Atlanta grad rate doesn't add up

Thousands of high school students vanished from the rolls of Atlanta Public Schools in the past eight years, often with few hints to where they went.

Schools recorded many of them as “transfers” to other systems, at times without proof that the students hadn’t dropped out altogether. In 2008, a consultant to the district estimated recently, school officials couldn’t document the whereabouts of more than one-third of the district’s departed students.

The mass exodus from Atlanta’s high schools may be the primary reason for one of the district’s proudest academic achievements: a dramatic increase in its graduation rate, an investigation by The Atlanta Journal-Constitution shows. District officials boast that the rate of students getting diplomas within four years has risen 30 percentage points since 2002.

But the rate’s only surge, from 43 percent to 72 percent, came between 2003 and 2005, the Journal-Constitution’s analysis of state data found. During that time, the district removed from its rolls about 30 percent of all pupils in grades nine through 12 — roughly 16,000 students.

As a result, most of those students no longer figured into the district’s calculation of what Superintendent Beverly Hall has described as the “all-important” graduation rate: The fewer students being counted, the fewer graduates needed to make the rate higher. A student listed as a dropout would count against the rate. A transfer would not — even if school officials didn’t know, or didn’t try to find out, where a student went.

The district has sustained its higher graduation rate — 69 percent in 2009 — despite a two-thirds increase in dropouts since 2003.

A majority of the dropouts occurred at one school: an “alternative” campus for students with academic and behavioral problems. Many of the students most at risk for quitting were concentrated there, allowing other schools to keep their graduation rates up.

Questions about the graduation rate follow investigations into unlikely gains at many Atlanta elementary and middle schools on a state-mandated standardized test in 2009. After the Journal-Constitution published two articles reporting statistically improbable increases at some schools, state education officials found suspicious erasures that boosted scores on the Criterion-Referenced Competency Test, or CRCT, in hundreds of classrooms in 58 Atlanta schools. An examination commissioned by Atlanta’s school board concentrated on 12 schools where it said cheating may have been the most pervasive.

Taken together, inquiries into the CRCT, the graduation rate and dropouts raise doubts about the academic turnaround in which Atlanta’s school system has basked. The turnaround has attracted millions of dollars in corporate grants for a district where students historically have struggled and has garnered numerous national awards for Hall. At the same time, the apparent success has created pressure inside the district to continually produce more achievements.

“Over the past 10 years, our school district has made reforms that are multilayered and comprehensive,” Hall said in a speech last year. “As a result, by state and national standards, very few school districts in the United States can match the rate of student progress in Atlanta.”

In an interview Friday, district officials said no one had tried to rid the district of low-performing students, and the data do not suggest otherwise.

Since 1999, when Hall became superintendent, “we have been on a mission to make sure more of Atlanta’s students graduate and go to college,” Kathy Augustine, the district’s deputy superintendent for instruction, said in the interview.

But experts in dropout prevention say the district should have been more curious about the spike in the graduation rate from 2003 to 2005. Instructional reforms alone typically don’t produce such gains so quickly, said Jay Smink, executive director of the National Dropout Prevention Center/Network, a South Carolina-based clearinghouse.

“In general it’s very unusual to see that kind of a change,” Smink said. “There are rarely significant jumps unless several things happen.”

For instance, an influx of immigrants might drive the rate down. Or a change in how schools calculate the rate might push it up.

“Whenever you see blips you need to do a drill-down in that data to see if you can explain it,” Smink said. “There’s got to be a reason, [although] I’m not sure what it is.”

Graduation rates have been the subject of debate in education circles nationally. Officials disagree over the best formula to track students who move away or simply disappear from schools. Georgia, to comply with federal rules, is switching to a new equation that aims to better follow students through the upper grades.

The formula that Georgia now uses presents Atlanta in a favorable light. Other education researchers, however, have pegged the district’s graduation rate as far lower. In a report issued last year, the America’s Promise Alliance, a Washington-based nonprofit group devoted to children’s issues, said Atlanta’s 2005 graduation rate was 43 percent — 45th-lowest among the nation’s 50 largest cities.

Dwindling enrollment

When they entered the ninth grade, Atlanta’s class of 2003 stood 4,583 strong. Four years later, when 2,366 seniors remained, the district calculated a graduation rate of 43 percent.

The class of 2009 experienced comparable attrition of the rolls: 4,332 freshmen dwindled to 2,351 seniors. But that year, the district said, the rate was 69 percent.

Atlanta’s overall high school enrollment dropped 12 percent from 2003 to 2009, state data show. But since 2003, according to the Journal-Constitution’s analysis, each year’s senior class has been, on average, 45 percent smaller than the corresponding freshman class four years earlier.

By contrast, districts such as Cobb County and Fulton County recorded declines less than half as large. Among the six largest districts in the metro area, only Clayton County lost a greater proportion of its students; Clayton also lost its accreditation during that time.

When a student leaves a school, officials are supposed to document a reason: as extreme as death, as commonplace as moving across town.

In Atlanta, large proportions of students transfer not just from one city school to another, but withdraw altogether from the district. In 2005, for instance, more than 5,500 Atlanta high school students — about 30 percent of all those enrolled — moved out of town or dropped out in a single year.

Such numbers attracted the attention of state education officials, who last year asked the district for an explanation.

Consultants to the district sampled two dozen student records at each of four high schools. In 35 percent of the records from 2008, they found no documents that supported codes showing the students had transferred to schools outside the district. Documentation was missing in 21 percent of transfers examined from 2010.

Older transfers couldn’t be examined because the district had recently installed new student record-keeping software, said Dan Bugler a senior research associate in the Atlanta office of WestEd, the nonprofit research group that conducted the study. The new software counted fewer transfers; this, Bugler said, suggests duplicate entries might have inflated the previous years’ figures. The newspaper’s analysis of state data weeded out duplicates, however, and counted only one record per student identification number.

The district has undergone “massive efforts” to improve data collections, Augustine said. But she said tracking students is difficult for school districts where most children are poor. Three-fourths of Atlanta students, or more than 35,000, qualified last year for a free or reduced lunch, compared to a statewide rate of 53 percent.

The transient nature of many poor families makes tracking students particularly difficult for urban school districts, Augustine said.

“Families sometimes move month to month,” enrolling children into new schools each time, she said. “We just know that is a reality to us. We know that is one of the reasons we have transfers.”

Clearly, not all the withdrawals are misreported dropouts. Nevertheless, experts said, the district should view irregularities as more than quirks in their data or inevitable byproducts of evolving demographics.

“We’re talking about children’s lives,” said Jimmy Garcia Arispe, president of the National Coalition for Exemplary Schools, a Smyrna-based consulting firm. “It may translate just into numbers. But we’ve missed someone through this process. That’s the really sad thing. What happened to them? Where are they?”

‘Set for failure’

Crim High is where students go to quit school.

The alternative campus, off Memorial Drive in East Atlanta, accepted 673 transfer students from other Atlanta schools in 2009, according to state data. By year’s end, state data show, three-fifths of them had dropped out. Most lasted at Crim little more than two months.

When WestEd looked into other withdrawals from Crim, administrators couldn’t prove the whereabouts of two-thirds of the students who transferred in 2008. The consultants’ findings suggest the school’s dropout rate may be even higher than the already astronomical figures previously reported: 51 percent in 2008, and 62 percent in 2009.

The district converted Crim from a traditional high school to an “open campus” in 2005 that offered night classes, accelerated course work and other accommodations for students — such as those who need a job, or who have children of their own — who might not succeed in a regular classroom setting.

Crim, Augustine said, gives students flexibility “so you can be as successful as possible.”

“You go to get what you need,” she said, “and you leave.”

But as what Augustine called a “second-chance school,” Crim draws the students most likely to drop out, hundreds of whom cycle through each year, many of them barely leaving a mark. Nearly 1,500 enrolled during the 2009 school year, for instance, but neither of the state’s semi-annual attendance tallies counted more than 525 at a time.

Last year, 847 students dropped out of Crim — more than 70 percent of all Atlanta pupils recorded as quitting school. The district’s 21 other high schools reported 349, combined.

Crim’s dominance of the dropout numbers cannot be easily explained, Augustine said. But she added: “There is no systemic effort in the district to shift children to Crim for them to drop out.”

Some students and parents at Crim are skeptical.

“I don’t know how to explain it, but it’s like [the school] is set for failure,” said Michael Lancaster, 17, who entered Crim after he was kicked out of Douglass High.

Melinda Jones, 36, is the mother of two Crim students, one 19 years old, the other 18. Neither, she said, seems to have progressed academically.

“I’m just hoping and praying this will work out,” she said. “They want to learn. They want their diploma, but I just don’t know.”

District officials discount the scope of the dropout problem, contending they have a minimal effect on the graduation rate.

“We’re not talking thousands of kids,” said Lester McKee, the district’s head of research, planning and accountability. “We’re talking several hundreds.”

At least once, the superintendent refused to acknowledge the problem at all.

“The dropout rate is down,” Hall wrote in an op-ed piece published in the Journal-Constitution in February — even though, in the previous school year, the district had posted its highest dropout rate since at least 2002.

District officials on Friday acknowledged that Hall’s article was wrong, and Jeff Dickerson, a public relations consultant working for the district, offered an explanation: He ghost-wrote the article for Hall and overlooked the inaccuracy. He wasn’t sure, though, whether Hall reviewed it before publication.