The education policy community is abuzz about the scoring irregularities that surfaced in the Race to the Top competition. (Governor Bill Ritter of Colorado, angry that one reviewer tanked his application, memorably told the Times that ???It was like the Olympic Games, and we were an American skater with a Soviet judge from the 1980s.")
The New Teacher Project (TNTP) recently published a particularly exhaustive review of the ways that rogue raters and imprecise guidelines adversely affected the outcome. TNTP made several recommendations to the Department for Round Two, including providing additional training to reviewers, dropping the scores of outliers, and norming scores across panels (in case one panel is full of particularly tough graders). [quote] With billions of dollars at stake, these are all reasonable suggestions. But take special note of TNTP's last recommendation:
Remember Secretarial Discretion. Though the opinion of application reviewers must be weighted heavily in final decisions, Secretary Duncan retains final authority over awards and should be prepared to intervene in Round 2 if best attempts to strengthen the process are insufficient.
But "keep in mind," they write:
There could be enormous political fallout if the Secretary is forced to deviate from peer scores in selecting winners. This is a last resort.
You can say that again. In fact, we've seen this movie before, and it was called Reading First. The five-year, $5 billion Bush Administration program set out to change "Washington business as?? usual"--just like the Race to the Top. In his must-read Fordham report about the program, Too Good to Last: The True Story of Reading First, Sol Stern writes:
President Bush...stated that under his proposed reading legislation the Department of Education would be able to leverage its spending power to prod state education agencies and local school districts to choose reading programs backed by science and evidence??? a historic first for the?? federal government. As one of the meeting participants recalled, Bush said there was ???no need to throw good money into programs that don't work. We've tried that before.???
Sound familiar?
The conventional wisdom is that Reading First was brought down by the weight of scandal, with Bush Administration officials improperly steering dollars to preferred reading programs. Arne Duncan and his team--seeing the personal toll such allegations of scandal can take--wisely sought to avoid any such appearance in their implementation of Race to the Top.
That's understandable and largely appropriate.?? But it misses the real lesson of the Reading First story. My colleagues in the Bush Administration didn't do anything scandalous; they simply exerted their own judgment over that of a peer review panel. (They also tried to ensure that true reading experts were on said panel--the horror!) When the process didn't work--and states were on the cusp of getting federal funds for unproven reading programs--they intervened and made states re-apply. When they got word that districts were purchasing reading programs with no evidence of effectiveness, they took action. And guess what? There was enormous push-back from the vendors of shoddy reading programs to this noble effort to do what's right. And the bad guys won.
Duncan didn't have to make these sorts of tough calls in Round One because he was inclined to fund just a few states and keep most of the money in the kitty for next time. But I suspect in Round Two it won't be so easy. What if reviewers somehow manage to keep Florida and Louisiana out of the winner's circle again? What if they continue to mark states down for the lack of "stakeholder support" beyond what the guidelines allow? What if the reviewers simply get it wrong?
Over-ruling them should be a "last resort." But the option should be on the table. And if Duncan has to use it, we pundits should hold our powder before accusing him of using "the nuclear option" for political or scandalous reasons.
One last point: just as the Race to the Top process for judging states can be improved but not perfected, the same is true of systems for evaluating teachers. We can bring greater evidence into the process, but at the end of the day we can't replace human judgment. In the Race to the Top, that judgment is Arne's. And in the teacher quality sphere, that judgment should be the principal's. No statewide, from-on-high, formulaic evaluation system can cut it on its own.
-Mike Petrilli
