Much as I would like Barack Obama to sprinkle magic dust all over the country, fixing racism, poverty, and absolutely everything we hated about the Bushies, each policy question will have to be tackled thoughtfully, one by one. Today's topic is the national education agenda.
A crucial issue here is the continuing mania for using public school children as a vast pool of customers for corporations specializing in both mass curriculum distribution and in the endless testing through which students -- on pain of humiliation, summer school, and being held back a grade -- are asked to regurgitate these educational products. (I use the phrase "educational products" consciously: currently, a standard curriculum in the United States is to education what Cheez Whiz is to cheese.) The sad backstory of test scores going up in any given school are the number of students who drop out, or are encouraged to drop out, because of their low test scores.
No Child Left Behind, built around a testing mandate, is a great example of "an expensive government program" (to use a distinction our new President made himself) that does not work and ought to be eliminated. While few federal tax dollars are spent directly on NCLB, its unfunded mandates put a huge financial burden on state and local education budgets. Under systems driven by high-stakes testing, students learn by rote, rather than becoming critical thinkers, many students drop out altogether rather than be left back a grade, and we are no closer to creating equal access to a good education than we were a decade ago. And yet, today's New York Times buries this issue with yet another assertion that the symbolism of an Obama presidency, rather than radical policy interventions, will fix what is broken in our nation by lifting up the self-esteem of African-American people. Our paper of record reports that education researchers tested a group of children before and after the presidential nomination. The test before the nomination demonstrated what is commonly called the "achievement gap" between African-American and white students. But in the test after the inauguration, presumably because their self-esteem was raised by watching an African -American man become president, the African-American students stepped up their game. On that second test, racial differences in the children's test scores were "statistically insignificant."
This, the researchers imply, potentially reinforces previous studies showing that white students do better on standardized tests because they believe that they will succeed; and that African-American students do significantly less well on the same tests because they believe they will do poorly. I am sure I am over-simplifying a large body of research, but as a historian, to design one's research around such a notion seems pointless, particularly since this longstanding belief has never done anything to actually make Americans a better educated people. I am reminded of Booker T. Washington's firm belief that education was useless to formerly enslaved people until they had acquired self-esteem through physical and moral uplift (personal hygiene, and specifically proper deployment of the toothbrush, was one way to develop self-esteem; another was to learn how to make an excellent brick.) Such arguments have been by no means solely the province of African-American social scientists and educators. Post-1945 white liberals like historian Kenneth Stampp were entranced by the notion that persistent social inequalities are a result of low self-esteem, rather than ongoing racial and class inequality. By the 1940's, self-esteem arguments had moved into progressive, anti-racist social science through the black doll-white doll test designed by influential African-American psychologists Mamie and Kenneth Clark, used as evidence in Briggs v. Elliott (1952) and other lower court cases folded into Brown v. Board of Education (1954). As Waldo E. Martin has noted, this test was not without critics on both sides of the issue, many of whom saw methodological flaws in the Clarks' experiments. However, the gravity of the moral issues at stake, Martin writes, meant that arguments about self-esteem resulting from the study nevertheless became powerful visual evidence in what stands out as a classic liberal decision by the Warren court.1
Brown, of course, eliminated de jure segregation, which was important. But it failed to address other educational questions that contribute to what we now call the achievement gap, because the scope of the decision, influential as it was, left undisturbed the connections between segregation, institutional racism and economic inequality. Furthermore highly theatrical demonstrations, filmed and widely distributed, of serious African-American children unerringly choosing the white doll, became one of the great reproaches to liberals of interventions too long delayed and compromises too often made, just as southern whites with hate-twisted faces spitting on African-American grade schoolers is a past that conservatives will have to endure as part of their history. To this day arguments about self-esteem dominate our discussions about the education available to children of color. While not entirely unimportant, they have a tendency to occlude other kinds of questions that should be asked about why our public education system is in such terrible shape.
In the spirit of full disclosure, we at la casa Radical wish that Obama had selected someone for this important Cabinet-level position like Stanford's Linda Darling-Hammond, who was briefly considered for the post and had been chosen to head up Obama's education policy working group. But things could have been much worse. Joel Klein, superintendent of New York City's school system was under consideration, and of course, there is always Michelle Rhee, a strong proponent of free market solutions for public education (read: siphon tax dollars to corporations) right outside the White House gates.
So it could have been much worse than Arne Duncan, formerly CEO (yes, that was his title) of the Chicago Public School system. Time Magazine recently characterized Duncan as friendly to several camps of school reformers. That is all good, and it will help him be effective in a highly contentious and ideology-driven professional world. But one of the challenges Duncan needs to be asked to deal with is the dominance of standardized testing as an assessment method. Under the Bush administration, high-stakes testing has been the centerpiece of education policy. Literally billions of education dollars now go into creating, administering, and grading these tests -- dollars that could go into the teaching of critical thinking; that could be keeping libraries open; that could be funneled into tutoring programs and smaller classes. Perhaps we could use a billion to provide regular dental and health clinics for children in each school so that they can see the blackboard and are not in physical pain in the classroom.
In conclusion, I would argue that the research cited by the New York Times is useful, but probably not in the way the researchers imagined. Forget self-esteem for a minute, as well as the problem of using the so-called achievements of white people (good test-taking skills) as the measure of intellectual success for everyone. If the inauguration testing experiment means anything, it certainly tells us something about the researchers. The resilience of assumptions about the importance of self-esteem in educational research design has caused them to leak information, as if it were an exciting new intellectual path, about a piece of research that even they admit is incredibly scattershot. But more importantly, this little experiment shows that testing children doesn't tell you what they actually know, and therefore, testing is failing at the most basic level that any policy or practice can be evaluated. If test scores are so dramatically affected by social and psychological factors like whether a child is excited that someone of his or her so-called racial group is President of the United States, then in fact, the one thing we know is that standardized tests are a bad tool that do not give us any reliable gauge of whether teachers are teaching well or students are learning.
Cross posted at Cliopatria.
1 Waldo E. Martin Jr., Brown v. Board of Education: A Brief History With Documents, Bedford Series in History and Culture (New York: MacMillen, 1998), p. 28.