School reform pays off

Education reform, the mantra of the 1990s, is beginning to take effect in the classrooms of America.

From North Carolina to Texas, states long seen as laggards on reform are now seeing gains in student achievement twice the national average. And nationwide, math scores are rising significantly faster than they had during the previous two decades, with marked improvement among Hispanic and black students.

None of this progress is obvious from national test scores, which have often been declining or flat for decades. Nor is it clear from international comparisons, where American 12th graders still rank near the bottom of the world in math and science.

But the view from the states opens a much more positive window on American public education. And while the progress can vary greatly from state to state, the overall picture is improving. Indeed, new evidence challenges the view that public schools can't reform - or that only students from the right kind of family can hope to succeed.

A new study, released yesterday by the RAND Corp., takes a look at state results from the National Assessment of Educational Progress (NAEP), a voluntary national test funded by Congress. The NAEP test is most often described as the nation's report card. RAND researchers, however, caution against looking only at national averages, saying that obscures significant differences among states. And some state policies are clearly more effective than others in improving student achievement.

Some key findings:

The math scores are rising at a rate of about one percentage point per year. But gains in some states are twice the national average, while others showed none at all.

Those differences are even more dramatic when you compare how states are educating their most disadvantaged students. Texas and California have a comparable mix of poor, black, and Hispanic students, but Texas students score 11 percentage points higher on NAEP math and reading tests than their California counterparts.

Here's what works: lowering class size, expanding prekindergarten programs, and making sure teachers have the resources they need. Also, setting clear standards and holding schools accountable for meeting them.

What doesn't work? Just raising teacher salaries or increasing the number of teachers with advanced degrees.

The RAND report comes at a critical moment in this latest reform cycle. The new evidence it produces could help shore up faltering reform movements.

In states such as Massachusetts and Virginia, where standards and testing regimes are well established, there's a backlash building from parents, students, and teachers groups who argue that high-stakes tests are stressing out students and the system. In conventions this month, the leading teachers unions both warned that the standards movement was making unfair demands on teachers.

At the same time, latecomers have committed huge amounts of money to reform. California, for instance, has set aside $4 billion to reduce class sizes - setting off a scramble for classroom space and qualified teachers. But it has yet to see clear results. In fact, California was singled out for its "dismal" record by RAND, although the report only covers test results up to 1996.

"Data confirming that accountability helps poor kids is a vital message," says Douglas Carnine, director of the National Center to Improve the Tools of Educators at the University of Oregon in Eugene.

"There will be clear signs of progress in California by the next [NAEP] test," he says. "The programs and standards there are still new. What's important is that people get used to thinking about what kids are learning - and not learning - and doing something about it."

For critics of public education, though, it's the ability to allocate resources effectively that is most in doubt.

"Given the current incentives in public education, there is no presumption that money will be spent well," says Eric Hanushek, senior fellow at the Hoover Institution at Stanford University in Palo Alto, Calif. "Nobody in the system has a job that rises or falls on student performance."

"As pointed out by this report, the Texas accountability system and other states doing similar kinds of things have the potential to use that information very productively," he adds.

For example, the report points out that some 100,000 master's degrees in education are awarded to US teachers every year, at a cost to teachers and school districts of about $2 billion. It amounts to "one of the least-efficient expenditures in education," because the degree seems to have "little effect" on a teacher's ability to raise achievement, the report says.

What does make a difference is whether teachers feel they have the resources they need to do their job, it adds.

"Research shows that teacher certification and master's degrees don't have much bearing on student achievement," says Chester Finn Jr., president of the Thomas B. Fordham Foundation in Dayton, Ohio. "But a teacher's basic knowledge and intelligence does."

Yet in a rebuke to reform plans linking teacher pay to student achievement, delegates to the National Education Association's annual meeting rebuffed a plan that would link teacher pay to anything but educational degrees and seniority.

At the outset in the 1970s, the NAEP test was intended to sample national results. At the time, nearly every state claimed to be at or above average on some test. Without a common test, there was no basis for evaluating those claims. As early as 1985, poor Southern states urged expanding the NAEP test to publish results by state, so that educators could have a better measure of progress.

"We knew that Southern states would be behind. But until state NAEP results, we didn't have a legitimate way to measure improvement," says Mark Musick, chairman of the federal governing board that sets policy for the NAEP and president of the Southern Regional Education Board.

"The most powerful part of the [RAND] study is that it shows that what schools do makes a difference," he adds. "Results are not preordained. Students from lower socioeconomic backgrounds do not have to have lower test scores."

(c) Copyright 2000. The Christian Science Publishing Society