Wednesday, June 2, 2010

Answering your graduation rate questions—sort of.

I am frequently asked what the U.S. dropout rate is. That’s like asking how you make chocolate ice cream; there are so many possible methods and outcomes. It is easy for me to make recommendations when it comes to ice cream: Alton Brown if you have a lot of time or David Lebovitz if you don’t. Graduation rates, however, are more complicated.

Today NCES released its most recent calculations of graduation and dropout data, taken from the Common Core of Data. The dropout rate—the percentage of students enrolled in 2006-07 who were not enrolled in 2007-08 and had not completed school—was 4 percent. The average freshman on-time graduation rate—the percentage of high school students who entered ninth grade in 2004 and graduated in 2008—was 75 percent, up from 74 percent the previous year. Keep in mind that this doesn’t count the real people who entered school and graduated. It just compares the total number of students at the start  to the number at the finish line. If 100,000 students drop out sophomore year and 100,000 exchange students from Belgium arrive junior year and get diplomas, that’s considered a wash. And having 4 percent of students drop out each year leaves more than 75 percent left to graduate, yes?

There are other wishy-washies you can read about in the methodology section.  They are sort of inevitable, I guess, until we have data systems that actually track individual students. But the info we do have is still worth paying attention to. Take a look; you might find something intriguing. Like, why is ninth grade such a flashpoint for dropping out in Louisiana, as opposed to the later grades in other states?Or depressing: Why do practically half of your students fail to graduate on time, Nevada? I do wonder if some of the variation between states can be attributed to reporting issues rather than simply who does better or worse by its students.

P.S. to Vermont and South Carolina reporters: Your states were missing from the some of the counts because of missing data. Why don’t you find out what’s up with that?


  1. Alexander HoffmanJune 2, 2010 at 4:31 PM

    Richard Elmore -- one of the smartest guys in ed policy -- makes clear when asked about the impact of X or Y on graduation rates that he does not believe anything about such claims because he's seen the condition of graduation rate data around the country.

    We don't know the graduation rate, in large part because there has not been a consistent defiintion.

    In one state -- I don't think that it was just one district -- they measured the number of kids graduating as a % of the kids still school at the end of senior year. That is, they totally ignored the drop outs. That's about as bad as *I*'ve ever seen.

    Otherwise, until states track individual students longitudinally, the states are too hard to track. If families move teenagers south, then nothern states will suffer in graduation rates (because some of their one-time nine graders move south before they graduate) and southern graduation rates will be inflated (because many of their graduates don't start school there. So many kids repeat 9th grade that comparing # of 9th graders to # of graduates double counts a large number of kids -- a well known problem.

    Even when those technical problem are overcome, we've got the issue of what counts as graduation. When states move the bar, how do we want to intepret that? How do we want to count kids who needed an extra summer (i.e. for summer school for that one class?). How do we want to count such things as "credit recovery" in NYC? Do we want to insist on 4 years, or is 5 ok?

    Think hard about that last one. If a kid totally screws up 9th grade and has to repeat it, but then goes through grades 10, 11 & 12 like gangbusters, what do you want to do that that? I would count that as a success for the school because what matters is the quality of the graduates and that's a kid who figured it out. But others might think differently. That not a technical question, but rather a personal judgement.

    A year or two ago, the states agreed on a common defition. It would interesting to see a piece on how that implementation process is going, and on what that has done to graduation rates.