In 1998, Major League Baseball burst out of its post-strike slump when fans were captivated by the record-setting home run battle between Mark McGwire and Sammy Sosa. Over the next few years, a home run-hitting frenzy, led by Barry Bonds, drove baseball to new heights of popularity.
We now know that those homers are tainted by steroid use, but at the time, no one cared. Fans loved the game again, and owners loved the record ticket sales. They weren’t thinking about the long-term damage to the sport from depending on artificial enhancement.
American education is in a similar position today. According to state standardized tests, students are doing great. Parents, school officials and politicians all want high test scores, so few people question what’s making them high. But like baseball’s astounding home run production, the numbers are artificially inflated.
The National Assessment of Educational Progress, the well-respected national test, tells a different story from the state tests. For example, according to Wisconsin’s test, 85 percent of eighth-graders were proficient in reading in 2005, but only 35 percent scored high enough to be considered proficient by the NAEP. Such large discrepancies are common in many states.
According to the Fordham Foundation, almost 20 states reported gains from 2003 to 2005 in the percentage of eighth-graders passing their own state reading tests. None of these states, however, showed any improvement in eighth-grade reading at NAEP’s “proficient” level. At the fourth grade level, the story was similar.
This is not to say that the NAEP is the perfect test, only that we should be concerned by the large discrepancies. We are not interested in high scores per se; we are interested in high scores that reflect high levels of knowledge. When students can do extremely well on state tests but not the NAEP, it suggests that the states are fudging the numbers.
We shouldn’t be surprised. The No Child Left Behind Act tells states to make annual progress in student achievement, but for the most part, leaves them alone to design their own measuring sticks. And creating a grading scheme that pumps up scores — by lowering the minimum passing score or watering down the difficulty, for example — is a lot easier than pumping up students’ knowledge.
It’s tough for a state to set stricter standards. Identifying more students as “failing” draws lots of negative attention; inflating the numbers doesn’t. Since this dynamic makes it unlikely that states will toughen up, the federal government needs to set national standards to put an end to the shenanigans.
National standards would set the bar, and then let the states do whatever they need todo to clear it. As some analysts have described it, NCLB is too tight on means and too loose on ends. The federal government mandates specific requirements for hiring teachers, adopting curricula, calculating yearly progress and remedying poor performance. But it says little about how students should be tested, what should be on the test, and what should constitute a passing score.
Federal education policy needs to be reversed. The ends should be ironclad, but the means flexible. The federal government should be clear and firm about what states need to achieve, but let them figure out how to reach that goal.
Critics often argue that national standards violate our cherished principle of federalism by allowing the national government to dictate educational policy to the states, but a new combination of flexibility and accountability could actually facilitate work in the “laboratories of democracy.”
It would follow the model of charter schools, which are held accountable for results but given considerable flexibility to achieve them. They have used that freedom to experiment with relaxed teacher qualifications, Montessori programs, and arts-based curricula, for example. With greater freedom of means, states could also become more innovative.
Furthermore, a common achievement standard could help states share ideas. Ideally, we would like states to experiment, and then copy good ideas from each other. But how can a state know which policies are working if every state measures success differently? Thanks to federal definitions of poverty and murder, states can effectively compare welfare and crime policies. But in education, there are 50 different measures of educational achievement. A federal achievement measure would make cross-state comparisons easier.
After the steroid use came to light, Major League Baseball implemented a league-wide anti-drug policy. Now that there is growing recognition of pumped-up state test scores, the federal government needs to take similarly decisive action. Next year’s NCLB reauthorization provides the perfect opportunity to scrutinize how the current set-up encourages gamesmanship at the expense of real learning, and how national standards combined with greater flexibility in means could ameliorate the problem.
Kathryn Newmark is a graduate student in public policy at Georgetown University
