Resident Scholar Frederick M. Hess

The No Child Left Behind Act is currently up for reauthorization, giving congressional champions of educational accountability a chance to address a major problem in its design.

When it took effect in 2002, No Child Left Behind required every state to adopt an accountability system with regular testing in reading and math while leaving each state free to decide what material tests should cover, which tests to use, and how to define and measure student "proficiency."

Congress also mandated that schools and districts be evaluated based primarily on how many students are "proficient" in reading and math with consequences for schools and districts that fail to meet state performance goals.

Congress then added a third ingredient--the requirement that 100% of students be proficient in reading and math by 2014. While states initially enjoyed a relatively free hand in determining what percentage of students had to be proficient, they are ratcheting that number up as 2014 draws closer.

Unfortunately, state-by-state flexibility on defining proficiency, coupled with consequences for "failing" schools and a focus on the 100% target, encourages governors and state legislators to boost student scores by adopting easier performance standards, setting lower passing scores, making tests easier or otherwise gaming the system.

Today, we see the results. Oklahoma claims 93% of its fourth-graders are proficient in reading and 86% in math. But the federally administered National Assessment of Educational Progress--the closest thing we have to a national benchmark--finds that just 26% of Oklahoma's fourth-graders are proficient in reading and 29% are proficient in math.

In fact, just 61% scored "basic" or above on the NAEP reading test, meaning one in three children that Oklahoma labels proficient was deemed "below basic" by NAEP standards.

Massachusetts, which led the nation on the latest NAEP, reports that, based on state tests, only 50% of its students are proficient in reading and 40% in math. The reward for having set this demanding bar is that Gov. Deval Patrick must explain to voters why Massachusetts schools seem to be faring only half as well as those in Oklahoma. How many governors standing for re-election will prefer Massachusetts' course to Oklahoma's?

Not surprisingly, many more states look like Oklahoma than like Massachusetts. What's needed is an approach that focuses states on educational quality rather than manipulating outcomes.

The first step is to abandon the grandiose 100% target.

A second is requiring states to calibrate their results against the NAEP in some fashion. A more ambitious approach would entail gradually moving toward rigorous national tests in reading and math.

A third step is encouraging states to adopt sensibly designed "value-added" systems that evaluate schools based upon whether students make reasonable progress during the school year rather than whether their students clear arbitrary hurdles.

An alternative course is to embrace a more modest federal role, let states set their own goals and charge the feds with illuminating how states are doing.

Absent such changes, state reports of rising achievement will be as useful and reliable as those old reports by Soviet commissars, claiming food production had doubled in line with the Kremlin's five-year plan--and never mind those empty store shelves.

Frederick M. Hess is a resident scholar and director of education policy studies at AEI.