SCHOOLS & LEARNING
'Safe Harbor' Offers Shelter From Strict 'No Child' Targets
Monday, April 7, 2008
Hundreds of schools in Maryland, Virginia and the District were judged to have made adequate progress last year under the No Child Left Behind Act even though they failed to meet performance targets for all groups of historically underperforming students, the requirement at the heart of the law.
The schools -- 153 in Maryland, 100 in Virginia and 11 in the District -- satisfied the law under a "safe harbor" provision. It forgives a school for low test scores from one or more subgroups if those students show yearly improvement and if the school scores well on the whole.
The extent to which schools rely on the provision illustrates the challenge posed by the No Child Left Behind law, which requires that poor and minority students meet the same performance goals as the overall school population, with the target rising by a few points each year. In Maryland, where students are being tested this week, schools must raise proficiency rates by five points in reading and seven points in math this year to demonstrate adequate progress. Testing begins April 22 in the District and in May for most Northern Virginia schools.
The safe harbor provision is a loophole, essentially, in the education law, which sets a national goal of 100 percent proficiency by 2014 for all students, including seven demographic subgroups defined by race, language, poverty level and disability.
But almost no one in the education field appears to dispute the value of the provision. Critics of No Child Left Behind embrace the provision, regarding it as a vital relief valve in an inflexible law. The safe harbor concept might loom large in the debate over the future of the federal law, which has been broadly criticized for judging schools based on performance targets and ignoring academic growth over time.
"People who complain that No Child Left Behind doesn't reward schools that are making great gains, well, it does, and this is how it does that," said Michael J. Petrilli, vice president for national programs and policy at the Thomas B. Fordham Foundation, a nonprofit organization dedicated to education reform.
The federal law, enacted in 2002, measures schools annually on the share of students who are judged proficient in reading and math. In Maryland, the targets this year are 69.7 percent proficiency in reading and 60.9 percent in math. In Virginia, the targets are 77 percent proficiency in reading and 75 percent in math. Targets for the District were not available.
Many schools will meet the targets, and many will not. The law provides two exceptions for schools that fall short. One is a statistical "confidence interval," a concept commonly used in polling: If a school tests just six special-education students, for example, the margin of error is wide and the school is given the benefit of the doubt if even one or two of those students pass the test.
The other exception is safe harbor. In this instance, the school must also demonstrate good attendance, or, in the case of a high school, a solid graduation rate.
At Tilden Middle School in Rockville, the student population has never rated lower than 74 percent proficient in reading or math in five years under the Maryland School Assessment system. But the school failed to make adequate progress in 2005 and 2006 because of poor performance by a large population of special-education students and, in the latter year, by low-income students. Tilden's special-education students have not met the proficiency target in reading or math in three years.
In 2007, safe harbor came to the rescue. The share of students rated proficient in reading rose from 31 percent to 40 percent, enough growth to satisfy the rule. Math scores missed the target but fell within the statistical margin of error for adequate progress. Tilden used the safe harbor provision to have low scores in two other subgroups, black and low-income students, forgiven.
The school celebrated finally making adequate progress. Principal Jennifer Baker said she has done her best to explain safe harbor to parents "in a way that people don't fall over asleep." She hopes the school can make adequate progress this year, by safe harbor or any other means, as proficiency targets inch upward. "It is a very hard goal to reach," she said.
As the bar rises, Maryland school officials are seeing an increasing number of schools that score well overall but don't make adequate progress because of a single subgroup, usually special-education students. Nearly all the 153 Maryland schools that invoked safe harbor last year struggled with special-education scores.
A school that repeatedly fails under the No Child Left Behind law, even because of a single test score, "could eventually end up in restructuring, which might call for a new principal, a new staff and so on," said Ronald A. Peiffer, deputy superintendent of Maryland schools. To many educators, that doesn't seem fair.
The U.S. Department of Education has announced a series of pilot programs that give schools more flexibility in satisfying the law.
One initiative, announced in 2005 and expanded in December, allows states to measure schools by the growth of individual students over time. The original No Child Left Behind law, by contrast, measured schools mostly on test-score targets, not on growth or decline in scores. Many educators hope the law will eventually allow all schools to demonstrate adequate progress either by scoring well or by showing improvement.
"If we're going to talk about scores, we need to talk about growth: where the students are now, where they were," said Brenda Foxx, principal of University Park Elementary School in Prince George's County, which made adequate progress last year under the safe harbor provision.
Another initiative, announced last month, invites states to create strategies to deal with schools that fail to make adequate progress. The proposal, called "differentiated accountability," addresses a perceived flaw in the federal education law: Schools that miss targets are treated largely the same way and face largely the same penalties, no matter how far they are from meeting the goals.