For the past six years, Education Secretary Arne Duncan has pressured states to use standardized test scores of students to evaluate their teachers. He has liked the idea so much that he made it a condition for states who want certain pots of federal funds as well as federal relief from the worst mandates of No Child Left Behind.


Secretary of Education Arne Duncan speaks in the East Room of the White House on Nov. 19, 2014. (Brendan Smialowski/AFP/Getty Images)

Desperate for the money, some states created “accountability” systems that evaluated teachers on the scores of students they didn’t actually have, and with scores from subjects they did not teach. In one urban school system, every adult in the building was evaluated for a few years in part on student test scores, including the custodians and the lunch servers. When one state (Florida) required a boy who was born without the cognitive portion of his  brain to take a version of the state’s standardized assessment test, the U.S. Education Department said nothing. (Really, really and really.)

During all of this time, various assessment experts — including the American Statistical Association, the Board on Testing and Assessment of the National Research Council of the National Academy of Sciences and the Educational Testing Service’s Policy Information Center — have warned that using “value-added” measurements that link test scores to individual educator evaluations is not reliable or valid. But the Education Department and Duncan carried on anyway with their support of “value-added measurement,” which purports to take student standardized test scores and plop them into a complicated formula that can supposedly determine the “value” a teacher has added.

Just as the 2014-15 school year was beginning this past August, amid a growing anti-testing rebellion across the country among parents, students and educators, Duncan wrote a post on his department’s Web site that told teachers he was listening to them. Duncan told them that he “shared their concerns” about too much standardized testing and test prep, and that he believed “testing issues today are sucking the oxygen out of the room in a lot of schools.” He said he would give “flexibility” to most states that wanted to wait a year, from the current school year to 2015-16, to comply with a federal requirement that they use test scores to evaluate teachers.

Then, in October, Duncan and Obama separately praised an announcement by two organizations representing top education officials in each state and major urban school districts that testing policies across the country would be evaluated and redundant and/or low-quality exams would be dropped.  It should be noted that the statement on assessment by the Council of Chief State School Officers and the Council of Great City Schools has no binding requirements and the organizations have no enforcement mechanism, but Duncan  wrote an op-ed  published in The Washington Post in support of it anyway. In that piece, he conceded that some students are being overtested and that “policymakers at every level bear responsibility here — and that includes me and my department.”

Was the man who had for six years been pressuring states to use standardized test scores of students to evaluate teachers really backing off?

The short answer: No.

Duncan told teachers — before the mid-term elections, it is worth noting — that he heard their concerns that his education reform policies have resulted in obsessive testing in public schools, but he hasn’t taken any steps at the federal level to reverse course. When state and city education officials announced that they would review testing policies, Duncan said the federal government would offer “support” and “technical assistance.” Nothing more.

It is true that some states have imposed more standardized tests than mandated by the federal government. But if he really believed his department was responsible in part for the overtesting of children, shouldn’t he be doing something at the federal level to stop it? I recently asked the department about why Duncan didn’t advance his own proposal for states and districts to evaluate the amount of testing, and this was the response, from Massie Ritsch, then the acting assistant secretary for communications and outreach:

In an area — testing — where redundancy is so often a problem, developing duplicative recommendations for how to do assessment well didn’t seem to add value when the leaders of state education systems and the nation’s largest school districts had already been so thoughtful on the topic. As Secretary Duncan said at the time of CCSSO and CGCS’s announcement, the Department will provide support and input where they think we can be helpful. We want to follow the lead of the people doing the work at the state and local level to ensure that assessment enhances instruction and provides useful information to educators and families.

Okay then.

In the latest test-obsessed move by the department, Duncan late last month released draft regulations for colleges of education that call for evaluating these programs in part on how well the students of their graduates do on standardized tests. Yes, Duncan wants education schools to be rated on how well the students of their graduates do on standardized tests, and he is linking some federal aid if the graduates’ students don’t improve on their tests. As Sarah Blaine, a mom, former teacher and attorney wrote here:

Never mind the myriad chains in the causation link between the program’s coursework and the performance of its graduates’ students (presumably on standardized tests). Duncan somehow thinks that he can proximally — fairly — link these kids’ performance not just to their teachers (a dicey proposition on its own), but to their teachers’ prep programs.

Apparently Duncan can magically tease out all other factors, such as where an alumna teaches, what her students’ home lives are like, how her students’ socio-economic status affects their academic performance, the level of her students’ intrinsic motivation, as well as any issues in the new alumna’s personal life that might affect her performance in the classroom, and, of course, the level of support provided to the new alumna as a new teacher by her department and administration.

As any first-year law student can tell you, Duncan’s proposal is preposterous, as the alumna’s student’s test results will be so far removed from her teaching program’s performance that ascribing proximate causation from the program to the children’s performance offends a reasonable person’s sense of justice.

Apparently not everybody’s.