The media recently reported high school graduation rates for each U.S. state courtesy of Educational Projects in Education (EPE), publishers of Education Week magazine. But don't be fooled---they are not to be believed. The respected Economic Policy Institute has labeled the report as, "exceedingly inaccurate." Why is it so inaccurate? Please read on.

If 90 students graduated from the local high school this year, can you tell me the graduation rate? No? Why not? Because you have to know how many of them started out in the 9^th grade. What if I told you that I could compute the graduation rate without knowing that information? You would say I was daft, and you would be 100% correct.

Yet, that is precisely what EPE has done. Without any knowledge whatsoever of the history of today's graduates, none whatsoever, EPE essentially "makes up" the academic history of today's graduates. By only knowing how many graduated this year, and nothing else about them, EPE publishes graduation rates for every state in the nation to a tenth of a percentage. VaVoom, now that's magic.

Well, this magic is not real and the graduation rates published by EPE cannot be taken seriously. EPE computes graduation rates using what it calls a cumulative promotion index (CPI) which uses enrollments in a school or a state for this year and last year only. EPE claims that the formula it uses to compute graduation rates, but it does not. The formula for calculating 2005 graduation rates is: (see chart several paragraphs down)

Now hold your hat because the following true statement will blow your mind: Regardless of whether the class of 2005 started out in the 9th grade with 100, 500 or 1,000 students, it would not affect EPE's graduation rate for the class of 2005. That is, if EPE computed the graduation rate for the class of 2005 to be, say, 65%, the EPE graduation rate would still be 65% whether that class began in the 9^th grade with 100 or 500 or 1,000 students. This should make you dizzy because it is so unbelievable. EPE's 2005 graduation rate depends only upon the number of students in the high school in 2004 and in 2005. Thus it is preposterous that EPE advertises its numbers as actual graduation rates. They are not and EPE should make clear to the public that they are only EPE's version of the probability that a 9^th grader will finish high school. Yet EPE publishes "graduation rates" for cities, states and the nation to the tenth of a percentage point. How incredible.

Let's see what this means in the real world. Suppose that 100 students began in the 9^th grade of our hypothetical high school and 90 finally graduated four years later. The actual graduation rate, therefore, is 90%. But let's see what EPE calculates:

The fact is that EPE's graduation rates may have little or nothing to do with the real world. Nevertheless, EPE claims them as "actual" graduation rates which are unfortunately being accepted as gospel by the news media.

To demonstrate how far-fetched, misleading and inaccurate are EPE's numbers, EPE estimated the national graduation rate for whites and blacks to be 71% for whites and 55% for blacks (1). On the other hand, the U.S. Census Bureau periodically samples the entire U.S. population to determine what percentage of 18-24-year olds have completed high school. For 2004, it found that 91% of whites and 83% of blacks had completed high school, some of whom held GED's (2) And the American Council on Education estimated that in 2000, 87% of whites and 79% of blacks had high school diplomas excluding GED's.(3) One might posit that the rates were even higher in 2004 since graduation rates have been rising year-after-year.

Of course logic would dictate that EPE's claims of graduation rates are not meaningful due to the fact that they have nothing to do with the actual academic history of students and are only probabilities. I conclude with the final straw from EPE's philosophy and, perhaps, a clue to why EPE has gone so obliquely off track. In a statement that defies reality, EPE states, "[EPE's CPI] represents graduating from high school as a process rather than a single event." No!  Graduation is, in fact, a single event---it either happens or it does not! 

gined@epe.org and insist that EPE stop parading these probabilitiesas actual graduation rates and to tell the public that these numbers are simply EPE's attempt at estimating the probability of a 9^th grader finishing in four years, and nothing more.

(1) http://www.edweek.org/ew/articles/2008/06/05/40execsum.h27.html

(2) http://nces.ed.gov/pubs2007/dropout/Table2.asp?Table=9&referer=ListOfTables&Error=B-8

(3) http://www.acenet.edu/bookstore/pdf/GEDASR06.pdf

Dr. Solomon is a retired University of Kentucky Professor and can be reached at mbsolomon@aol.com

Published June 26, 2008