Response to Robert Gordon's cover story in The New Republic

As a recovering journalist who has spent the last five years working on education reform, first in the Atlanta Public Schools and then at the state level in Georgia, I agree with Robert Gordon's recommendations in TNR to develop high-quality national tests (a la NAEP) and to focus on improving teacher quality. The latter goal is less likely to be reached by selective pay raises--Gordon's principal suggestion--than by improvements in three complex systems: teacher training, principal certification, and curriculum development. The policies and practices of all three systems are based more on conventional folk wisdom than on rigorous empirical evidence of what works and doesn’t work to raise student achievement. Indeed education today is where medicine and public health were in the 18th century–in the pre-professional stage. Grover Whitehurst, head of the federal Institute for Educational Sciences (IES), estimates that only 10 percent of decisions in education today are based on rigorous data–-the other 90 percent are based on unwarranted assumptions. The great majority of education’s products and
practices–-from reading curricula and math pedagogies to science textbooks and ed-leadership degree requirements–-are developed, approved for use and implemented without adequate field-testing or other evidence to prove their efficacy.

This pre-Copernican way of doing business has survived for generations for two reasons: the adults who benefit from it financially (and psychologically) are more numerous and powerful than the students who are harmed by it, and the consequences of the system’s “failings” are less dramatic than are the failings of medicine, public health, the airline industry or any other
competitive, high-stakes profession. We care more about surviving surgery and arriving safely on our flight from NY to DC than we do about whether someone else’s child learns to read. This is only natural.

One thing a Democratic Party hungry for ideas might want to propose is the creation of an independent federal consumer-protection entity modeled on the FDA that establishes guidelines for curriculum product development and standard protocols for the teaching of basic skills such as reading and math. States would be free to ignore these standards but would lose federal
funds if they did so. States embracing the standards would receive extra money and technical assistance to help realign their systems of teacher and principal training and certification; textbook evaluation and adoption; and student monitoring and evaluation to make all of them more evidence-based. In a spirit of bipartisanship, the Democrats could embrace the IES’ What
Works Clearinghouse while noting that it lacks teeth, is underfunded, and needs a bigger educator-training component. Democrats could likewise embrace promising, federally-funded alternative teacher training/certification regimes such as the American Board for Certification of Teacher Excellence (ABCTE) and push for their adoption at the state level, where resistance
(bipartisan) has been stiff. They could unambiguously embrace the findings of Project Follow Through, the largest study ever conducted on approaches to reading instruction, whose results (direct instruction works best by far) were disowned by the Nixon administration after pressure from educators whose approaches (including whole language) fared poorly. More generally,
Democrats could send the message that they are on the side of science and against junk science in education, and on the side of parents who care, students who work hard, and educators who do what works. They and only they deserve greater federal support.

The unions need not be a fatal obstacle. There is plenty of money in the system to keep a lot of people happy-–the challenge is to steer that money increasingly toward those activities and products that actually increase student achievement, and toward well-designed R&D that can identify same. Right now, the vast majority of funds are spent to support poorly designed ed research; teacher training programs that are based on unfounded theories; and, in effect, experimental treatments using curricula and teaching techniques that have never been shown to work, particularly for the bottom 40 percent of the student population. Either we need to pay and train the same players in the game to do what works, recruit new players to the game
(ie. via charter schools, alternative training and certification programs, vouchers, etc.), or do as much as is politically feasible of both.

In any case, we will need both Democrats and Republicans to sustain this effort. Here's hoping Gordon's piece is the first of many that ups the ante for both parties. Politicians competing to help kids: wouldn't that be nice?