Debate on Graduation Rates Heats Up

Estimated rates vary widely, gaps have narrowed, and more accurate measurement is on its way.

A June 20th Education Week report estimated that 1.2 million students – a full 30 percent of the class of 2006 – failed to graduate on time this year. The reported estimates rekindled the on-going debates on graduation rate trends, gaps, and measurement methodologies.

The Ed Week Report

The report came as a surprise to education officials in many states, almost all of whom had previously reported strikingly higher state graduation rates. New Mexico and North Carolina, for example, reported graduation rates that exceeded the Ed Week report’s estimates by more than 30 percentage points. Graduation rates are cause for concern for many – particularly given recent statistics that confirm the importance of a high school diploma in the modern economy. In 2003, the median income of high school dropouts age 18 and over in 2003 was $12,184, while for graduates with a high school diploma or GED, that same figure was $20,431. High school dropouts are more likely to be unemployed, incarcerated, and even end up on death row, according to a recent NCES report.

The Ed Week report found that males are much less likely to graduate than females; this difference is most pronounced among blacks: in 2002-2003, 57.8 percent of black females will someday march with their classmates to the tune of “Pomp and Circumstance,” but only 44.3 percent of black males will do the same. In districts where most students are members of ethnic or racial minorities, graduation rates are almost 20 percent lower than majority-white districts. Rates are similarly divergent between high-poverty and low-poverty districts.

The report also found that graduation rates vary markedly across the nation’s largest districts, from a high of 82.5 percent in Fairfax County, Virginia, to a low of just 21.7 percent in Detroit. Differences among states were also sizable – with a high of 84.5 percent in New Jersey and a low of 52.5 percent in South Carolina.

Long-Term Progress

Two other recent reports, which received considerably less press coverage, address longitudinal graduation rates. The first, by the National Center for Education Statistics (NCES), was released on the same day as the Ed Week report. It found that the average graduation rate in 2002-03 – defined as the percentage of students who graduate in four years with a regular diploma (the same definition used by the Ed Week report) – was 73.9 percent nationwide. Generally, both reports ranked states at similar levels.

The NCES report is distinguished by its examination of the “event dropout rate” (the number of students who leave school each year without a high school diploma or equivalent) over a long period of time. In 2003, nationwide, this rate was just 4 percent, the lowest it has ever been in the thirty years the study assessed, down from 6.1 percent in 1972. For blacks, in the same period, this figure decreased from 9.5 percent to 4.8 percent; for Hispanics 11.2 percent to 7.1 percent. For low-income districts, the rate decreased from 14.1 percent in 1972 to 7.5 percent in 2003, for middle-income districts 6.7 percent in 1972 to a low of 4.6 percent in 2003. Additionally, since the 1991-92 school year, only three states – Delaware, Utah, and Louisiana – have seen increases in their dropout rate. Over this longer time span, it appears that both the nation and individual states have made great strides towards decreasing dropout rates, particularly among poor and minority students.

A book by Lawrence Mishel and Joydeep Roy, Rethinking High School Graduation Rates and Trends, released last April by the Economic Policy Institute, also suggests a national graduation rate that has increased over time to its current 82 percent. Examining longitudinal data that tracks individual students, the authors found that the black-white graduation gap has shrunk considerably, from 27.6 percentage points in 1960 down to just 10 points in 2000, while the Hispanic-white gap has also shrunk over the last 10 years.

The Response

Most state school officials give similar explanations for the disparities between state-reported graduation rates and these independent studies. First, officials say, it is not fair to compare states that have different requirements for graduation and different student demographics. In the Post and Courier, Inez Tenenbaum, state superintendent of Education for South Carolina, offered, “We’re proud that a high school diploma in South Carolina means something.” South Carolina had the lowest reported graduation rate in the country, but requires 24 credits (although top-ranked New Jersey has a 22 credit requirement). Others note that the states with the highest reported graduation rates also tend to have relatively few black and Latino students, while the states near the bottom of the list tend to have relatively higher proportions of ethnic and racial minorities. But maps of graduation rates by district and state provided in the recent Ed Week report suggest that high graduation rates also tend to correlate with higher per-pupil spending.

Second, many state officials suggest that the Ed Week report overestimated the number of dropouts for several reasons. The report included among “dropouts” students who changed schools or moved, received a vocational diploma, later received a GED, were incarcerated, or even passed away. Thirty-two states try to account for “alternate diplomas,” transfer students, and other factors like these. These states report graduation rates that are slightly higher, which they claim to be more accurate.

These methods also attempt to compensate for what experts call the “ninth grade bulge,” something that the Ed Week report was not able to do. The bulge refers to inflated ninth-grade enrollment (of poor and minority students in particular) because more students are held back at the end of that year; thus, the number of entering ninth-graders could actually be considerably lower than the “enrollment” might indicate. For example, in 2000, nationwide, there were 13 percent more ninth than eighth graders, and 26 percent more black and Hispanic ninth graders than eighth graders. Since most methodologies, including the Ed Week report, attempt to measure the probability of a student moving from ninth grade to graduation successfully, artificially high ninth grade enrollment numbers tend to lead to understated graduation rates.

Resolution to Come?

In the future, such a controversy may not exist: governors from all 50 states and leaders of 12 national organizations signed a compact in 2005 to adopt a common definition of a high school graduation rate. That definition follows the lead of South Dakota and 17 other states that have begun to use data to track individual students, instead of using enrollment rates to develop estimates. As South Dakota Education Secretary Rick Melmer said in the Argus Leader, “The difference between what they [Education Week] have done and what we do, is that they’re just using numbers… we track individual kids.” Melmer and other education department officials recently admitted that South Dakota’s previously reported graduation rates were artificially high because they had lacked the data to calculate graduation rates based on individual students.

Soon, more accurate data collection will allow for more precise calculations. Until then, most experts agree that energy should be re-focused on increasing – instead of estimating – graduation rates.

Prepared by Charley Cummings, June 27, 2006