So, looks like we’re getting back into NCLB reauthorization mode. I laid out some of the broad context on Monday. While nobody is thrilled with NCLB, there are concerns that the Senate Republicans are going to go too far in “retreating” from the appropriate federal role. Today, I want to set aside for the moment philosophical arguments about the federal role, and talk about the design problems of NCLB, and why it’s essential that any vision for reauth steer clear of repeating those.

Checker Finn and I argued six years ago in Education Next that NCLB’s critical flaw was its pie-eyed, overwrought ambition. As we wrote, “NCLB is, in fact, a civil rights manifesto masquerading as an education accountability system. Its grand ambition provided a shaky basis for policymaking, rather as if Congress asserted in the name of energy reform that America will no longer need to import oil after 2014 or fought crime by declaring that by that date all U.S. cities would be peaceable kingdoms.”

All the same, we were not unsympathetic. We wrote at the time: “NCLB’s backers can legitimately argue that they had already spent nearly two decades asking state and local officials and education leaders to address mediocre school performance and stubborn race- and class-linked inequities... In that light, the passion-drenched unseriousness infusing NCLB is forgivable, even honorable. And NCLB indeed has virtues: it produced long-overdue school transparency, focused unprecedented attention on achievement, [and] created urgency where lethargy had ruled.”

That said, the design failings of NCLB were notable. Unaddressed, they infused the Harkin-Enzi bill that emerged in 2011, and will continue to haunt the Democratic proposal. Checker and I pointed out that NCLB sought to do three different things -- each sensible enough in its own right, but a Rube Goldbergesque hodge-podge when combined. As we wrote:

“Embedded within NCLB’s accountability system are three distinct, discernible models of educational change that have been awkwardly welded together.

Model one would make transparent the performance of students across the nation, providing an X-ray to show parents, educators, and policymakers how different schools and groups are performing in key subjects. Model two would deploy “behavior modification” accountability methods, refined through decades of public sector reform, to force low-performing schools and districts to set goals, assess effectiveness, and do better. And model three would set “shoot-the-moon” targets and use the federal bully pulpit to exhort leaders in states and districts to improve.

Each of these approaches is plausible on its own terms. And each has a place in federal policy. But they cannot reasonably be linked to one another, as NCLB tries to do. They entail discrepant views of the federal role in education and employ discordant mechanisms. The result isn’t working.”

We pointed out, “The value of an ‘X-ray’ of the nation’s school performance has long been recognized. NCLB’s dictate that all states regularly test students in key subjects marked a historic success. The accuracy of the picture is compromised, however, when this cross-sectional look at student achievement becomes the basis for gauging the performance of schools and educators, much less for triggering interventions or remedies. We don’t judge doctors based on whether their patients are sick today but by how much patient health improves under their care. Judging professional performance on the basis of a one-moment-in-time X-ray encourages questionable behavior, leads states to play games with standards, and threatens to discredit the X-ray itself.”

Similarly, when it came to the idea that the feds needed to force states to act, we wrote “Prodding public sector institutions to set goals, monitor performance, and then reward excellence and address mediocrity has been a signal success for reformers on both the left and the right... Sensibly structured accountability systems encourage self-interested workers to take goals seriously, focus on outcomes, and employ all the levers at their disposal to produce those outcomes. But we compromise such ‘behavior modification’ when those on the ground view the targets as unattainable. If workers know they are unlikely to succeed, the goal becomes to avoid trouble when they fail. By making failure inevitable, unrealistic goals have the perverse effect of focusing employees on compliance.”

What to do about all this? It requires teasing apart these three roles. The federal government can and should insist on transparency in return for federal funds. It’s fine for the Secretary of Education to be a cheerleader and appropriate for the SecEd to use moral suasion. But it’s a mistake to tie artificial goals to pleasing sentiments. And it’s a mistake for Uncle Sam to try and get into the business of fixing schools, no matter how much he distrusts state and system leaders. After all, Uncle Sam can’t fix schools. All he can do is pass laws, which makes ED write regulations, ordering states or districts to alter policies, in the hope that these change practices in schools and classrooms.

Now, for instance, the feds requiring states to set performance targets instead of setting a national 100% by 2014 target is an invitation to repeat some of these same problems, just with a new wrinkle. States will be pressed by the Secretary of Education to set pie-in-the-sky growth expectations for gap-closing... and then all the pathologies of NCLB repeat. Unless and until someone proffers workable suggestions here, such a proposal tells me that the lessons of NCLB haven’t really been learned.

A decision to focus NCLB reauthorization on promoting transparency, honest measurements of spending and achievement, and on ensuring that constitutional protections are respected ought not be seen as a retreat from NCLB but as an attempt to have the feds do what they can do sensibly and well.