(Guest post by Greg Forster)
My column on NCLB is on Pajamas Media this morning. A sample:
When you set aside all the implausible multi-year plans, toothless sanctions, easily evaded school choice requirements, and other window dressing, NCLB boils down to one simple commercial transaction: the system got a big cash payoff, in exchange for which it agreed to give standardized tests and release up-to-date information on how students are performing.
Before NCLB, many states didn’t give standardized tests at all, or didn’t release the results in a timely and publicly useable format. Now they all do. And all 50 states now participate in the Nation’s Report Card, a single national test of a representative sample of students, which allows researchers to conduct cross-state comparisons.
This transparency represents an incredible boon. The amount of empirical research done on education has been growing at a breathtaking rate. Before NCLB, education was a fringe element at best in economics, political science, and other social science disciplines. Now it’s everywhere. A lot of that research is due to the data made available by NCLB.
Jay P. Greene's Blog 
I agree with some of your comment – but your are wrong about “a single national test of a representative sample of students, which allows researchers to conduct cross-state comparisons”. The part missing is a single national test. This should be added and all schools, public and private, should be required to take it.
The national test is the Nation’s Report Card, also known as the National Assessment of Educational Progress (NAEP). It’s administered to a representative sample of students in each state. This allows cross-state comparisons of performance on NAEP scores. And researchers have been linking each state’s NCLB test to its NAEP scores in order to allow cross-state comparisons of the NCLB tests as well.
Before NCLB, many states chose not to participate in NAEP. Now it’s mandatory. So NCLB did establish a national test – not in the sense that every child takes the test, but in the sense that there is a test given in every state that allows cross-state comparisons.
As for requiring private schools to take tests, that’s not a good idea. Believe me, I’m sympathetic to your concerns. Transparency is good – I’m a scientist and data put food on my table – so I’m always glad to see kids taking tests. But freedom is more important than transparency. And test-taking does change the educational environment. A school with testing will have a different curriculum from a school without testing. And a sizeable minority of parents say that their children will learn better when they’re not subjected to testing. I trust them to know better than I do what will help their kids learn. Research shows that accountability testing increases learning for most kids. But what helps “most kids” doesn’t necessarily help every child without exception. Requiring private schools to test would eliminate a valuable choice that parents should have – the choice to educate their kids without testing.
Requiring all private schools to take the same national test would be even worse. Now you’re not just talking about taking away freedom to choose your educational environment in one respect (whether you have testing). Now you’re talking about de facto imposition of a national curriculum on all students. That’s not only going to reduce parents’ freedom to choose the right educational environment that will work best for their children, it also raises constitutional concerns. The Supreme Court was dead right in 1925 when it ruled that parents have a fundamental right to control their children’s education. The law it struck down in that case was a bigoted attempt to stamp out a minority language and culture by forbidding parents to pass their culture on to their children. Obviously what you’re proposing is not nearly the same thing, but if you give the government the power to require all students to take the same curriculum, you’re giving government power that it can abuse.
NAEP warns users not to make cross-state comparisons with its data. And I don’t know about other states, but Kentucky still doesn’t get its results out in a timely or usable fashion. A big problem we have seen in recent years is dumbing-down of our state test.
NAEP can be used to make cross-state comparisons, regardless of whatever politically convenient boilerplate is placed on it. Even the agency that conducts NAEP uses it for cross-state comparisons. For example, if you go to the NAEP section of the NCES website, you can see State Profiles and State Report Cards that show you each state’s performance compared to that of other states.
I agree that not all states are as timely or make the data as user-friendly as they could. But the difference between before and after NCLB is huge. NCLB has made a big difference in this area, even if more improvement would be desirable.
As for dumbing down of tests, that has always been a problem, and I see no reason to think the problem is any worse now than it was before NCLB. I’ve seen a lot of people who have systematically compared test rigor across states. What I haven’t seen is somebody who has systematically compared test rigor before and after NCLB.
If you count all the states that gave no test before NCLB as having test rigor zero, I’ll bet NCLB has been a lot more positive than negative for test rigor.