One in Four U.S. Students Drops Out

One in Four U.S. Students Drops Out
February 1, 2001

Although a recent Census Bureau publication has been widely reported as showing more Americans than ever graduating from high school, with gradually improving dropout rates nationwide, data from the U.S. Department of Education make it clear that the U.S. high school graduation has fallen precipitously since 1993. Today, one out of every four U.S. high school students drops out of school before graduation.

The U.S. high school graduation rate peaked in 1992 and 1993 at 81.2 percent, then fell to a low of 74.7 percent in 1998, according to the November 2000 report Dropout Rates in the United States: 1999, from the National Center for Education Statistics. The graduation rate recovered slightly to 76.8 percent in 1999, meaning only three out of every four ninth-graders ultimately receive a high school diploma, an absolute tragedy in an Information Age.

In California, ground zero for the cyber revolution, high school graduation rates are even lower, falling from 76.1 percent in the 1974-75 school year to just 68.4 percent for the 1998-99 school year.

What also peaked in 1992, at 86.4 percent, was the nationwide "high school completion rate," a statistic often confused with high school graduation rate. The completion rate is the total of high school graduations plus completion through GED or other means, which in 1992 added 5.2 percent to the high school graduation rate of 81.2. By 1998, post-dropout completions of high school education had soared to 10.1 percent of all high school completions, masking the fall in the high school graduation rate to 74.7 percent.

While adding GEDs and other post-dropout completions produces a relatively stable high school completion rate, they aren't as highly valued as a real high school diploma. As education researcher Lance Izumi from the Pacific Research Institute has demonstrated, the employment market doesn't value a job applicant with a GED or a high school equivalency certificate much more than a dropout, with the additional qualification producing an insignificant increase in salary.

Census data for high school graduation and dropout rates are inherently unreliable, as the data are self-reported: They simply rely on what respondents say about their own educational attainment levels and those of their children. The Census Bureau's internal analysis of its methodology acknowledges it has never verified whether respondents are telling the truth. Moreover, its data are inconsistent with attrition rates, the complement of high school graduation rates.


Dropout Rates Misleading

Much of what is reported on dropouts is incorrect at best and highly misleading at worst.

Dropout rates tend to be off the radar screens of families who live in "good" neighborhoods and send their children to "good" public schools, because dropout rates usually are low in such schools. Many Americans don't realize that many inner-city high schools lose two-thirds of their students to dropouts year after year--a tragedy that is the root cause of inner-city disintegration and the skyrocketing prison population.

Dropout rates vary tremendously among ethnic groups and by socioeconomic status, producing significant differences in high school graduation rates. For example, the high school graduation rate for black students in 1999 was 72.9 percent--3.9 points lower than the national average of 76.8 percent. By contrast, the graduation rate for white students was 82.0 percent, and that of Hispanics only 54.9 percent.

The dropout figures produced by the public school establishment hide this tragedy from the public. The statistic most commonly used by public schools to report dropouts is the event dropout rate, in which an attempt is made to count the individual dropouts from grades 10-12 as they leave school. This rate is most often presented to the public as the official dropout rate, when in fact it is a number that is just a fraction of the truth. The national event dropout rate for 1999 reported in the November 2000 NCES report was 5.0 percent.

Event dropout numbers are almost never audited and almost always misleading. Districts usually fail to count the following classes of students as dropouts:

  • students who drop out during summer vacation;

  • students who are sent to county continuation schools and then drop out; and

  • students who "transfer" to other high schools, even if they never make it to the next school.

The recent NCES dropout report illustrates an even bigger problem with the event dropout rate. Nowhere in the report is any mention made of the fact that this measure of dropouts is not a four-year figure, but rather a one-year figure.

To understand how misleading the event dropout rate is, take a high school with four grades of 100 students each and 40 dropouts in one year. The event dropout rate is 40/400, or 10 percent. However, if all these dropouts occur in the 12th grade, then the school's graduation rate is only 60 percent, or an actual dropout rate four times higher than the event dropout rate.

The four-year derived dropout rate is the event dropout rate multiplied by four to estimate dropouts occurring over the four-year high school career of the average student. Based on the 5.0 percent event dropout rate reported by NCES, the national four-year derived dropout rate for 1999 was 20.0 percent. However, the NCES dropout report makes no mention at all of this derived dropout figure.

Another commonly reported dropout statistic is the status dropout rate: the number of young adults aged 16 to 24 who are out of school and have not yet completed a high school program. According to NCES's November 2000 dropout report, the national status dropout rate for 1999 was 11.2 percent. But the data underlying status dropout rates are derived from Census Bureau surveys, meaning the cautions mentioned above regarding self-reporting bias also apply to this rate.

Attrition Rates More Accurate

None of the dropout rates reported by NCES provides a reliable estimate of actual dropout rates. The hard number that counts in calculating dropout rates is the high school graduation rate: the percentage of children who graduate from high school compared to those who started ninth grade four years earlier. Its complement, the attrition rate, is the best available estimate of the dropout rate.

With a 1999 national high school graduation rate of 76.8 percent, the national attrition rate is 23.2 percent. This is significantly higher than both of the official dropout rates reported by NCES: 5.0 percent for the event dropout rate and 11.2 percent for the status dropout rate. It is, however, quite close to the derived dropout rate of 20.0 percent--which NCES did not report.

For different ethnic groups, national attrition rates for 1999 are 18.0 percent for white students, 27.1 percent for black students, and 45.1 percent for Hispanic students.

Attrition rates don't count the following four "acceptable" reasons for not completing high school:

  • students who earn GEDs;

  • students who earn high school equivalencies;

  • students who start college without graduating from high school; and

  • students who die before completing high school.

However, all of these factors put together don't alter the dropout picture very much, since students who drop out rarely return to school. Further, the attrition rate often understates the true dropout rate, because it must be adjusted upward if enrollment is growing to reflect new students coming into the system and taking the place of those who dropped out. In California, for example, that upward adjustment for rapid population growth overwhelms the downward adjustments for GEDs and the like.

Attrition rates don't even count the small but especially tragic number of dropouts who leave junior high school and never even start high school.

Alan Bonsteel M.D. is president of California Parents for Educational Choice and coauthor, with Carlos Bonilla Ph.D. M.D., of A Choice for Our Children: Curing the Crisis in America's Schools (Institute for Contemporary Studies, Oakland, 1997). His email address is

For more information . . .

Use PolicyBot, The Heartland Institute's free online research service, to request the documents in Adobe Acrobat's PDF format. Point your browser to and click on PolicyBot. Search for the document numbers above.