These pieces originally appeared as a weekly column entitled “Lessons” in The New York Times between 1999 and 2003.
[ THIS ARTICLE FIRST APPEARED IN THE NEW YORK TIMES ON DECEMBER 19, 2001 ]
The education bill: Many trials ahead
The education bill written largely by the White House and given final Congressional approval by the Senate yesterday is a breathtaking intrusion of the federal government on states’ control of education.
For two decades, policy makers have wanted to make schools more accountable. Some states have developed creative ways of using tests to tackle that complex challenge; others have weaker systems. But the new legislation seems to impose on every state a Texas- type model that may actually retard sensible accountability in some places.
The bill requires every child to be tested in math and reading each year in Grades 3 to 8, and specifies state intervention in any school where children as a whole, and within each demographic group, do not demonstrate proficiency or make annual progress toward it.
Texas has had some success with that policy: more low-income minority children acquire basic math and reading skills there than before. But the Texas model also has flaws. There is some evidence that dropout rates have risen because students must meet higher standards without adequate preparation. Further, focus on basic skills has reduced teaching of analytic reasoning and interpretation, and Texas is now adding more complex questions to its tests to try to fix this.
Other states have tried to avoid such pitfalls, but some promising alternatives may not survive without bending the new federal rules.
Take, for example, the new federal requirement that each student get a separate score annually. Maryland leaders realized some time ago that individual scores on a single exam could be misleading, because one test cannot cover all the topics in the curriculum. And testing can be harmful if teachers emphasize to students only what is most likely to be on that single, narrowly drawn exam.
So in Maryland, different students are tested on different parts of the curriculum. Individual scores can’t be compared, because some students get harder questions than others. Rather, the state calculates statistically reliable schoolwide scores by combining answers from all students in a grade.
North Carolina, meanwhile, concluded that comparing students this year with those in the same grade last year may be pointless. Even in a school where instruction is superior, scores can be lower if students are less able than those the previous year. Punishing a school for such a decline could set back the cause of improvement.
So North Carolina holds schools accountable for growth of the same students: from last year’s third grade, for example, to this year’s fourth. This is more sophisticated than the Texas plan, but may not be permissible under the new federal rules.
In Kentucky, the state demands progress not only in reading and math but also in science, social studies, the arts and health education. Different subjects are tested in different years. Schools are accountable for improvement not only in test scores but also in student work samples and other indicators, like whether graduates go on to college. Under the new legislation, Kentucky will have to add what its policy makers consider needless extra tests in reading and math.
The federal bill permits a state to define its own proficiency standards, provided the same standard applies to all students in a grade. It also limits the time for achieving those standards to 12 years, a provision that makes most sense when, as in Texas, the standard itself is low and focuses mostly on basic skills. California, however, has set a standard equivalent to university eligibility requirements. Because this is so much higher than standards in most other states, California requires schools to get needy students up to only 80 percent of the goal, and each year schools must narrow the gap between present achievement and the target by 5 percent.
At that rate, California schools would not nearly eliminate the gap in 12 years. The state would be able to comply with the federal timetable only by lowering expectations for achievement. Reducing unrealistic standards may be wise, but we can’t be sure of its wisdom if the government prohibits such experimentation.
Some state systems may be better than the new federal prescription. Some may be worse, so the federal rules could be beneficial. But until now, Congress assumed that accountability was best left to states. That way, superior systems might be copied elsewhere; inferior ones could be abandoned without imposing the mistake on the entire nation. State educators with authority to fine-tune systems could take pride in their work, leading to better results.
Secretary of Education Rod Paige could preserve state flexibility by issuing interpretive regulations that stretch the legislation’s literal provisions. Otherwise, some states may ignore the new law or beseech the government for waivers. Either way, the story of the president’s education program is not yet over.