Over at New Talk, a public policy discussion project that purports to bring in “experts…who have different points of view…and a commitment to the kind of open discussion that might take place around a dinner table,” anything but an open discussion is going on about the No Child Left Behind Act. Yes, the e-talkers range from neoconservative Chester E. “Checker” Finn to American Federation of Teachers President Randi Weingarten, but in one way or another all of the discussants believe in significant federal involvement in education. That’s left out one very important perspective: the Constitution’s.

Here’s moderator John Merrow’s reason for keeping away those who think that the federal government has neither the Constitutional authority, nor the ability, to run American education:

Any talk of abandoning No Child Left Behind is foolish because NCLB is the continuation of a long trail of federal education legislation that traces back to the Elementary and Secondary Education Act of 1965.

Congress and the next Administration must do something, but what? That’s the question posed to a remarkable roster of deep thinkers and activists.

Merrow’s question framing has, predictably, produced not the “open discussion” promised by New Talk, but a very narrow exchange dominated by “national standards” chat. Indeed, former Columbia Teachers College president Arthur Levine remarked in his first contribution to the forum that he “was surprised at the commonality of views the group shares.” He shouldn’t have been: For the most part, the group is a collection of education policy insiders with a big bias toward government “doing something,” and the moderator made certain that asking whether the feds actually can do something — at least of value — was off limits.

Of course, it isn’t rational to just assume that NCLB — or any federal interference in education—will remain forever just because it is the continuation of legislation dating back to 1965. Consider that as recently as 1996 the Republican Party had eliminating the U.S. Department of Education in its platform. Or that two major pieces of legislation that would essentially dismantle NCLB — the LEARN and A-PLUS acts — are working their ways through Congress. Or that several states have seen considerable efforts to leave NCLB behind, despite the sizable financial blow doing so could entail. Or that in polling Americans often express great distaste for NCLB.

So what about the question of whether the federal government is actually capable of producing good education? The evidence, whether we bury our heads in the sand or not, is that it can’t. Since 1965, real federal funding for elementary and secondary education has grown nearly six times larger, while achievement has been essentially flat or declining. Under NCLB, what scores were increasing have seen improvements slow while others have seen losses. And why is this? Politics. Washington, like all government, is more responsive to special interests that make their living off of government programs than the people those programs are supposed to help. So the feds keep lavishing cash on the public schools, while the states keep setting almost subterranean “proficiency” bars, using statistical gimmicks to exclude the groups the law is supposed to make visible, and the administration cheers NCLB’s “success.”

The one highlight of the New Talk discussion so far (it runs through tomorrow) has been education historian Diane Ravitch, whose past calls for national standards I’ve taken to task. She’s still calling for national standards, but she has nonetheless been the most realistic of the discussants. After asserting that “there is little to commend NCLB,” she made clear that any national standards would have to be disconnected from federal sanctions because “states and localities” — not Washington — “are closer to the schools and likelier to come up with workable reforms.” She also deflated the soaring rhetoric of those who talk about NCLB as an historic shift because it has finally shone light on the kids left behind (though she also leads one to conclude that she might actually want stronger national tests than she outlined earlier):

OK, so NCLB is historic. No doubt about it. Never before has the federal government reached so deeply into each and every public school in the nation. There was a fundamental error, however, in allowing states to define their own standards and write their own tests. As a result of this error, the public does not have the information that Checker speaks about; instead, in most states, the public gets a wildly inflated picture of student performance.

Unfortunately, as Ravitch being the highlight portends, another perspective is almost wholly absent from the discussion: that accountability exercised by parents through universal school choice, not continued top-down accountability from states or Washington, is the key to truly effective education reform. Consumer choice is the primary driver of accountability for almost everything in America that we take for granted — consumer electronics, package delivery, automobiles, greeting cards, sneakers, home construction, and on and on — but not so for schooling. Maybe that’s why over the decades education hasn’t progressed at all, while almost every other good or service has gotten much better. Maybe that’s also why school choice should be a part of any “open discussion” about how best to deliver education.

New Talk could have put together a truly valuable forum on federal education policy if it had included all perspectives. Sadly, it chose not to.