William J. Mathis
“As I see it,” wrote Paul Krugman, “the economics profession went astray because economists, as a group, mistook beauty, clad in impressive-looking mathematics, for truth.” Krugman himself is, of course, an economist (and a Nobel Prize winning one at that) which demonstrates that economists can indeed spot the pitfalls of their field. As an educator, when I review economists’ education reform ideas they, all too often, show manifestations of Krugman’s syndrome. They confuse mathematical symmetry with truth.
When applied to education, the pathology takes three identifiable forms: unitheoryia, measuremyopia, and regressionia.
Unitheoryia manifests itself in the narrow vision of education as an economic free-market commodity. The afflicted reflexively turn to privatization and choice as ways to improve the quality of schooling. Those in advanced stages tend to explain all human and social behavior as the pursuit of self-interest through competition. Eruptions about international competitiveness being dependent on standardized test scores are a common symptom.
The debilitating aspect of unitheoryia is that it results in a complete misreading of the purposes of schools in a democratic society. In checking these purposes in the fifty state constitutions, there’s a remarkable absence of clauses centered on “education for economic competition.” Instead, I found phrases like “Intelligence and virtue being the safeguards of liberty and the bulwark of a free and good government . . .” (Arkansas), “Knowledge and learning, generally diffused throughout a community, being essential to the preservation of a good government; . . .” (Indiana), and “. . . the encouragement of virtue and prevention of vice, . . .” (Vermont). Virtually every state also has an equal protection clause, which is quite the theoretical obstacle when extolling the values of competitive economic Darwinism.
Measuremyopia is characterized by a fixation on indicators that are easily quantifiable and reliable. For traditional economics, that indicator is money. For schools, this leads researchers (of many persuasions) to worship at the altar of the standardized test score. The intellectual rationale seems to be, “It might not be a valid measure but it’s the best we got and nobody can say we didn’t uniformly apply it!” Sometimes, a researcher afflicted with measuremyopia will throw in graduation rates or attendance data, since these are also quantifiable.
Constitutional phrases like “civic virtue” and “preservation of a democratic government” are simply beyond the conceptual consideration of most economists. These terms are too value-laden to be represented by well-behaving mathematical variables. Furthermore, the effects of schools, parents and communities are all inter-related in a tangled way. (Some acute observers may point out that test scores actually present the same problem, but this truth is forbidden in the orchid hothouse of economic analysis).
Sometimes, the real success of a school is in the shy child who bloomed and became a community leader, the artist who transformed a city-center, the needy child who found a positive role model, or the quirky kid whose creativity later led to a medical break-through. The fundamental flaw is that the really important things about schools (and about life) don’t lend themselves to economic analysis.
Regressionia reveals itself when, to the exclusion of all else, the afflicted seek a statistical model to understand schools. This often leads to the use (and misuse) of various forms of a technique known as linear regression, applied in infinite color and hue. (Regression is a correlational approach based on how variables change with each other). Beta weights, dummy variables, hierarchical models, and regression discontinuities are paraded about in formulated exotica ornamented with Greek symbols. It’s all very impressive and creates the impression that truth can be found here.
Sadly, this over-reliance on regression models is encouraged because it is so easy to do. No knowledge of the subject being studied is required. You don’t actually have to go inside a school. It only requires a data base, a packaged statistical computer program, and a slice of the unitheory to be tested. Not surprisingly, economic (as well as other) studies tend to confirm their theories. Scientific American refers to the epidemic of “false positives and exaggerated results.”
One of countless examples is found in an Education Next article ballyhooing the successful effects of school competition. The economist authors found that having a private school within 1.1 miles of a public school increased test scores by 1.5% of a standard deviation. Such findings can easily be “statistically significant” in a state with 2.6 million students. However, from a policy point of view, these effects are trivial. This study hits the Trifecta of measuremyopia, using regression to force-fit a unitheory. Does the study add anything to the policy discussion of schooling other than a “beautiful” model (to paraphrase Krugman)? Not really.
Fundamentally, the problem with regressionia is that none of the variables are particularly good indicators of what they are supposed to measure. For example, test scores are only a small part of schools and seniority is not a good measure of teacher quality. Further, the basic model assumes that the world is linear, relationships remain static, and that groups and individuals behave in predictable ways. This means that the exercise is often modern day numerology. The result is that a really robust regression equation might show school effects accounting for only a fourth of the variance. Now a curious person might want to know in what black hole the other three-fourths of the variance can be found.
Cohesiveness, cooperation, altruism, caring and the common good are vital to a democratic society. But these are alien concepts to economists’ visions of school reforms. Thus, the richness of what education is and should be is unseen, oblivious to e.e. cummings message “While you and i have lips and voices which are for kissing and to sing with, who cares if some one eyed son of a bitch invents an instrument to measure Spring with?
Be wary of economists bearing reforms.
William J. Mathis is the Managing Director of the National Education Policy Center, University of Colorado at Boulder. He previously served as a school superintendent in Vermont and is a member of the state board of education. The views expressed are his own.