I would like to welcome you to AEI today for this conversation about No Child Left Behind--and whether it needs to be mended, ended, or simply given some time to work.  If you are here, you are well aware that NCLB is the most significant federal education legislation in decades, the hallmark domestic accomplishment of the Bush administration's first term, and an on-again, off-again topic in this fall's presidential election.

This session was occasioned by a piece that Checker Finn and I recently wrote for The Public Interest on NCLB--copies of which you will find in the folder on your chair.  It occurred to us that it might be constructive to hold a kind of open conversation on the same questions we address in the piece on where we are with NCLB, the challenges we are addressing, and where we go from here.

So I have asked a panel of smart, knowledgeable, and plainspoken people to join us, and I hope that the conversation may prove informative and useful. From your left, the panel includes Chester E. Finn Jr. of the Thomas B. Fordham Foundation, Joel Packer of the National Education Association, Ross Wiener of Education Trust, and Michael Petrilli of the U.S. Department of Education.

The way this will work is that each of us up here will speak for no more than 8-10 minutes and then we will open it up for questions.  I will speak for a few moments, then we will hear from Checker, Joel, Ross, and Mike, in that order.

Perhaps what is most interesting about NCLB, at least as a political scientist, is that it is viewed as a "Bush bill"--due to its strong support from the administration and its emphasis on accountability.  In truth, NCLB was not crafted as a coherent package--but is an amalgamation of ideas from left, right, and middle and is about as consistent with Al Gore's 2000 proposals as with the president's.  Ultimately, like any major legislative effort, it includes a slew of sometimes-discordant parts.

In fact, NCLB's towering promises, extensive implementation challenges, and expansion of federal authority call to mind Lyndon Baines Johnson's Great Society legislation--much more than the conservative critiques which faulted those efforts for insufficient attention to the realities of human nature and organizational behavior.

In this reenactment, the president is playing the role of LBJ--championing a heartening vision of equal opportunity while wrapping his agenda in the mantle of racial and social justice.   Playing the part of yesterday's War-on-Poverty critics are today's NCLB critics, warning of runaway federal authority, naïve expectations, and unfunded mandates.  Some critics go so far in flinging every imaginable charge, however silly, at the administration and at NCLB that they recall the demagogic willingness of the "Southern Manifesto" segregationists.

The academic field of "implementation studies" emerged in the 1960s and 1970s as scholars tried to explain the disappointing results of widely hailed Great Society programs.  They found that, too many times, Washington's money and direction were overwhelmed by real-world complexities.  The lesson we learned is that federal statutes don't work because of good intentions--but because they are sensibly designed and make measured use of mandates and incentives. 

Looking at NCLB

In that spirit Checker and I have turned a social scientist's beady eyes to NCLB.  Now, NCLB is a lot of things; heck, it is a telephone book.  But the engine of NCLB is its accountability and remedy provisions--that these are the pistons intended to drive school improvement.  So these are the pieces that we focus upon.

Let me begin by acknowledging the immense, critical achievement of NCLB--the ample sunshine that it beams upon student, school, district, and state performance in reading and math. 

Ultimately, Checker and I suggest that NCLB should be pushed back towards the tight-loose formulation that gave birth to the education accountability movement and that has revolutionized modern management.  Today, NCLB is too prescriptive about means and too hands-off about ends. 

NCLB dictates how adequate yearly progress is calculated, how schools are labeled, the timeline for improvement, and the imposition of remedies.  At the same time, this otherwise prescriptive statute is mute about what a state's academic standards should look like, laid back about how high the proficiency bar should be set, and agnostic about how students are tested.  In short, NCLB is highly explicit about the process of accountability while remaining quite lax about its substance. 

The rules invite finagled timelines, eccentric assumptions about how to measure improvement, and passing rates that appear to rise as "cut scores" quietly decline.  This "flexible" arrangement is likely to increasingly encourage state officials to flout the spirit of the law while nominally complying with its letter.

The reforms we suggest presume that Uncle Sam needs to set clear expectations for states, schools, and districts but then leave enough play in the joints that Washington bureaucrats are not routinely asked to make nuanced decisions about state or local educational practices. 

I will take a moment to suggest three general suggestions for retooling the accountability provisions, and then Checker will take up the subject of NCLB's remedies.

The National Assessment of Educational Progress

First, NCLB is today too lenient about the skills and knowledge that students must acquire and too prescriptive about calendars, state improvement targets, and school sanctions.  This is backward.  Washington should offer stricter guidance regarding the essentials that students are to master while being flexible about how states, districts, and schools produce those results--and address failure. 

We suggest there is a reasonable level of nationwide agreement as to what children should learn in reading and mathematics.  Federal lawmakers should take advantage of that consensus.  Using the NAEP as a benchmark, Washington ought to establish clear and uniform expectations regarding student master in reading and math at the fourth-, eighth-, and perhaps twelfth-grade level. 

Critically, this would allow the Department of Education to use existing NAEP tests to gauge student performance in math and reading without requiring states and the Department to engage in the extensive negotiations now taking place over assessment systems.  Meanwhile, states should have a free hand to craft standards and tests for other subjects.

Gains

Second, the performance of schools and districts should be judged primarily on how much students are learning while in school--not on the absolute level of student achievement.  Measuring aggregate levels of achievement encompasses three things--what a student has learned this year, what they've learned in all previous years of schooling, and everything else that has transpired in the child's life--of which only the first is relevant to gauging whether schools and educators are performing adequately.  While we obviously care about the level of student performance, sensible accountability recognizes that employees can be competent without being Herculean.

Today, NCLB is not designed to capture the value that schools are adding--how much students learn in the course of a given school year.  However, despite assorted technical concerns, the capacity is developing.  Today's NCLB is hostile to value-added analysis.  That should change.

Triage

Third, NCLB should replace its all-or-nothing adequate yearly progress calculation with a triage model that distinguishes among schools more effectively than does the current system.  Today, too many adequate schools are being flagged as "needs improvement" due to vagaries in the way that AYP is calculated--particularly its creation of dozens of subgroups, failure in any one of which means a school is failing to make AYP.  An alternative is advisable.  One might, for example, distinguish among schools that are making progress overall and in 90 percent or more of their subgroups; those making progress overall but in fewer than 90 percent of subgroups; and those failing to make acceptable overall progress.  Such a triage system would reduce the vast number of mostly okay schools that are now being flagged by NCLB.  It would distinguish between those on the verge of succeeding and those that are catastrophically inadequate and enable states and districts to focus on preparing the latter.  And it would bolster the confidence of parents and voters in the NCLB branding system.

Frederick M. Hess is a resident scholar at AEI.