(Guest Post by Matthew Ladner)
On March 9, 1974 Yoshimi Taniguchi, a Japanese book merchant and former Major in the Japanese Imperial Army, traveled to the Philippines in order to order a former subordinate, Second Lieutenant Hiroo Onoda to stand down from combat operations. Lieutenant Onoda had received news of Japan’s surrender in World War II, but had concluded that it was mere enemy propaganda. Firm in this misapplied conviction, Onoda carried on a long since concluded war for almost 30 years.
Onoda may have misspent three decades in an island jungle, but on a positive note he at least inspired an episode of Gilligan’s Island (Ginger in fact defeated his doppelganger in a judo battle). More disturbingly, he may have also passed his die-hard spirit on to NCLB’s last stalwart defenders of achievement gap mania and the dream of mandated perfection.
I should begin by saying that I do believe that achievement gaps are very important. I went back to review a post I wrote a blog post three years ago called In Defense of Achievement Gap Mania and found nothing that I had changed my mind about on this subject. Getting low-performing American students on a faster academic pace is of the utmost importance.
I am however mystified by the Onoda-style defense of NCLB’s unworkable division of schools into multiple subgroups with targets for narrowing and ultimately closing all achievement gaps on the equivalent of a train schedule. That would have all been enormously beneficial if it had worked, but let’s just say that the trains weren’t showing much sign of arriving on time, sometimes at all. We could discuss various ways states found to escape from the NCLB subgroup noose at length (hello n-size manipulation!) but there are deeper problems to discuss.
Today we find NCLB diehards sprinkled throughout the K-12 reform conversation archipelago. In this recent post on Eduwonk, Anne Hyslop goes Banzai! corrects factual problems in a recent New York Times story focusing on the flaws of NCLB. My own take on this is that Hyslop is probably completely correct in her assertions. I however believe that they are largely irrelevant to the bigger picture. When I read a paragraph explaining how technical mumbo jumbo safe harbor confidence intervals actually mean that the mandated 100% proficiency mandate doesn’t actually arrive until 2016 instead of in 2014, it makes me chuckle. Do we imagine that when educators read the fine print they will rush to embrace NCLB’s machine of mandated universal proficiency with open arms? Or would it be more accurate to say that many educators never took AMO schedules seriously in the first place, confident that they would be dropped (they were- through the waiver process).
Ed Trust released a report recently critical of NCLB waivers. I personally don’t like NCLB waivers either given the Secretary’s lack of obvious authority to grant conditional waivers, but that is not what has Ed Trust excited. They think that the Secretary ought to have used his non-existent conditional waiver authority to mandate gap closing measures into state accountability systems.
Stand down Lt. Po…that’s an ORDER!
Ed Trust is careful towards the end of their study to say that they are NOT calling for a return to NCLB’s multiple pathways to failure based on myriad subgroups in pursuit of mandated improvement on a schedule. Hey its 2014 and they aren’t that crazy, you see, they just seem to want Secretary Duncan to work something out with these states that will be the functional equivalent of NCLB’s multiple pathways to failure based on myriad subgroups in pursuit of dramatic improvement by mandate on a schedule. That would do just fine.
Ed Trust (and others) seems sufficiently wedded to NCLB-era mechanics that they dislike an elegant improvement-the super subgroup. Florida policymakers grade schools half on proficiency (the % passing state exams) and half on student progress over time. They double-weighted the importance of the progress of the lowest performing 25% of students from the last year’s test. The students falling behind thus constitute the “super subgroup” and they became the most important students in the building for determining a school’s grade. They count against all three of the main three components of a school grade: overall proficiency, overall growth and the growth of the students who have fallen behind.
The super subgroup doesn’t ask whether you are White, Black, Asian, American Indian, economically disadvantaged, an English Language Learner or a child with a disability, blue-eyed or left-handed. It simply identifies the lowest performing children in the school and puts a special emphasis on their academic gains over time. It doesn’t create a perverse incentive to ignore an academically struggling child because he or she happens to be White, or because his or her parents make a little more money than this year’s Free and Reduced Lunch standard or because you’ve been reclassified out of SPED. Done properly, super subgroup creates a powerful incentive to identify struggling students regardless of their appearance and/or circumstances and get them making progress.
Oh, and the Ed Trust’s own previous research would lead one to the conclusion that it can help reduce achievement gaps. Not eliminate achievement gaps on a train schedule, mind you, but to make substantial progress on them by creating an incentive for schools to get struggling students to catch up. Who are the kids struggling? Why it is the kids on the short end of the achievement gaps as a matter of fact. Florida kept A-F school grading up over a good period of time and you see gaps narrow in the best way possible- bottom scoring kids making greater progress than the still progressing top scoring kids.
The Ed Trust report tut-tuts things like Black students in A graded Florida schools scoring lower than White students in C graded schools as evidence that we ought to be including gap closure in state accountability systems.
Should we work ourselves into a froth about this? I personally don’t think so. Ed Trust’s own research has documented Florida’s overall progress in narrowing achievement gaps, but it’s not like they have eliminated achievement gaps. Would anyone be shocked to learn that ELL students at A graded schools score lower than non-ELL students at C schools? What about children with disabilities? Low-income children? Ed Trust focused only on three states, but you could find similar results in any state.
The super-subgroup mechanism creates an incentive to get all students who have fallen behind to make academic progress. A moment of reflection regarding grading schools based upon various achievement gaps would give any thoughtful person pause. Do we really want to bake perverse incentives to stall the progress of high performing students into state accountability systems? Under the super sub-group, schools have any incentive to get any child that has fallen behind back on track. If states began rating them based on trends in achievement gaps, they could create perverse incentives to ignore their plight if they happened not to have a disability, or if they were a native English speaker, if their family made too much money for a free or reduced lunch, or if they were White.
Against this backdrop, the Ed Trust report seems strategically vague- not in favor the NCLB AYP system, but vaguely in favor of including achievement gaps in state grading systems. The fuzzy nature of these recommendations deftly avoids discussion of how to avoid creating cringe-inducing perverse incentives.
We live in a nation where Black and Hispanic students score closer on PISA to students in Mexico to those in South Korea or their own Anglo classmates. Mexico, btw, has far greater poverty and far lower public school spending than the United States. This is sickening, but we should exercise good judgment in addressing it. Previous Ed Trust research and the NAEP both show that it has achieved commendable gains in narrowing achievement gaps in Florida. In the country as a whole, not so much.
Thinking more broadly, we should recognize the NCLB era as a decentralized learning process. While NCLB created a general accountability rubric, many states had already created accountability systems of their own, creating the opportunity to learn from variations in policy approaches. Florida paid far more attention to school grades than to NCLB’s AYP and achieved greater than average gains among traditionally disadvantaged student groups. I’m not a fan of conditional waivers, but we need to study and learn from the successes and failures of the diversity of approaches as best we can as well. It is understandable that there are many with a deep investment in NCLB, but we should not allow that attachment to blind us to something more effective at achieving its aims. The importance of achievement gaps should lead us to adopt the most effective methods for reducing them rather than pining for the ones we had hoped would eliminate them in short order.