done. If educators are not held to task for helping the most disadvantaged students, their attention will remain elsewhere.

The past five years have proved the law's framers right beyond anything they could have imagined. The problem is a classic case of misaligned incentives. No Child evaluates schools by the percentage of students passing state tests. Imagine that students must answer 70 percent of the questions correctly to pass. Schools get no credit for moving a student from a 15 to a 69, or from a 70 to a 95. Yet, if educators nudge a student from a 69 to a 71, the school's passing rate increases.

The stakes for schools are enormous. So it isn't surprising that many educators game the system by reaching first for the low-hanging fruit, the students closest to passing. Dubbed the "bubble kids," because their scores put them on the bubble of the passing mark, these students give schools the biggest bang for the buck. In response to this incentive, many schools have rationed out practically all of their resources to these students. Meanwhile, the lowest-performing students, the "hopeless cases," languish. So do their high-performing classmates, who are relegated to the waiting room while the bubble kids are cured.

It would be easy to question the ethics of educators engaged in triage, but they are doing exactly what the No Child Left Behind Act asks them to do. Policymakers, not teachers, must be held accountable for implementing a policy that rewards schools for privileging some students at the expense of others.

The solution is to tweak incentives to encourage educators to improve the achievement of all students. Congress will have the opportunity to revisit this issue as the act comes up for reauthorization early next year. Unfortunately, Education Secretary Margaret Spellings recently scoffed at the idea of making significant changes.

We need to look closely at the act's impact on the lives of real kids and educators. Both the co-author of "The Bell Curve," Charles Murray, and the Civil Rights Project at Harvard University have argued that the losers in a system based on passing percentages are minority students. It is an approach that not only creates perverse incentives to focus on students close to passing but also underestimates the size of the achievement gap, creating the illusion of progress where none exists.

Numerous studies from the trenches reveal the painful compromises teachers make to maximize passing rates. The research substantiates the proposition that unrelenting attention to passing rates turns educators' attention away from the law's intended beneficiaries, the lowest-performing students, and hurts high-performing students as well.

One policy change that would go a long way toward addressing this problem is measuring educational growth. When students improve on their previous performance but don't clear the passing threshold, schools still deserve credit. We also need incentives for educators to further improve the performance of students already passing state tests. As it stands, they have no reason to invest in average and high-performing students.

If we don't get the incentives right this time, we face five more years of lost opportunities for America's children.

Jennifer Booher-Jennings is a doctoral candidate in sociology at Columbia University. She wrote this commentary for the Washington Post.