Testing Mania and Uncle Sam's Clumsy Reach
Last week, a lot of good people insisted that they are shocked--shocked!--to see that more than a decade of federal pressure to test has led ... to too much testing. The kerfuffle was triggered when the Council of Chief State School Officers (CCSSO) and the Council of Great City Schools (CGCS) issued a sensible joint statement on testing. The actual CCSSO-CGCS statement was pretty straightforward, declaring that assessments should be "high quality," "part of a coherent system," and "meaningful" for teachers and parents. (Not a lot to argue with there. I mean, I meet hardly anybody who insists, "I'm in favor of lousy, incoherent, meaningless tests!") The document got a fair bit of attention, though, because it acknowledged that testing has gotten out of control and called for states and districts to evaluate the tests they're giving, streamline them, and make sure they're useful.
The truth is that a lot of today's tests just aren't that useful. The Data Quality Campaign put it well in its reaction to the CCSSO-CGCS statement, observing, "The major backlash against student testing is because teachers and families are getting little value out of it. If a test is to be worthwhile, it needs to be producing information that's useful in classrooms and at kitchen tables." Parents generally don't mind when their child's pediatrician draws blood or does an examination, because they're confident that the purpose is to help their child. But, in schooling today, parents rightly intuit that much of today's testing has little to do with making sure their child is learning.
Unfortunately, Secretary of Education Arne Duncan responded on Sunday with a response that left something to be desired. Duncan, who has embraced testing in Race to the Top and in his NCLB waivers, took to the op-ed pages of the Washington Post to share his disappointment that there's so much testing and that, "in some places, tests--and the preparation for them--dominate the calendar and the culture of schools."
Duncan's defense of testing, however, boiled down to him explaining, "As a parent, I want to know how my children are progressing in school each year. ... The more I know, the better I can reinforce at home each night the hard work of their teachers during the school day." This is all well and good, in principle. The problem is Duncan's persistent unwillingness, in practice, to distinguish between tests that might be useful in this way and more than a decade of federally-mandated tests that are not. (I've no clue how a federally-mandated test each spring--the results of which generally aren't known until after the end of the academic year--helps parents support day-to-day classroom instruction.)
What Duncan and so many well-intended federal "reformers" have missed during the past decade--whether Democratic or Republican--is that Uncle Sam can't handle nuance when it comes to schooling. The federal government issues rules and regulations that try to make states make districts make schools do the right thing. By the end of this game of telephone, it shouldn't surprise that federal directives have been magnified, blunted, and garbled.
They're magnified because running afoul of federal rules and requirements can seem a career-threatening mistake to state and local officials, yielding fear-based overcompliance. They're blunted because rules have to be applied to 100,000 schools serving 50 million students, making it impossible for federal guidelines to have the precision that one might wish. They're garbled because requirements get confused as they're codified, bureaucratized, and translated from one level to the next, and as the original rationale gets lost over time.
What is Duncan's practical response to this state of affairs? He promises to "support state and district leaders in taking on this issue" and to "provide technical assistance." He celebrates ED's spending $360 million to create new Common Core tests (though he strategically avoids mentioning PARCC, SBAC, or the Common Core). And he congratulates ED for giving states more "flexibility" on using test scores to evaluate teachers (which they're only rushing to do in the first place because he made doing so a dubious condition for receiving an NCLB waiver).
Look, I believe testing can be good and useful. And, like everyone else, I'm for "smart" testing and against "too much" testing. This is all cause to be careful about issuing broad directives from Washington, where policymakers are far removed from the practical impact of their decrees and where problems and unintended consequences are tough to address in a nimble or timely way.