As reviewed by Jay P. Greene
Diane Ravitch’s new book, Reign of Error, is obviously not a work of scholarship, nor is it intended to be. The first half of the book is a rehashing of arguments against “corporate reform” that were mostly presented in her previous book, The Death and Life of the Great American School System. And the second half makes the case that until a wide variety of social ills are addressed, it is unreasonable to expect much improvement from the traditional public-school system. Only after the revolution can real progress be achieved.
Ravitch, who has long been an effective polemicist, must have felt increasingly irrelevant and ignored over the last decade, as rigorous quantitative analyses, which she is not capable of producing or even understanding very well, increasingly displaced clever rhetoric as the primary mechanism for influencing education policy. She has, at least temporarily, regained the spotlight by appealing to a new audience indifferent to the standards of quality social science. With this new audience in mind, Reign of Error is not designed to be a thoughtful and balanced piece of social science. It is meant to be a call to arms, to rally supporters and inform them of Ravitch’s views so that they are more likely to prevail in policy disputes. But the book is not even that. Given how tendentious and hyperbolic her arguments, and how selective and distorted her reading of the evidence, this book is unlikely to influence any policy discussion. It speaks only to those already converted to “the Cause.”
Reign of Error should be understood as a form of therapy. It soothes the outraged educator by articulating that anger and giving it legitimacy. And educators have some reason to be outraged. They are losing autonomy over their daily work life, as control over education is increasingly centralized and politicized. They are under pressure to teach according to some script meant to increase performance on standardized tests. They feel threatened with consequences if those test results are not favorable. They see young twerps from elite colleges with little classroom experience assuming positions of power in state departments of education and cash-gushing foundations.
Reign of Error is a venting of collective anger, but it is not a productive catharsis. Ravitch is so reckless in her interpretation of evidence that she and anyone citing her would lack credibility in policy discussions with those possessing a passing familiarity with the research. The selective and faulty reading of evidence is so pervasive in Reign of Error that it would take a volume of equal or greater length just to document and rebut all the instances of it.
Let me illustrate by highlighting one example: how Ravitch distorts the evidence on private school vouchers. More than a dozen published analyses of random-assignment experiments reveal the effects of winning a voucher in a lottery on educational achievement and attainment. In the chapter in which she claims “no evidence” that vouchers are beneficial, she mentions only one of these experiments. And when she describes the results from the federally funded Washington, D.C., experiment, she focuses only on the fourth-year achievement results, which showed a positive effect but fell short of the conventional standard for statistical significance by having a p = .06 instead of .05. Ravitch does not share that detail. Nor does she mention that after the third year, the D.C. program did produce statistically significant gains before sample attrition made the same effects more difficult to observe with confidence in the fourth year.
More importantly, she ignores the large and statistically significant improvement in high-school graduation rates resulting from vouchers in that same study. If she included the other random-assignment studies, her readers would learn that only one shows null results, and the rest demonstrate significant benefits, at least for African American students. For example, receipt of vouchers to attend a private school in New York City significantly increased high-school graduation and college-attendance rates for African American students.
Ravitch addresses only one other study in her review of the evidence on vouchers. This matched-sample analysis from Milwaukee shows significant gains in high-school graduation rates for students receiving vouchers. She tries to discredit those results by noting the high rate of attrition from the voucher program during the high school years. What Ravitch does not understand is that this is an “intention to treat” analysis, in which all students who started in private schools via the voucher program are counted as if they had remained there, even if they transferred into public high schools. She writes, “This very high rate of attrition very likely left the most motivated students in the voucher schools and certainly raised questions about whether the voucher program had any effect.…” But in this type of analysis, outcomes for even the possibly less motivated students who transferred to public schools for some of their high-school years would still be credited to the voucher program. This means that the Milwaukee attainment analysis almost certainly underestimated the benefits of remaining in private schools throughout high school by diluting the private school group with students who transferred to public schools. Intention to treat is the conventional and appropriate type of analysis; it is designed to produce a conservative estimate of effects. Ravitch does not understand the direction of the potential bias from attrition.
For anyone who knows the research literature, reading Ravitch is downright infuriating. But her devoted followers couldn’t care less. She gives voice to their suffering and crowns their preferred policy positions as the ones supported by “evidence,” so she must be right. This raises questions about Ravitch’s earlier historical scholarship. Was Ravitch the darling of the Right during the 1980s and 1990s when she attacked multiculturalism and progressive education because she fairly and exhaustively described the historical record or because she just drew conclusions that the Right preferred? With historical research, it is more difficult for the general reader to check the original materials to see whether authors have been selective or distorted in their interpretations.
One suspects that Ravitch was once more careful, but it is hard to be sure. Her writing was certainly better in the past, with a more even-tempered tone and greater nuance. Reign of Error reads like a string of her hyperventilating blog posts. Her Twitter obsession, launching 140-character missives on average every 46 minutes of her waking life, has reduced her prose to a preponderance of short, overly broad declarations with a good deal of contempt for disagreement.
I’m sure Reign of Error will be a commercial success. It will sell plenty of copies for its publisher and yield high-fee lectures for its author. But in scholarly and policy terms, this book is a failure. It offers readers little more than primal-scream therapy. Aggrieved teachers deserve a better champion, one who can provide a fair and comprehensive reading of evidence.
Jay P. Greene is professor of education reform at the University of Arkansas.