Adding It All Up

To the uninitiated, calculating a high school’s graduation rate probably seems easy. Just divide the number of graduates by the total number of students in a class.

But accounting for the constant change that high schools experience—students who transfer out or in, others who drop out only to return later, and still others who need five years to earn a diploma—makes producing accurate rates far from easy.

Pressure to come up with more accurate numbers has been mounting, though, because of the federal No Child Left Behind Act. The 2001 reauthorization of the Elementary and Secondary Education Act judges the performance of high schools and districts based, in part, on their graduation rates. The provision was intended to balance the law’s focus on raising test scores with a requirement that would discourage schools from pushing out low-performing students.

But because the U.S. Department of Education has permitted each state to propose its own method for calculating graduation rates, and its own targets for how quickly or how much those rates should improve, critics contend that the provision has failed to live up to its promise.

“The implementation of the NCLB graduation-rate provisions has been a travesty,” says Bethany Little, the director of federal advocacy and policy development for the Alliance for Excellent Education, a Washington-based group pushing to improve high schools. “The department approved state plans with graduation-rate calculations that would be laughable if they weren’t gambling with the education of at-risk students.”

Overstating Success

For years, states have published graduation rates using a variety of measures. A state, for example, may simply divide the number of graduates by the number of students who started the year in the senior class. This approach misses the majority of dropouts, who leave school before the 12th grade.

Other states track students as they move from school to school and know whether they drop out or continue toward a diploma. But they may still consider students who earn a General Educational Development credential as “graduates”—a practice that many experts say exaggerates states’ success in ensuring that students have successfully completed rigorous high school coursework.

In recent years, researchers have documented states’ systematic underestimation of the percentage of students who fail to complete high school. By dividing the number of graduates by the number of students who entered 9th grade four years earlier—or at least an estimate of that number—those researchers now suggest that between 25 percent and 30 percent of those who enter high school fail to earn a diploma within the standard four years.

While many researchers agree on the general conclusions based on the data—with a few exceptions—significant differences remain over how to estimate graduation rates and how those estimates should inform policies to improve high schools.

“This is a vibrant discussion,” says Kerri Briggs, a senior adviser to Deputy U.S. Secretary of Education Raymond J. Simon. “There are a lot of differences of opinion about how to track graduation rates.”

In reports to the Department of Education required under the No Child Left Behind Act, most states say that their graduation rates are 80 percent or higher, with some publishing rates as high as 95 percent.

But researchers say those reports range from optimistic to wildly inaccurate.

Jay P. Greene, a senior fellow of the Manhattan Institute, a New York City-based think tank, and Christopher B. Swanson, formerly of the Urban Institute and now the director of the Editorial Projects in Education Research Center, have developed their own methods of estimating a state’s graduation rate using enrollment and diploma data from a federal database.

While the two methods differ slightly, they use the same basic data and a similar strategy. Their methods both avoid using the unreliable dropout data that experts believe may inflate state-reported graduation rates.

While both researchers acknowledge that their figures are estimates, they believe their calculations are closer to the truth than the figures most states publish.

Using his method, Greene says that the nationwide graduation rate was 70 percent for the class of 2003. The gaps between graduation rates for members of minority groups and whites were significant. Seventy-eight percent of white students earned a diploma in four years, while 72 percent of Asian-Americans, 55 percent of African-Americans, and 53 percent of Hispanics did.

In his research for Diplomas Count,Swanson estimates the rate was 69.6 percent for the class of 2003.

One prominent critic, though, argues that Swanson’s and Greene’s estimates are not accurate when compared to other data.

“Their answers are profoundly wrong,” says Lawrence Mishel, the president of the Economic Policy Institute, a Washington think tank.

Surveys conducted by the U.S. Census Bureau and labor economists regularly find that between 85 percent and 90 percent of adults have earned high school diplomas, he notes. A survey by the federal Department of Education that tracked a nationally representative sample of 8th graders starting in 1988 found that 78 percent graduated on time, and that 83 percent eventually earned a high school degree or General Educational Development certificate.

But that longitudinal survey was not designed specifically to produce graduation rates, and Census Bureau and other surveys aren’t accurate measures because respondents often lie or exaggerate their accomplishments, responds Greene, who is also an associate professor of education reform at the University of Arkansas at Fayetteville.

Governors’ Agreement

While researchers debate the best way to estimate graduation rates, policymakers are adopting methods that are similar to Greene’s and Swanson’s.

In 2005, the National Governors Association persuaded all 50 governors to work toward changing the way their states calculate and report graduation rates, based on the number of students who enter 9th grade and exit high school four years later. The rates also will reflect the students who transfer in and out of the schools.

Before the states can fully carry out that agreement, though, they will need to upgrade their data systems significantly. Right now, just 17 states have all the elements in their data systems needed to accurately calculate a graduation rate using the NGA formula, according to the Data Quality Campaign, a group of policy and research groups working to improve state data systems. The campaign’s members include the NGA and the Council of Chief State School Officers.

But states have made great progress in producing the data systems needed to calculate graduation rates based on actual student results. For example, the number of states with a code to track students from one school to the next has increased dramatically, from eight in 1999 to 44 now, according to the EPE Research Center. Once those “identifiers” are in place, states will need to link that data to transcripts and other high school data.

The Data Quality Campaign is pushing states to have the data capability to measure graduation rates and other indicators of school improvement within three years.

“Be patient. It does take time to build these systems,” says Dane Linn, the NGA’s education policy director.

But advocates for high school improvement say the problem is a long way from being solved.

“This is the right first step,” says Little of the Alliance for Excellent Education. But with 36 gubernatorial elections in fall 2006—and many governors not seeking re-election—the implementation of the NGA agreement, she points out, will be in the hands of many new governors who didn’t sign on to the deal.

NCLB’s Role

Meanwhile, the No Child Left Behind Act, which President Bush signed into law in January 2002, will continue to play a big role in graduation-rate policies.

The law requires high schools to use graduation rates as an indicator of whether they’re making adequate yearly progress toward the academic goals set for the 2013-14 school year. It says states and schools must base their published rates on the percentage of students who graduated four years after their class entered 9th grade.

But Little and others contend that the federal Education Department has been lax in carrying out the law. Federal officials have approved some states’ plans for NCLB compliance even though those states won’t be working toward a 100 percent graduation rate. Some states haven’t even committed to raising their current rates.

That’s inadequate, Little says, because the law requires states to aim for 100 percent proficiency in reading and mathematics for students who take annual state assessments in grades 3-8 and once in high school.

States have “set graduation rates at such an unbelievably low level that it’s been irrelevant [for schools],” Little says.

Organizations such as the Alliance for Excellent Education will continue to push the Education Department for more ambitious targets on graduation rates, both as the current version of the law is being implemented and as Congress considers making changes to it, starting in 2007.

“This is going to be a huge issue” for the reauthorization of the law, Little says.