Defining 'Failure' Critical To Bush Testing Plan
President Bush has pledged to get tough on schools that are failing to educate their students. But how do you determine when schools are "failing"?
As Congress discovered this month, coming up with that answer is no easy task.
Last week, a bipartisan compromise reached by senators and the White House after extensive negotiations drew stinging criticism from one advocacy group, which charged that it was "fatally flawed."
The Washington-based Education Trust concluded that the composite formula the Senate has proposed to identify failing schools could actually mask achievement gaps between poor and better-off students—the very gaps that Mr. Bush has promised to close.
"Moreover," read a statement issued by the organization, which works to promote high achievement for all students, "its sheer complexity will undermine the message of high achievement for all."
Lawmakers and members of the Bush administration agreed that the compromise language in the Senate bill, S 1, for identifying subpar schools is complicated. But they argued that it's still more rigorous than existing law.
"We need to both have noble, high goals that mean something and, yet, at the same time be realistic about the kind of progress that we can get out of the system," said Sandy Kress, a Bush education adviser. "Because if we don't really have both of those features in this legislation, folks in the field won't take it seriously."
The bill is currently under consideration on the Senate floor, and debate is expected to continue at least through this week. The House Education and the Workforce Committee, meanwhile, approved its version of the legislation, HR 1, last week. The more stringent requirements in the House bill would give states two options for how they identify schools needing improvement.
Congress is debating the measures for holding schools accountable for improving student achievement as part of the reauthorization of the Elementary and Secondary Education Act, the main federal law for K-12 education. President Bush has made annual reading and math tests in grades 3-8 and a stricter accountability system a centerpiece of his education agenda.
At issue is how states identify schools and districts as "needing improvement" because they have not made "adequate yearly progress" under the Title I program for disadvantaged students.
Under the proposed legislation in both chambers, failure to make such progress would trigger a range of interventions for Title I schools, from providing extra technical assistance; to using public funds for students to transfer to other public schools, or to buy private tutoring services; to closing and reopening the schools as charter schools. President Bush has proposed going beyond existing Title I law, by requiring schools and districts to make annual progress not only for their total student populations but also for subgroups of their students, such as poor or minority youngsters. States, districts, and schools that failed to make progress would face losing federal aid for education.
Both the House and Senate bills go one step further. They set a specific timeline—10 years in the Senate, 12 years in the House—for having each subgroup of students meet the "proficient" level on state tests: economically disadvantaged students, students with disabilities, students with limited English proficiency, and students by racial and ethnic group.
But some lawmakers and the White House became concerned the original language in the Senate bill could have required schools to meet specific numerical improvement targets for each subgroup each year to be considered adequate; failure to meet the growth targets even slightly for just one subgroup would have identified the school as failing.
The result, said Mr. Kress, could be an unrealistically high number of schools identified as needing improvement. "I think, at bottom, the president is just not prepared to say that 80 percent of the public schools in America are failing," he said.
Governors, including some of Mr. Bush's staunchest Republican allies, also complained. "If you set this adequate yearly progress at a somewhat unrealistic standard, and then you're going to tie rewards and sanctions to that, you could potentially create some perverse incentives for both states and local school districts to try and game the system," said Scott Jenkins, the education policy coordinator for Republican Gov. John Engler of Michigan.
So senators and administration officials came up with a formula—what some view as an excessively convoluted formula—for identifying schools and districts that need to improve.
Under the Senate plan, each state would have to devise a detailed system for determining adequate yearly progress that weighed whether each subgroup of students was making progress toward the 10-year goal and that gave extra weight to two groups: those that were furthest from the "proficient" level, and those that made the greatest improvement.
As a further backstop, districts and schools still would be identified as failing—even if they met that formula—if each subgroup achieved less than a 1 percent gain toward the "proficient" level in reading and math annually.
"It is complicated. This is not ideal," said Dan Gerstein, a spokesman for Sen. Joseph I. Lieberman of Connecticut, a leading moderate Democrat who has been involved in the Senate negotiations. "While this is not as strong, or as clearly linked to closing the achievement gap as we would have liked, it's still a reasonable middle ground. It still, in our view, will drive real change."
But Amy Wilkins, a policy analyst for the nonpartisan Education Trust, said the formula is so "overly complicated" that the public and educators wouldn't be able to understand it and, therefore, wouldn't accept it.
In addition, an analysis by the group found that the process of combining the scores of poor and nonpoor, minority and white students, into a composite formula would wash out the differences between groups and disguise achievement gaps, particularly in schools where poor children make up less than half the enrollment. About one-third of all Title I students attend such schools.
A hypothetical case study the group did of one such school found that the school could meet its "composite score" if achievement rose for both its poor and nonpoor students, even if achievement rose more for its nonpoor students, actually increasing the achievement gap.
Under another scenario, the school could be punished if all groups made progress, but the scores of the lowest-performing group improved most. That's because the group's score would be given so much weight it would pull down the composite average.
Mr. Kress, the president's adviser, described such scenarios as "highly unlikely." But the issue underscores the hoops that legislators are trying to jump through to draft workable language.
"This is what happens if you lock people in a room for two or three weeks; [their plans] get increasingly complex and convoluted," said William L. Taylor, the acting chairman of the Washington-based Citizens' Commission on Civil Rights, a watchdog group. But, he added, "contrary to what's being said, I don't think they've punched a big hole in 'adequate yearly progress.' As I read it now, it says you have to make continuous and substantial progress. It keeps the 10-year period and says everybody has to get there."
In contrast to the Senate bill, the House bill would require states to set annual, numerical objectives for each subgroup in one of two ways: They could require a minimum percentage of students to meet the proficient level on state tests, with the percentage the same for each subgroup; or they could require the achievement of each subgroup to improve by a certain percent every year, so that the rate of improvement for each subgroup ensured that they all reached the proficient level within 12 years.
A 'Good Goal'
That would be a big stretch for many states, some educators say.
"The bigger issue is how many states currently have a goal—no matter what their timeframe is—of bringing 100 percent of their kids up to proficient," said Margaret E. Goertz, a professor of education at the University of Pennsylvania. "You've got a whole bunch of states out there who don't even have this as a target."
Currently, only 14 states require that in the future nearly all their students attain a targeted level of performance, according to the Education Commission of the States in Denver. Twenty-five states require only a portion of their students to reach the target, and 13 don't specify any percentage.
Mr. Kress described the 10-year mark as a "good goal," but not necessarily one all schools would meet.
"There's not a death penalty at the end of 10 years," agreed Mr. Gerstein, the aide to Sen. Lieberman. "It's a benchmark you're supposed to be working towards."
Similarly, only a handful of states now report student test results by race, ethnicity, or income for every school and district. Only two states incorporate such breakdowns of data into their measures of adequate yearly progress, Ms. Goertz said.
Moreover, using the scores of every subgroup in a school to determine whether it has made adequate yearly progress may be "fairly unreliable," because of the inherent imprecision of testing, cautioned Brian Gong, the associate director of the National Center for the Improvement of Educational Assessment, a nonprofit group in Portsmouth, N.H.
In addition, many states are not helping schools they've already identified as failing. A study by the federal Department of Education found only about 40 percent of schools that have been identified as needing improvement are currently getting extra help from their states.
The final act of this dizzying drama will be written in the conference committee that hammers out the differences between the House and Senate bills. And there's a lot more work to be done.
"I think when we get to conference, there will be more discussion on it," Mr. Kress said. "It would certainly be desirable if it could be simpler."
Even then, others note, what any eventual changes actually mean for districts and schools will depend on how vigorously the administration enforces the law.
They point out that all states were supposed to have statewide testing and accountability systems in place by this school year, under the existing Title I law. But the Clinton administration gave full approval to only eight states' testing systems by mid-January, and had not yet reviewed most states' accountability plans.
Vol. 20, Issue 36, Pages 27, 32Published in Print: May 16, 2001, as Defining 'Failure' Critical To Bush Testing Plan