A year after all 50 governors agreed to take on the high school dropout crisis by counting their graduates using a single formula, two states - North and South Dakota - already have abandoned plans to use it.
South Dakota Secretary of Education Rick Melmer said the decision probably will make his state look "like we're rebels without a cause." But he said he fails to see much difference between his state's current formula for tabulating graduation rates and the uniform graduation standard that governors pledged last summer to adopt.
"I'm disappointed that we're looked at as an outlier, as not playing ball with the rest of us," Melmer said. "That isn't my take on it at all."
Three states - Hawaii, Illinois and Washington - have not decided whether to adhere to the Graduation Counts Compact formula the governors signed, while 13 are ready to report their graduation rates. By 2010, at least 39 states are expected to be using the new method, with more states conforming later, according to a new report by the National Governors Association (NGA) on where states are in implementing the formula.
Experts say the first step toward improving the nation's high school dropout rate - which a recent Education Week report put at about 30 percent - is to accurately and uniformly track how many students who start high school actually finish in four years.
States have been using a hodgepodge of different methods to calculate this, and their self-reported rates have differed hugely from independent estimates. For example, Indiana reported a graduation rate of 91 percent for its 2002-03 class, compared to Ed Week's estimate of 73 percent.
Some states did not differentiate students who took more than four years to graduate, while others counted credentials such as the General Education Development (GED) certificate as a diploma - both methods that inflate a state's graduation rate.
"Comparisons get made, inevitably," said Bridget Curran, the author of the NGA status report. "As long as that's the case, it's important to have something that's reasonably comparable so that we're looking at apples to apples and not apples to oranges when we compare states' graduation rates."
The formula in the compact announced by the NGA in July 2005 counts the number of students who graduate within four years with a regular or advanced degree - not a GED - and divides that by the number of first-time ninth-graders four years earlier. The freshman figure can be adjusted for transfers, while special education students and those with limited English proficiency can be counted in different classes to give them more time to graduate.
The formula requires states to set up student identifiers to track students, or to collect accurate data from the districts on student enrollments. States also should be able to verify schools' claims that some students are transfers and not dropouts.
Melmer of South Dakota said the main difference between his state's formula and the NGA formula is philosophical: South Dakota does not count those who take more than four years to graduate as dropouts.
If students are still enrolled in school, they will be counted with some graduating class, although not the one in which they entered high school. Melmer said this past school year had two high-profile cases of boys who couldn't graduate on time because one had leukemia and the other had a stroke.
"Neither of those young men graduated on time. They had circumstances that caused them to miss classes," Melmer said. "In our estimation, we don't believe those kids are dropouts. They ought to be allowed to finish their coursework, but the NGA formula would count them as dropouts."
According to Curran, the NGA compact does not automatically count such students as dropouts, but recommends that states tally five- and six-year graduate rates in addition to regular graduation rates.
Both North and South Dakota base their graduation rates on the number of dropouts. Instead of using enrollment counts to calculate a graduating class' size when it entered high school, as the NGA rate does, the Dakotas estimate a freshman class by adding the number of dropouts over the past four years to the number of graduates.
The problem with that method, according to the NGA, is that it doesn't include students who unofficially leave the school system or whose whereabouts are unknown, thus inflating the graduation rate.
In every state except North Dakota, students who disappear from the school system are listed as dropouts. In North Dakota, they're listed as "unknown" and taken out of the graduation rate equation.
"Now if you treat them as a dropout, you're making a presumptive decision there," said Greg Gallagher, the state's director of standards and achievement.
Students may slip through the cracks not because they dropped out but because of an administrative error, Gallagher said, adding that North Dakota is increasing its oversight of transfers within the state to avoid losing track of students.
Melmer questioned whether the remaining 48 states actually will follow the compact formula to the letter, or whether some of them will adhere but with differences.
In fact, three states with formulas that differ slightly from the compact formula have not decided yet whether they will fall in line. Washington state's system follows the recommendations of the compact but is based on estimates; Hawaii's formula only adjusts for transfers out of the system and not into it; and Illinois, like South Dakota, does not distinguish those who take more than four years to graduate.
Although some states only need to tweak their previous formula, others will see dramatic changes. New Mexico's previous way of calculating graduates took the number of 12 th -graders in the fall and counted how many of them received diplomas in the spring. The state reported an 89 percent graduation rate in 2002-03, but Ed Week found New Mexico's graduation rate to be only 57 percent that year.
North Carolina's previous formula counted how many students in a graduating class took only four years to finish high school. It reported a graduation rate of 97 percent for its 2002-03 class, while Ed Week calculated that only 66 percent graduated.
Previously, Massachusetts didn't even count its high school graduates. Instead, it released a "competency determination rate" that counted the number of students eligible to graduate based on how many passed the state's exit exam. For its 2003-04 class, Massachusetts reported a graduation rate of 96.2 percent.
Other highlights from the NGA report: