The thing to understand is that national solutions do not work. Demographics, culture, and economics vary drastically from region to region, and one-size-fits-all approaches simply can't accommodate these complex differences. As attractive as "universal" (as in universal health care) programs may sound, they rarely deliver on their promises and usually make things worse. Indeed, one large step toward improving education (and most things) in America would be to place it completely beyond the influence of Washington.
So the real consideration is whether the reforms being sought -- either a free market approach or an overhauling of the existing government school curriculum -- can practically be implemented at the state level. In the case of free market reform, it's reasonable to expect so. States are already moving down this path, experimenting with various forms of school choice, while the federal government hasn't done much in this area.
Curriculum reform, on the other hand, is more apt to play out at the national level. After all, states have been in control of their curricula for years, with little to show for it. There has been an increasing call to "try something else." Unfortunately, that something else has taken the form of increased federal control, and that trend seems likely to continue.
Here's what Andrew J. Coulson, Director of Cato's Center for Educational Freedom has to say on the matter:
More central planning is not better than less central planning. . . The instructionists who seek to homogenize American education are detached from reality, imagining that new national standards, testing, or curricula would invariably be better — and remain better — than their counterparts adopted at the state level today. There is no reason to believe this. The same dysfunctional incentive structure that has produced the faddism and quackery in state curricula. . . would apply at the national level.
Furthermore, national standards/curricula would raise the stakes for the school wars that have been dividing communities in this country since the inception of state schooling. What history should we teach? Should math be fuzzy or traditional? Should reading instruction be phonetic or wholistic? Add to all this the fact that the U.S. Constitution grants the federal government no mandate [I would say no authority] to meddle in matters instructional, and there is every reason to oppose calls for national standards/curriculum/testing.
Coulson goes on to point out that without free market pressures, any progress made within the current system (even at the state level) are likely to be meager and temporary.
Trying to “fix” the education being provided by a monopoly school system is like trying to “fix” a command economy. While occasional improvements will certainly be possible, ultimately, the effort is doomed. Even when excellent, proven methods or curricula are adopted in state schools, the incentive structure of the system provides no support for retaining them.
Hence the doubly ill-advised No Child Left Behind act, which has always struck me as a positive step in the wrong direction. Introducing accountability into a system where none exists can't help but yield improvements, as we see with NCLB. But such a system -- monopolistic and administered at the national level -- is too easy to game. Improvements become increasingly difficult to retain and build upon as loopholes are found and relaxing standards go unenforced.