Arne Duncan’s useful exaggeration

This from Jay Mathews at WaPo's Class Struggle:

As my colleague Nick Anderson reported recently, Education Secretary Arne Duncan estimates that 82 percent of U.S. public schools will miss their academic targets this year as the federal No Child Left Behind law raises the bar higher than most schools can jump.

This was predicted by many experts when the law took effect in 2002. Its demand that nearly all students be proficient in math and reading by 2014 was hopeless, but unavoidable. Members of Congress knew they would be ridiculed if they passed a law saying that 15 percent, 30 percent, 50 percent---pick your favorite number—of children would not learn.

What is most interesting about Duncan’s estimate, however, is not that so many schools are unable to meet the law’s standards, but that so many experts insisted Duncan’s expected 82 percent failure rate for 2011 was wildly exaggerated. Only 37 percent of schools missed their targets last year. This year’s percentage will be more than that, experts said, but nowhere near that much.

What is going on is an old clash of statistics and politics.

What KIPP is Yet to Prove

Until KIPP tries to succeed within an entire, single community,

it is, for all its remarkable rise and deserved praise,

just another model program that has

yet to prove it can succeed with all

—or even most—disadvantaged children.

This from Sara Mosle in Slate:

The Educational Experiment We Really Need

What the Knowledge Is Power Program has yet to prove.

In his new book, Work Hard. Be Nice., Jay Mathews claims that the Knowledge Is Power Program is the "best" program serving severely disadvantaged, minority-group students in America today.

Let me begin—before I'm denounced as a traitor to the cause of educational reform—by saying that I'm inclined to agree. The improbable story of how KIPP was founded in 1994 by David Levin and Michael Feinberg, two young Teach for America alumni in Houston, is thrilling and worthy reading. KIPP's mission has been akin to putting the first man on the moon: an all-out education race, requiring extraordinary, round-the-clock dedication from parents, students, and teachers alike.

But the program is not the proven, replicable model for eliminating the achievement gap in the inner city that Mathews imagines, and this distinction is crucial. KIPP may be something more important: a unique chance to test, once and for all, the alluring but suspect notion that there actually is an educational panacea for social inequality. As of yet, the evidence for such a thing doesn't exist.

There have always been model school programs that work. There have even been some that have been successfully replicated in different parts of the country. But no program has shown it can work for all, or even most, disadvantaged children within a single city or neighborhood.

Instead, as critics point out, such model programs tend to skim off those kids who are already better positioned (thanks to better home environments, greater natural gifts, savvier or better-educated parents, etc.) to escape the ghetto. Meanwhile, regular public schools are left with a more distilled population of struggling students...

Mathews insists that KIPP has solved this riddle.

It's true that perhaps no other model program has risen so far so fast, with such consistently strong test scores. KIPP now has 66 academies in 19 states. Still, 66 academies amount to just three schools, on average, per state. Houston has far and away the highest concentration with, currently, seven middle schools, three elementary schools, and one high school. But this is in a school system with 200,000 students, nearly 80 percent of whom qualify for reduced or free lunches.

At the moment, like every other model program before it, KIPP serves only a tiny fraction of disadvantaged students within any given district. And as education researcher Richard Rothstein has rightly noted, comparing students from different schools, even within the same disadvantaged neighborhoods, is very difficult to do in a rigorous, scientific way. Just because KIPP uses a lottery for admissions, for example, does not tell us anything about the self-selecting nature of the pool from which this lottery is drawn. (Rothstein's own research—here and here—has shown that KIPP students come from families that are better off, or better educated, than their regular public school or special-education counterparts.) ...

KIPP teachers typically work 65-hour weeks and a longer school year. Recognizing that students need more out-of-school aid to supplement their educations, the program also requires its staff to be available to students by phone after hours for homework help and moral support. For this overtime (which represents 60 percent more time in the classroom alone, on average, than in regular public schools), teachers receive just 20 percent more pay. Unsurprisingly, turnover is high. The program has relied heavily on the ever-renewing supply of very young (and thus less expensive) Teach for America alums, whose numbers, while growing, are decidedly finite...