A presentation at the 18th Education Trust National Conference, Nov. 9, 2007, Washington, D.C., by E. D. Hirsch, Jr.
© 2007 Core Knowledge Foundation. Not to be copied or reproduced without permission from the Core Knowledge Foundation, 801 E. High Street, Charlottesville, VA 22902.
I am grateful to the Education Trust for inviting me to give this talk. It’s an honor and a kind of homecoming. We at Core Knowledge feel great affinity with The Education Trust with its focus on narrowing the unfair achievement gap between groups. That injustice was my reason for leaving academic pursuits and entering education reform in the 1970s.
I won’t distract you with the intricate details of my experiments on literacy some 35 years ago beyond observing that they were first done at U VA, and then at a mainly African-American college in Richmond. I described the results in two technical publications that are virtually unknown. But they have colored all of my subsequent work. Anyone who bothers to read those reports might be surprised to discover that it was empirical science and not ideology that originated Cultural Literacy and the Core Knowledge movement. The ideological controversies surrounding Cultural Literacy during the 1980s and ’90s were gripping but, to my dazed mind, essentially off point. For, the key educational issues we faced urgently both then and now are less connected with ideology than with empirical reality.
I’ll very briefly describe the discovery that shocked me into education reform. The African-American students at the Richmond college (It was the Sargeant Reynolds Community College.) could read just as well as UVA students when the topic was roommates or car traffic, but they could not read passages about Lee’s surrender to Grant. Their performance on that particular text shook me up the most. For they had graduated from the schools of Richmond, the erstwhile capital of the Confederacy, but were ignorant of the most elementary facts about the Civil War and other basic information that is normally taken for granted in writing. They had not been taught the various things that they needed to know to understand ordinary texts addressed to a general audience. The results were shocking. (What had the schools been doing???). I decided to devote myself to helping right the wrong that is being done to such students.
Let me explain my title: “Narrowing the Two Achievement Gaps.” The sort of gap usually meant by the phrase “achievement gap” is the one between whites and African Americans or whites and Hispanics, or more generally between high- and low- income students. Let’s call this “the fairness gap.” But there is an equally fateful achievement gap between our students and those in other developed nations. Let’s call this “the quality gap.” My first theme in this talk is that these are not separate problems. The solution to the fairness gap is also the solution to the quality gap, and vice versa.
I will focus on the verbal achievement gap, which is critical to academic performance, later income, and general competence. I want to show that if we raise the average verbal achievement for all groups of students we will, by that very deed, also narrow the fairness gap, killing two birds with one stone.
There were already inklings of a correlation between educational quality and gap closing in the famous Coleman report of 1966, called Equality of Educational Opportunity, in which Coleman and his colleagues showed that good schools had a greater positive effect on disadvantaged students than they had on advantaged ones. Really good schools, he showed, are inherently compensatory. According to his massive data, a school that achieves higher scores for advantaged students also narrows the fairness gap for disadvantaged students. So much emphasis has been placed on Coleman’s finding about the importance of family-background that this correlation between school quality and social equity has been widely overlooked.
We can see further evidence for the quality/equity correlation in recent international studies. The OECD, the organization of developed nations, has issued a report showing that those nations which achieve the highest verbal scores are also the ones that most effectively narrow the verbal gap between groups of students.
In this chart, gap-narrowing is measured from left to right, and verbal scores from bottom to top. You will see that the best-performing nations, Finland, Canada, Ireland, Korea, Japan, and Sweden not only score high verbally but also score well — off to the right — in reducing the gap between schools and between students within schools. Low scorers like Germany and Luxembourg tend to be shifted to the left. The USA is in the middle in average verbal scores, and in left field with regard to equity. There are some interesting exceptions to the overall pattern, but the general trend is from bottom left to top right. Lower-scoring nations tend to offer less equal educational opportunity, and high-scoring nations tend to offer more.
But let’s quickly leave international comparisons. They are unpersuasive to the majority of Americans, who say that these other nations are not truly comparable to the United States — our apples being unlike their oranges. I had the following thought. Why not look at individual states within the United States, and not compare the states directly to each other, since that would lead to the same apples-oranges objection. Why not compare each state to itself by looking at its NAEP reading scores, first in 1998 and then in 2005 to see whether a rise in quality also produced a rise in equity.
Now, according to NAEP, only 10 states showed any significant change in eighth-grade reading scores between 1998 and 2005. Three of them rose significantly — Delaware, Massachusetts, and Wyoming. And seven showed significant verbal decline: Connecticut, West Virginia, New Mexico, Arizona, Nevada, North Carolina, and Rhode Island. So, let’s see whether the states that raised average verbal scores also narrowed the achievement gap between the 25th and the 75th percentiles of students. And let’s also see whether states that declined increased the gap between the 25th and 75th percentiles. The answer is an emphatic “yes” to both questions.
A statistical analysis of this quality/equity nexus shows a .92 correlation and a probability to the .0001 level — a near certainty — that there is been a non-random connection between a rise in quality and a rise in equity.
These are sensational results supported by a number of further statistical analyses. Delaware seems to be on top, because it changed most. But while it can be very proud of itself, it did start from a very low point in 1998. The hero of the chart is Massachusetts which rose from being in the mid-top range of states to being at the very top in absolute verbal scores — without neglecting low-end students. Massachusetts can of course do better on both metrics, as can all the states. But, right now, if you are poor, or black, or Hispanic and have a school-age child, the state to be in is Massachusetts, where both the average reading scores and also the scores of the lower-percentile students are higher than those of any other state. Coleman’s finding of a connection between quality and equity is thus clearly reconfirmed both among the OECD nations and within the various states.