Education Opinion

The “90/90/90 Schools” Myth

By Justin Baeder — May 30, 2011 3 min read
  • Save to favorites
  • Print


One of the most widely cited studies on closing the achievement gap is Doug Reeves’ “90/90/90 Schools” research, which asserted more than a decade ago (as reported in his book Accountability in Action) that schools can and routinely do get all or nearly all students to meet standard, despite extremely high poverty rates.

Given the growing consensus that poverty plays an undeniable role in the achievement gap—meaning the gap can be closed, but not without addressing poverty a la HCZ—I’ve been wondering what to make of the 90/90/90 schools. So I decided to do a little digging and see if there has ever really been any such thing.

Reeves defines 90/90/90 schools as those in which:

More than 90 percent of the students are eligible for free and reduced lunch, a commonly used surrogate for low-income families More than 90 percent of the students are from ethnic minorities More than 90 percent of the students met or achieved high academic standards, according to independently conducted tests of academic achievement

When pressed for actual examples of 90/90/90 schools, Reeves provided a list of schools in Milwaukee that had 90% or more of students receive a “basic” or higher score on a certain reading test, the Wisconsin Reading Comprehension Test, in the 1998-1999 school year.

It’s important to note the rhetorical context. Reeves went so far as to imply that it’s racist to ask for the names of 90/90/90 schools:

When The Bell Curve (Herrnstein and Murray, 1994) was published with the widely accepted assertion that children who are black and poor perform badly on academic achievement tests, I cannot recall a single instance of demands for the names of students who were subjects of the studies cited. When I have demonstrated that poor and black children perform well, I am inundated with demands for verification. These demands speak volumes about our expectations of children based on their appearance and economic status. from Accountability in Action

Very well then. We all want to believe that all students can achieve at equally high levels, but experience bears out the truism that poverty impacts achievement. Reeves wants to tell us otherwise, using data from actual 90/90/90 schools.

But there’s a problem with Reeves’ data: “Basic” on the WRCT cannot by any stretch of the imagination be defined as a “high academic standard.” The state of Wisconsin defines “basic” achievement as

Somewhat competent in the content area. Academic achievement includes mastery of most of the important knowledge and skills. Test score shows evidence of at least one major flaw in understanding the academic content area tested.

In fact, it is expected that 80% of students in each school will be “proficient” or higher; the purpose of the WRCT was to identify students most in need of intensive support, not to set a bar for “high standards.”

To put this in context, an examination of the spreadsheet of all ‘98-'99 WRCT results shows that fully 347 of 1132 schools (over 30%) had not a single student scoring below basic. Another 199 schools had only one student score “minimal.”

In fact, only 195 of the 1132 schools with valid WRCT scores in 1998-1999 had fewer than 90% of their students score basic or higher on the test. In other words, Reeves set a bar for “high academic standards” that 83% of schools met.

Reeves could have used these data to make a good point: All kids can learn, and regardless of the demographic conditions in a school, all students can be expected to reach some important basic academic standards. Schools with similar demographics have widely varying WRCT scores, so surely there’s something to be learned from the more successful schools. We should always look for opportunities to improve.

If we want to those standards to signify high achievement, though, we’re going to have to look for different data to support the 90/90/90 claim in any place other than Harlem Children’s Zone.

Reeves’ belief that data and accountability can help all schools help all students reach high academic standards was codified in No Child Left Behind. The past decade in education has been a nation-wide experiment to see if Reeves was right. Apparently he wasn’t.

If there are other data that identify 90/90/90 schools (with the final “90" representing some meaningful standard of student achievement), I’d love to see them. More recent research, of course, has shown that very few schools (such as the KIPP academies) consistently outperform regular public schools in closing the achievement gap, after selection factors and other differences are taken into consideration, and none by huge margins.

Until then, I will have to conclude that Reeves’ 90/90/90 schools are a myth, calculated to instill hope in discouraged educators and to sell books explaining how our schools can single-handedly overcome all odds.

The opinions expressed in On Performance are strictly those of the author(s) and do not reflect the opinions or endorsement of Editorial Projects in Education, or any of its publications.