(photo by Gideon Tsang)
Reports of another "Texas Miracle" are making the rounds in the media, as the Lone Star state says that the dropout rate for the Class of 2009 was 9.4 percent. That is, only one out of 10 students in Texas who entered high school in the fall of 2005 had quit school four years later.
A widely circulated Associated Press article opens by saying: "State officials are claiming the state dropout rate declined by almost 11 percent over the last year, but critics say the data being used is flawed and doesn't accurately reflect what's going on in Texas schools."
An 11 percent decline in the dropout rate sounds fabulous. And in just one year? Wow! This, my friends, sounds like true progress.
But is it? How exactly was the figure calculated? Conveniently, both the AP article and the Texas Education Agency's press release provide a straightforward explanation: 45,796 students dropped out in 2007-08, whereas only 40,923 students dropped out in 2008-09. Voilà! A decline of 11 percent. Actually, 10.6 percent, but the Texas Education Agency can be forgiven for rounding to the nearest whole number.
The trouble is that calculating dropout rates is tricky -- and highly political. The above calculation, we'll soon see, is a gross oversimplification. The 11 percent decline in the dropout rate is "fabulous" in the truest sense of the word, which comes from the Latin fābulōsus, for "fable."
The National Center for Education Statistics (NCES), the primary federal body that collects and analyzes education-related data, introduced a uniform definition of dropouts that states are expected to adopt in the near future. Currently, about half of the 50 states use the definition. The Texas legislature, which adopted the NCES definition back in 2003, provides a helpful summary of the definition: "a dropout is a student who is enrolled in public school in Grades 7-12, does not return to public school the following fall, is not expelled, and does not graduate, receive a GED certificate, continue high school outside the public school system, begin college, or die."
(It's worth noting that for decades the federal government has collected data on four different but related rates: the event dropout rate, the status dropout rate, the status completion rate and the averaged freshman graduation rate. The distinctions are important but subtle; click here for the federal government's explanation of these rates.)
You'd think the new definition would make calculating dropout and graduation rates easy. Or easier. Think again.
Kids move and schools lose track of them. Parents take their kids out of public school to homeschool them. Some kids flee to private schools. It gets very complicated very quickly.
The problem is exacerbated because data are typically reported by districts, which lessens the data's reliability. School administrators and district leaders have clear incentives to under-report dropouts.
In fact, no one likes high dropout rates -- not the public, not teachers, not administrators, and certainly not the politicians who stake their careers on improved graduation rates.
Let's return to the NCES definition of a dropout to see why the color gray dominates in dropout-rate reports. The big loophole turns out to be the category of students who "continue high school outside the public school system." How do we really know if a student who leaves a school has continued his or her education outside the public school system? Short of implanting chips in students and tracking their every move -- which I don't hear anyone advocating -- the unfortunate answer is we don't really know. If a student or parent reports that the family is leaving the state or country, schools and districts believe it. Why wouldn't they?
They believe it because there's no surefire way to know otherwise. That is, they believe it because it's convenient to do so.
Now, it's time to look more closely at Texas. Under the NCES definition, students who legitimately leave a state's system aren't counted as dropouts. But what does it mean to leave a system "legitimately"? A student might graduate from high school early or complete a GED elsewhere -- or even die young. Such "leavers" cannot be held against a school, district or state and shouldn't be counted as dropouts. But few students fall into these highly specialized categories.
A student might also leave the public school system to be homeschooled, return to one's home country, enroll in a private school or enroll in a school outside the state. And these are the four categories into which the vast majority of "leavers" in Texas fall. These "leavers" aren't dropouts, but they aren't necessarily graduates either. They're just gone, someone else's responsibility.
The Class of 2009 in Texas, which started high school in the fall of 2005, was comprised of 308,427 students. State data suggest that four out of every five students, or 248,500 individuals, actually graduated in 2009. Other estimates put Texas' graduation rate closer to 70 percent, including new data from the U.S. Department of Education, the annual "Diploma Counts" study by Education Week, and figures from the Intercultural Development Research Association.
According to state data, the number of students who officially quit the Class of 2009 was 28,856, or 9.4 percent of the total. (The rest of the students -- 30,000+ strong -- either earned GEDs or started on their fifth year of high school; they don't count as dropouts.)
But here's what is potentially troubling: there were another 61,179 students in the Class of 2009 who were written off as "leavers," and they don't show up in any of these calculations. These students ostensibly left the Texas public schools to return to their home country, to be homeschooled, to attend private schools or to attend public schools in other states. But did they? All of them? How do we know?
That's the $25,000-question.
Let's be generous and say that fully 95 percent of the students reported as "leavers" really did leave the state system. What if five percent didn't? What if instead they just dropped out? That'd be another 3,000 dropouts for the Class of 2009, which would push the dropout rate over 10 percent. It would also erase any claims about improvement between last year and this.
Finally, here's another puzzling piece of information. In claiming that the state's dropout rate fell by 11 percent between 2007-08 and 2008-09, the Texas Education Agency had to use data from the 2004-05 academic year, a time when the state's dropout rate was calculated differently.
So no comparisons can be made. Apples to apples this is not.
The TEA admits as much in its press release, in fact, but not until the penultimate paragraph:
Because of the phased in collection of the data, the Class of 2009 data is the first set of graduation statistics available using four years of data collected under the NCES definition. Therefore, statistics for the Class of 2008 are not directly comparable because they are based on three years of data collected using the NCES definition and one year using the previous state definition.
So why the direct comparison in the press release's second paragraph?
Hmm. You'd almost think this is an election year with an incumbent fighting for his political life!