Secretary Arne Duncan recently announced his plan to judge teacher education programs by their “results,” including the test scores of the students taught by their graduates. If the Ed Schools can’t produce teachers who can raise test scores, Duncan said, they should go out of business. Spoken like a true businessman.

Mike Rose, celebrated author and professor emeritus at UCLA, has six questions for Arne.

He writes:

Six Questions for Secretary Duncan

1. Will you be evaluating with the same metrics all teacher preparation programs, alternative as well as traditional, Teach for America as well as California State University at Northridge or UCLA?

2. If the Department of Education will use close to $100 million per year on grants to forward its agenda, where will that money come from? From other educational programs that serve needy populations? If so, what services or funding will be cut or discontinued because of this reallocation?

3. Policy formation emerges out of staff research, consultation with experts, and political deliberation. What research and consultation leads you to the current project? I ask because your statement about teacher preparation programs needing to improve “or go out of business” as well as your general approach echoes last year’s report from the National Council of Teacher Quality, a report that has been roundly criticized by a wide range of experts.

4. The National Academy of Education recently issued a comprehensive report on evaluating teacher education programs that recommends an approach very different from yours. Have you read it or consulted its authors?

5. There is an increasing number of respected scholarly organizations—the National Academies Board on Testing and Assessment, AERA, the National Academy of Education, the American Statistical Association—that are advising caution in the use of procedures like value-added to evaluate teacher effectiveness. These organizations point to technical, logistical, and conceptual problems in doing so. One conceptual problem imputing causality between teachers’ activity and a test score, for so many other variables come into play. Your stated plan will use student test scores to not only judge teachers, but also the institutions from which they come, introducing another level of questionable causal attribution in your model. You will have a putative causal chain that goes from the student test score to the teacher to the teacher’s training institution. How do you plan to address this basic conceptual problem?

6. The implication in your plan that bad schools will go out of business assumes that all prospective teachers are the economist’s idealized free agents who can go wherever a highly rated program exists. But a number of prospective teachers from lower income backgrounds do not have the finances to travel—or cannot travel because of family obligations and expectations. How will you address the possible unintended consequence of your program placing burdens on this segment of the population?

Thanks, Mike. If I hear from Secretary Duncan, I will post his answers.