Posted November 20, 2008, 5:23pm
I want to thank the folks at NewTalk, once again, for hosting this discussion. No matter where people stand on NCLB, I think it is important to consider all possible fates for the law in order to best gauge what should be done. I also want to thank Jay for moderating. I thought he was very even-handed and tried his best to keep us cats herded.
Where I stand on what should be done with NCLB is probably clear—it should be scrapped—but I just want to reiterate one more time why. Top-down, political control of education does not work because it puts most of the power in the hands of the people employed by the system, not the parents and children the system is supposed to serve. This is an especially dangerous problem if all authority is concentrated at the federal level because if that power is abused, there’s nowhere to run. But this absolutely does not mean that I think local or state control is much better; the same power structure remains, only at different governmental levels. That is why the only way to truly cure what ails us is not to make tweaks to the fundamentally broken system, but to transform American education by giving control of funding to parents and letting all schools—public and private, charter and traditional public, existing now and created when there’s a market—compete for their business.
Posted November 20, 2008, 5:00pm
As this dialogue draws to a close, I want to thank everyone for their thought-provoking contributions. I think we've clarified the lines of debate about NCLB, which I hope will be useful as reauthorization approaches. Thanks again.
Posted November 20, 2008, 4:01pm
I agree with Elaine on emphasizing incentives over punishments, which is one of Lauren Resnick's key recommendations.
One parting thought: Albert Shanker, a strong proponent of standards, testing and accountability, predicted at a 1993 forum in Boston that the reform would take decades to perfect. I'm afraid his prediction is proving all too true.
Enjoyed the conversation with everyone. I learned a lot.
Posted November 20, 2008, 2:45pm
While a lot will depend on the details, Elaine, I think there is a great deal of merit in what you propose.
Posted November 20, 2008, 1:36pm
One reason, yet to be mentioned, not to scrap NCLB, is that educators will constantly say behind the backs of policy makers and superintendents, "just wait it out and this too shall pass." They also state that policy makers continue to change direction, create new forms, bureaucracy and demands. As one of these policy makers, I believe they have a valid point. Therefore, I believe we would be better off to fix NCLB, simplify it, provide more flexibility to states and yet be very clear in defining accountability and providing consequences for schools that fail.
There should be more of a focus on incentives rather than punitive measures, which in most situations have not created positive outcomes.
There should be MUCH less reporting and an adoption of the student growth model as the accountability measure.(See The Quick and the Ed , Tuesday, October 21, 2008, "A Thousand Words".)
Lastly, as described in the September 28, 2007 policy briefing memo, entitled "Keeping Achievement Relevant: The Reauthorization of 'No Child Left Behind" prepared by Democrats for Education Reform—one of the most fundamental flaws in U.S. education policy right now is that the students who most need good instruction are the least likely to get good teachers. This issue MUST be addressed in the reauthorized NCLB.
Posted November 20, 2008, 12:36pm
The concern about NCLB is precisely that it encourages evasion while simultaneously declaring, “Don’t worry! We’re here from Washington and we’re here to help!” It makes promises to parents and taxpayers on which it likely can’t and won’t deliver, while dangerously increasing centralization. And, of course, it does that at an ever-increasing price tag (and expect a lot more money soon!) while distracting us from the key to reform, putting power in the hands of the parents whose children the schools are supposed to serve.
Posted November 20, 2008, 12:28pm
And be careful what you call “the four strongest research studies” on vouchers. I’ll just point out the problem with one of your cited pieces, the Wolf Milwaukee study. It may very well end up being one of the strongest research studies, but Wolf and co-author John Witte themselves have written that from what has been released to date, “no reliable conclusions about the effectiveness of the choice program can or should be drawn….We provided that important guidance throughout our reports. Nevertheless, many commentators chose to ignore it.”
Finally, let me address the question of NCLB’s effect on standards. Without question during NCLB’s tenure there have been some movements to push standards up (as with Achieve) and some down (as with Michigan, Texas, and Colorado). Not surprisingly, whether state standards have risen or fallen has been a decidedly mixed bag. But NAEP data reveals no great improvement under NCLB (indeed, improvements appear to have been coming a bit faster before it) and we have good evidence that almost no state has set its standards high, if by “high” we mean state “proficiency” equivalent to NAEP “proficiency.” Maybe that’s NCLB encouraging states to push standards down, or maybe just not push them up—or maybe it’s not NCLB at work one way or the other—but regardless, we see nothing like the “huge, positive contribution" you, Sandy, have attributed to the law. This is especially troubling since we’ve seen about a 40-percent increase in federal ESEA spending under NCLB, and because there is good reason to believe that many states would have been working toward higher standards without NCLB. Maybe they wouldn’t have kept them, but is anyone willing to bet that NCLB will remain strictly enforced?
I hope this dispels, for the last time, some of the confusion about what I and other have said that I continue to see repeated.
Posted November 20, 2008, 12:26pm
Sandy, let me lay this out one more time.
I will not re-print the actual NCLB statute as that would take about 600 pages worth of pixels. Perhaps the U.S. Department of Education’s 2005 “Road Map for State Implementation” will suffice to show that the feds have told states what many of their educational structures will look like. I’ll just give a couple of examples from the Road Map:
Assessments in Grades 3-8: What gets measured is what gets done. States must test all students annually in reading and math in grades 3 through 8 and once during high school by the 2005-06 school year—not every other year or every other class but all students every year.
Or how about this:
Proficiency by 2013-14: All education reform efforts, from federal and state policies and programs to individual classroom strategies, must strive to be informed by the best of what we know from research and must focus on the bottom line of raising student achievement and closing achievement gaps. States must include all students in school accountability systems and set targets for all students to reach state standards for proficiency in reading and math by 2013-14.
Now, agree or disagree with the worthiness of these goals, the federal government is telling states what their assessment structures will look like. And, of course, from highly-qualified teacher requirements to supplemental services demands, NCLB does a lot more than just this.
About me touting local control: I have never touted local control, I have only said it is preferable to NCLB and federal control. In fact, I wrote this morning, “people keep wanting to say that some of us advocate for local control as a solution to our woes, when we don't. We argue simply that local control would be better than NCLB.” Indeed, starting with my very first post, I have said that widespread school choice, which fundamentally changes the incentive structure in education, is the only solution.
About the voucher research, I wrote the following this morning:
As for vouchers, as our esteemed moderator has documented, almost all "gold-standard," randomized field trials of vouchers have found a benefit for voucher users. Moreover, as I have linked to in previous posts, lots of research from here and abroad shows that the more free-market-like an education system, the better the performance.
Oh, and look what else I wrote:
But I have also stated that we shouldn't expect major positive effects from the choice programs we have now because they are very small and hamstrung, and do little to move the vast majority of power away from politicians, bureaucrats, and special interests and toward consumers.
Posted November 20, 2008, 12:18pm
Key features of NCLB that, in my view, have made the greatest impact and must be preserved (while being improved) are:
1) School success (i.e. ratings) must depend significantly on the success of disadvantaged students;
2) Measures of performance that are used in accountability must be valid, reliable, objective, and statewide in nature;
3) There must be meaningful consequences for performance, including better technical assistance for low performing schools and real choice for students stuck in persistently low performing schools;
4) Annual tests along with the data and transparency requirements must be retained (and improved) in order to inform parents, educators, and the public, and serve other purposes, such as making growth models workable;
5) The emphasis on good teachers, particularly in low income schools, must be retained, though, as previously discussed, must be revised to promote teacher effectiveness;
6) The expanded NAEP participation must be retained;
7) Greater honesty and transparency regarding graduation rates and their use in accountability must be retained; and
8) While continuing assessment problems must be addressed, the increased expectations and learning opportunities for disabled students and English language learners must not be lost.
I've described elsewhere advances and extensions that I think are important. And I'm sure I'm forgetting items in the rush to get this out. But this is a start to your question, Jay.
Posted November 20, 2008, 11:14am
Let me try to challenge each side in this debate. It seems to me that NCLB actually did very little—good or bad. 37 states already had some kind of accountability testing before NCLB was passed. And NCLB leaves it to the states to define proficiency, develop measures of that proficiency, etc… Lastly, the sanctions in NCLB have now been shown to be almost completely toothless. Students may be entitled to tutoring if their school fails to make AYP, but the schools are the dominant provider of that tutoring and have captured virtually all of those dollars. Students may be entitled to transfer to another public school, but the schools have effectively undermined the choice option by failing to inform parents properly and failing to allow transfers by claiming that schools are full.
In the end no school district, state, or educator has experienced any negative consequence from NCLB sanctions. No revenue has been lost. No teacher has lost his or her job. Nobody has experienced a cut in pay. Sure, schools are threatened with "restructuring" but that may be little more than Dean Wormer's "double secret probation" from animal house.
Now let me be clear that NCLB has produced some significant changes. It has entrenched the idea that we ought to measure achievement in some way and publicly report results. That is a huge shift in the culture around schools that is not going to go away, even if NCLB is greatly reduced in scope.
But if I am right that NCLB hasn't done very much, then why would we expect the perverse behaviors that the "scrap NCLB" folks are worried about? Why should anyone narrow the curriculum, teach to the test, focus on bubble kids, etc…?
Conversely, if NCLB hasn't done very much (other than entrench a culture of testing), what do the "keep NCLB" folks see in NCLB that is so important and worth saving? We probably wouldn't lose the measuring and reporting culture in the absence of NCLB because reporters, public intellectuals, and politicians are now all addicted to the information. If NCLB hardly does anything, why endure all the hassle of sustaining it?
Posted November 20, 2008, 10:53am
I would like to respond to Martin's suggestion that federal law ought principally to challenge the states to address the problems of a certain number of the lowest-performing schools—say, the bottom 10 percent.
This was actually an idea that the late Sandy Feldman and I pursued in negotiations on NCLB. It makes a lot of sense.
Posted November 20, 2008, 10:48am
Neal—would you take the time to cite provisions in federal law or policy to prove your assertion that the feds tell states what their standards, testing, teacher quality, and other educational structures must look like? You allege the horrors of federal infringement, yet the fact is that education in our country remains largely dominated by state and local policy. You tout local control, yet you refuse to acknowledge its problems and how they should be solved. You pooh-pooh research presented to show accountability works, then you dismiss the four strongest research studies that demonstrate how little effect your pet program has shown in lifting student achievement. I'm not sure we're doing the others a favor in continuing to argue how little we think of each other's views.
Let me return to Jay's challenge to stick to the details of issues around NCLB. I follow standards work in states extremely closely, and, other than some gaming around the edges, I find virtually no erosion in standards around the country in the wake of NCLB. Many states have actually raised the level of their standards. Take a look at Achieve's recent reports. And as I've pointed out earlier, any state that wants to raise standards can do so as "advanced" or under any other such banner and meet federal requirements under current lower levels of "proficient."
Testing policies are totally in the hands of states. States, on their own, emphasize math and reading. And most of them are using science and social studies tests increasingly. Many have gone to ACT both for exit level tests and for secondary level tests. Others are looking at standards and tests tied to being on path to post-secondary readiness. Others have developed regional collaboratives to revise standards and tests. Others have gone to end of course exams. State experimentation is more vital now than before NCLB.
Teacher quality? The teacher quality provisions of NCLB are criticized mostly for their weakness, not because of their "control" of state policy. After a flurry of somewhat constructive action and tension in 2002, the states used the HOUSSE loophole and now do exactly what they want to do.
The problem with NCLB is not that it intrudes too much. It does place pressure on the states, and that's been to the good. (The letter from the civil rights groups on the proposal to weaken NCLB says it better than I.) But its principal flaw is that its effect is not as strong or effective as it should be.
Posted November 20, 2008, 10:03am
Thanks, Richard K. for the appreciation and "loving the irony that one of our country's leading free market school voucher advocates is put in the position of having to regulate and direct the discussion!" I have to say that I enjoy being the moderator because it allows something like the word moderate to be close to my name. : )
Speaking of moderation, I agree with Richard R. that we should stick closer to the issues in dispute. Things often get heated in education debates because people care intensely about these important matters, but I'm sure that everyone will cool down here this morning. In fact, I already see that we are making some good, constructive progress this morning.
Posted November 20, 2008, 10:00am
I hate to disagree with Jay, but there is necessarily tension between choice and accountability if it comes down to punishing choosers for not picking the “right” thing. Indeed, countries with national curricula often allow choice—and as Woessmann has shown it seems to help—but the chosen schools must generally all teach the national curriculum. Once that is required, it seriously squelches competition and especially innovation, not to mention freedom. And, with NCLB, we are but a few steps away from a national curriculum. In fact, many people are calling for “national standards” because of NCLB’s inherently perverse incentives: “States, get everyone to proficiency or else, but you define ‘proficiency’ for yourselves.”
I also think choice has to remain in the discussion of what to do with NCLB. If we are to find a workable solution to our problems, and neither federal, state, nor local control hold it due to fundamental flaws in top-down power, we have to consider completely reversing the power structure of the system, and that means we have to continue to consider school choice (though not necessarily vouchers).
What about the accountability evidence? From my reading of it, accountability regimes seem to work to an extent to get kids to improve on easily quantified learning. The research in this country, however, is very thin concerning the ability to sustain accountability regimes, especially when it comes to reaching truly high standards. Moreover, even if it were to prove sustainable, the fact that such a narrow band of learning is fixated upon at the very least holds considerable risk that other material, and admittedly mushy “critical thinking,” could get short-shrift.
I should also make an important but overlooked point. If public sentiment behind reaching for tougher standards has gotten high enough that an accountability system can be sustained politically—and any success of accountability regimes very possibly reflects overwhelming sentiment for reaching certain standards, not the power of accountability systems per se—than an even better option is, again, school choice. You will get most people still seeking higher standards, but also competition, innovation, and the ability to tailor education to individual needs and desires. And you’ll get all of that without the same level of risk that a political change will take the rigor will go away. When lots of people get tangible benefits, they are much harder to take away.
Posted November 20, 2008, 9:10am
Jay, you're doing a great job of summarizing, and I'm loving the irony that one of our country's leading free market school voucher advocates is put in the position of having to regulate and direct the discussion!
The central question you've identified is whether we should "scrap" NCLB. My answer is no, because a) standards based reform makes sense; and b) the federal government has a unique role in ensuring equity. I agree with those who have argued for significantly improving the accountability system. I think it's to Sandy's great credit that a key architect of the act is willing to acknowledge that we need to fix it and improve it in significant ways.
Let me disagree, Jay, with your characterization of the discussion threads about pre-k and school integration/interdistrict choice as peripheral. I think they are central to fixing NCLB because to my mind, the major flaw with NCLB is that it tries to narrow the achievement gap while basically ignoring the critical findings of the legendary Coleman Report (1966), confirmed in subsequent research time and time again.
Coleman found that the socioeconomic status (SES) of the family is the most important predictor of achievement and the SES of the school a child attends is the second most important predictor. That's why pre-k, health insurance, housing and all the things Richard R. talks about are so important. And it's why NCLB's transfer provisions—which hold out the hope for reducing concentrations of school poverty—are also critical.
I'm also for fixing the standards, testing, and accountability provisions of NCLB, and Lauren Resnick has an excellent set of recommendations in her chapter in our book, Improving on NCLB. But at an even more fundamental level, fixing NCLB requires addressing Coleman's key insights. We have to address childhood poverty and we have to stop being content with doing our best at making Plessy work. The current strategy will never close the achievement gap.
Posted November 20, 2008, 8:10am
Thanks, Jay, for re-focusing our discussion. I certainly support a federal role in accountability (and apologize for leaving this in doubt) but think that the NCLB approach requires more than minor tweaks.
As Mike Petrilli has argued, the NCLB experience confirms that the feds are simply not in a position to drive broad district-level reforms. The public school choice provisions have been a failure, as has the mandate that persistently under-performing schools be restructured. Petrilli (and others such as Diane Ravitch and Checker Finn) argue that the feds should limit themselves to providing transparent information on student achievement and leave states and districts to intervene where needed.
I’m not willing to go that far. But I would limit federally mandated sanctions to a much smaller number of schools (no more than, say, 10%) where both levels and growth in student achievement are abysmal. The current restructuring timeline would be expedited and the “any other major restructuring” loophole eliminated so that the sanctions for these schools are quick and severe. States would essentially be left to deal with other schools as they see fit, though the feds could encourage various approaches through competitive grant programs modeled off of the Teacher Incentive Fund.
Of course, all this would be futile unless the law’s school ratings are revised so that they do a better job of identifying effective and ineffective schools. At a minimum, they need to be based in large part on student growth, pegged in some fashion to comparable benchmarks for achievement across states, make multiple distinctions among schools, and not premised on the utopian notion of near-universal proficiency by 2014. As Rick Hanushek points out, these are all “straightforward adjustments.” But each of them will draw opposition, and it will take strong leadership in Congress to get them done.
Posted November 19, 2008, 8:37pm
Jay raises a good question in his 3:33pm comment. I can’t answer it without reference to the problem I raised earlier, that a narrow, exclusively test-based accountability system will inevitably lead to goal distortion. There are many things we want schools to accomplish beyond basic skills in math and reading, but holding them accountable only for math and reading MUST lead to some abandonment of other goals. Now some of these other goals can also be assessed with standardized tests (historical and scientific knowledge, for example), so you might think that NCLB could be reformed to include testing in these other areas. But many of the other goals cannot presently (perhaps not ever) be assessed in this way, and require observation of student performance that is more difficult (though not impossible) to standardize. In practical terms, holding schools accountable for many of these other goals requires observation, and judgment of expert observers. In some cases, where we do not yet know how to measure outcomes, we might have to rely, for the time being on inputs (if we don’t know how to measure good citizenship, can we, for the time being, ask if a high school offers community service opportunities to all students?). I have suggested that there is an infrastructure already in place that could do this (the regional accreditation agencies) although they would have to be significantly reformed to play this role. But it is hard to imagine a set of federal rules that could guide this process, because considerable state experimentation is still required to develop such accountability procedures, and I don’t see the expertise available at the federal level to do this. I could see an NCLB revision that simply required states to develop such systems, but can’t imagine that it can get more specific than that. If it were possible to develop a federal procedure along these lines, it can’t possibly be accomplished in time for this ESEA re-authorization. In 2001, there was more interest in getting it done than in getting it done right. I think that those who favor reauthorization of NCLB sometimes seem to be reproducing this reckless sense of urgency. (For example, there are still too many unanswered questions about the reliability of value-added measures to make these the basis of a national accountability system.) Someone earlier (in this exchange) observed that education tends to get much more politicized at the federal level, and is addressed more pragmatically at the state level. I agree with this observation, and have more faith in the ability of state governors and legislators to move towards sensible accountability procedures. Some will get it wrong. Some will get it right. NCLB ensures that we all must get it wrong.
I also agree with Neal’s response to an earlier critique: that the alternatives are not federal or local control. I can say for myself, I am not arguing for local control, but for state responsibility. Many of our states are as large as other nations. There is no reason to assume they cannot fulfill this responsibility.
Sandy, when I agreed to participate in this dialogue, it was because I assumed there was an interest in a civil airing of views. I have no intention of responding to accusations of “bombast,” “extravagant ideological views,” and the like.
As to another personal (though not so offensive) reference, there is no “new 10th Amendment model Richard Rothstein.” I have been writing for years that accountability policy is too complex to be micromanaged from D.C. (See, for example, my exchange with Rick Hanushek, SHOULD THE FEDERAL GOVERNMENT BE INVOLVED IN SCHOOL ACCOUNTABILITY? in the Journal of Policy Analysis and Management, Vol. 24, No. 1, 167–178 , an exchange to which we were both invited because our views on this topic had been previously well established.) As the Department of Education and Congress have flailed around since re-authorization has become overdue, unable and/or unwilling to modify NCLB to respond to its obvious flaws, I think that the reasonableness of my views on this topic has been confirmed.
Posted November 19, 2008, 8:28pm
Sandy, this is turning out like the argument about local control. People keep wanting to say that some of us advocate for local control as a solution to our woes, when we don't. We argue simply that local control would be better than NCLB. Similarly, I haven't "cloaked" my "opposition to NCLB in the sudden rationale that research does not prove it effective." My rationale from the start has been that the incentive structure in NCLB, and all top-down education control, is wrong. I brought up research when you declared without support that NCLB has "made a huge, positive contribution" to improving education. I couldn't let such a statement go by, especially in light of all the evidence I and others have offered against it, without asking for support.
As for vouchers, as our esteemed moderator has documented, almost all "gold-standard," randomized field trials of vouchers have found a benefit for voucher users. Moreover, as I have linked to in previous posts, lots of research from here and abroad shows that the more free-market-like an education system, the better the performance. But I have also stated that we shouldn't expect major positive effects from the choice programs we have now because they are very small and hamstrung, and do little to move the vast majority of power away from politicians, bureaucrats, and special interests and toward consumers. Moreover, I have said that we do not have evidence in K-12 education of how a fully consumer-driven system would work—even though most competitor countries have more choice than we do—because such a system does not exist anywhere. But we can look all around us—at our computers, cars, package-delivery services, and on and on—to see how well free markets meet our needs and respond to customers. In contrast, to see how well top-down, government control works, we can look at the Soviet Union.
Finally, I have to reiterate a point I've now made several times. The federal government has clearly—BY DESIGN—taken control of state accountability, telling states what their standards, testing, teacher quality, and other educational structures will look like. True, Washington has left it to states to define important things within that structure, but all the incentives, enforced by the federal government and threats of punishment, are for states to set those things that they can control, most notably the definition of "proficiency," at the lowest level possible. So while I have never said that the states have lost their ability to set their own standards or hold themselves accountable under NCLB, I have said that Washington has created strong incentives for states to set their standards low, and that, I think, is very well established.
Posted November 19, 2008, 6:10pm
I agree with Rick H.'s earlier suggestion that there is no necessary tension between choice and accountability. Sandy makes a fine point about how research on vouchers is not a slam dunk just as it isn't on accountability (although I have a more positive reading of the voucher literature here). I think his point is that we have to rely on a mix of good theory and solid evidence, since all evidence is a bit shaky. But let's avoid a tangent on a voucher debate. Instead, let's address the point that Neal, Sandy and Rick H. are really debating: what kind of evidence do we have about the effectiveness of accountability regimes?
Posted November 19, 2008, 6:00pm
Well—let's go ahead and get into the voucher research in greater depth.
The Wolf study of Milwaukee—students with vouchers aren't doing much better or worse than students in MPS.
The Department study of DC vouchers—no significant difference between those who got the vouchers and those who didn't but were qualified.
The Indiana Center for Evaluation on the Cleveland voucher experiment—no statistically significant difference in achievement.
The competing work of Krueger and Peterson on NY vouchers—as Mathematica's David Myers concluded: in the overall group of students studied, obtaining vouchers provided no academic benefit.
So, please, Neal, if you want to promote vouchers and diss NCLB, have at it. But don't cloak your opposition to NCLB in the sudden rationale that research does not prove it effective.
As to the role of the feds in state accountability, I would remind you of all the ways in which states dictate their own systems. They set the standards. They make the tests. They set performance standards. States and locals almost entirely determine the consequences for performance. All curricular, materials, personnel decisions are made at the state or local level. Etc.
The feds ask in return for the billions they send essentially that schools that do not attain broad proficiency for their subgroups do something about it, something largely of the districts' choosing. I realize there's more to it, there's clear tension, and there's difficulty because of the inadequate ways in which these relationships have worked. But your characterization of the federal role as virtually displacing state accountability is inaccurate—to say the least.
Posted November 19, 2008, 3:33pm
We had a flurry of activity this morning and things seem to have slowed down a bit. I think everyone wanted to get off of their chest what they preferred on a variety of issues, some of which are only tangentially related to NCLB.
But let's get back on track. There are many things that people could want in education—more choice, universal pre-k, increased funding, etc… But the issue on the table is whether NCLB, a federal accountability regime, is a necessary and useful mechanism for promoting quality education. You may think other things would also be important or more important, but there is no necessary logical connection between those other policies and NCLB. One could be for (or against) expanding choice and still be for (or against) NCLB. One could be for (or against) universal pre-K and still be for (or against) NCLB. The same goes for funding reforms or many of the other issues that we've touched on.
While I think we have often strayed this morning, let me highlight points that people have made that are directly on topic. Neal has clearly come out against NCLB because he is against centralized accountability systems. Sandy, Rick (that's Eric Hanushek for those not familiar with everyone's nickname), and Elaine have clearly come out in support of NCLB because they both think accountability programs are helpful and think it has to be done at the national level. They still see room for improvement, but they primarily are defending the NCLB approach.
Andy has offered his insightful political analysis that NCLB isn't going anywhere in the near-term, but I didn't see him directly engage the debate about whether it should go somewhere. I don't want to engage in useless philosophizing any more than he does, but I think being clear about what is good and bad about the basic NCLB approach can help us figure out what to do within the political constraints he rightly recognizes. Joe, Richard K., and Andy have also strayed into discussions about integration, pre-K, and inter-district choice—all of which are interesting, but don’t necessarily address the central question about whether we should scrap NCLB and why or why not.
Richard R. has also clearly taken the anti-NCLB stance but I still don't quite see why he doesn't want NCLB to be revised to address his concerns about poorly conceptualized and designed measures of achievement and funding. Nor am I clear on why he would have more confidence in the states to develop accountability systems that would address his concerns—they certainly didn't do it before NCLB.
I'm not entirely clear where Marty is on these issues because we haven't heard too much from him.
I list all of this to push everyone to focus on the issue at hand (should we scrap NCLB) and to repeat back what I think I and (probably) other readers are seeing.
Posted November 19, 2008, 3:11pm
Let me begin by saying that I don’t think anyone has advocated local control. To some extent or another many have determined that federal control, at least as constituted under NCLB, is worse than local control, but they have not said that local control is sufficient.
Now, for the research. Thank you, Sandy, for citing examples of accountability at work. Unfortunately, I find it far from compelling.
I will begin by noting that NCLB did not simply extend “state accountability to the nation,” as Eric argues. Federal accountability is inherently different from state, because it dictates to all states the general framework under which they will operate. That is the very intention of federal accountability. Unfortunately, as I have documented, the effect of doing this has been to encourage states to either lower standards or keep them low. Moreover, extending accountability regimes used by some states to cover every state destroys the built-in protections of the federal system that lets some states experiment without all taking the risks. Add to this the much larger role of education in the states—and state politics—and it is clear that state officials must be more responsive on education than federal.
As for the research base, it seems rather flimsy. The CEP study on state-test-measured achievement, with its huge holes in data, proves more that NCLB has encouraged constant changes in metrics than sustained improvement, a big problem for a law that is supposed to bring “transparency” above all else. Carnoy and Loeb look at only math “basic” achievement over a short period of time—and my primary worry about government control is diluting rigor over time—plus the numbers of states analyzed are generally in the high 20s to low 30s, leaving the results subject to distortion by outliers and hardly universally applicable. Finally, Woessmann, at least based on his chapter in Cato’s book What America Can Learn from School Choice in Other Countries (sorry, I couldn’t get his other work quickly enough, and I don’t have the RAND studies at my fingertips), notes that “external exams” work best with school choice.
So why not national exams? For one thing, we’ve seen that in our uniquely diverse, free nation, few people can agree on what should be tested and how. Moreover, perhaps as a result of our diversity, we believe in much more than just testable material as the be-all-and-end-all of education. And that commitment, arguably, has worked well for us, fostering creative thinking necessary for success in our very free society and economy. Of course we could stand to do better in quantifiable, hard skills, but we don’t need the federal government to give us the “external exams” to help push that. Just as automakers are evaluated externally by Auto Week, appliance manufacturers by Underwriters Laboratory, and restaurants by Zagat’s, schools, in a system of widespread choice, would likely be externally evaluated by lots of sources, including external exams either now extant or to come. The standards and accountability would be there without the political perversion and bureaucratic waste of government controls.
Finally, one last note on school choice research. It is correct that evaluations of vouchers in the U.S. have found only marginal gains (though the best research pretty consistently does show gains). We should expect as much when the programs are as small and hamstrung as they are (unlike the nationwide NCLB). That said, most of the research from here and abroad on market-like delivery of education shows that the more freedom in education, the better. And we can look outside of education to see clearly that widespread choice and producer autonomy work much better than centralized control. Whether it’s your computer or Fed Ex, you’re seeing the market at work.
Posted November 19, 2008, 1:31pm
I have to say, I am somewhat startled at how some of you can advocate for complete local control with no oversight whatsoever by the federal government. When I began my tenure on the Denver school board about 10 years ago, I was all about local control. I have done a total turn around on that topic. I now believe that we need voluntary national standards and assessments. I agree with the notion that once these are established we should allow maximum flexibility to states in how they implement the standards. Think about how much money would be saved if there was an alternative to having 50 states spending millions of dollars developing and grading tests. Besides, what other country in the developed world has as much local control as the US with such poor results?
Our challenge is to figure out how to hold schools, districts and states accountable when schools fail. Accountability and consequences are key.
Posted November 19, 2008, 1:27pm
Sandy is right on this. It is not really possible to estimate the impacts of NCLB per se. The analyses of NAEP by Carnoy and Loeb and by Macke Raymond and me show that states with strong accountability got better NAEP results. These studies precede NCLB, but NCLB really just extended state accountability to the nation. The CEP analysis of trends in state test performance also suggests that the trend in performance tended to improve after the introduction of NCLB. This analysis is for a limited number of states (because tests have changed in a majority of states), and there is no way to say that it was NCLB and not something else. Nonetheless, the evidence points to a positive contribution of accountability in general.
Posted November 19, 2008, 1:16pm
Well, Neal, indeed—there have been no broad scientific studies on NCLB. But there have been few if any randomized trials on any of the ideas that have been mentioned in our discussions, much less the broad scale reforms such as NCLB. By the way, I happen to agree with some of your ideas about choice despite the fact that the IES randomized control trials say that vouchers make little difference. I'm very supportive of solid research, but sometimes we make the best decisions we can.
There have been fairly solid research studies on the benefits of accountability, the sort supported by NCLB. And I think there's strong data in 4th and 8th grade NAEP and some in 4th grade reading that standards-based reform (of which NCLB is a part) has made a difference in those grades to which accountability principles have been applied.
There are several studies, including some by participants in this discussion, that the sort of accountability called for by NCLB has made a difference in several states. The Carnoy and Loeb study in 2003 comes to mind. I would also reference the RAND study in the late 90s, discussing such reforms in Texas and North Carolina.
I don't want to swear by the research , but I don't think that the work of EdTrust, CEP, and the Council of Great City Schools should be ignored.
While the idea of federal initiatives is being pooh-poohed, let's also keep in mind recent OECD/PISA work that shows that countries with national standards, strong accountability, and local school autonomy are among the top performers. I also would commend the Woessmann cross-country study based on TIMSS.
Posted November 19, 2008, 12:55pm
I think the federal government has a legitimate role to play, under the 14th amendment, to ensure that states and local districts don’t discriminate in provision of services. I would say that it is far from conclusive, however, that IDEA has been a net gain. Its perverse incentives are huge for parents, schools, and all taxpayers, and, while what limited NAEP data we have for LD kids shows some improvement, it could very well be the result of massive increases in kids with at most mild problems being identified as LD, a major unintended consequence of the law.
Posted November 19, 2008, 12:34pm
Philip Howard’s Reader Comment poses a popular idea (“Why not transform NCLB into a standardized testing regime, but leave accountability local?”). If you look around the country and think that states are overall doing a pretty good job looking after the interests of low-income and minority youngsters it’s an idea you should support. But, if you look around and decide that the appalling dropout rates for minority kids (almost 50 percent) and the substantial gaps in achievement we see on state and national assessments between white students and minority students are in part the result of various state and local policies then that approach should give you pause.
A really good parallel here is special education and the Individuals With Disabilities Education Act. That law, while still full of problems and perverse incentives, has also really improved outcomes for special needs youngsters over the past generation. But does anyone argue that the best thing we should do with special ed kids is leave the states basically alone outside of information?
Getting that marriage of state flexibility and national oversight for populations that have traditionally not been well served right, striking the right balance, is an enormous challenge. But we can easily err on the side of too much federal involvement or too little.
Posted November 19, 2008, 12:06pm
I hate to keep pressing on this, Sandy, but there is nothing coming even close to conclusive evidence that NCLB has “made a huge, positive contribution.” I have pointed out the at-best-inconclusive testing data—both domestic and international—documented lots of evasion, and explained why the incentives behind NCLB would produce no great, positive contribution. Richard Rothstein has done the same. So what is the compelling reason anyone should believe that the federal government can push powerful, widespread, sustained reform? There might be some, but it hasn’t come from NCLB.
Posted November 19, 2008, 11:43am
Jay—obviously I'm a supporter of a federal role and believe NCLB, even with its flaws, made a huge, positive contribution. But I share concerns in all four areas you've identified.
As I've argued in previous posts, we need a much stronger federal/state/local partnership than we currently have. A better implementation of reforms in IASA and NCLB on all sides would have helped considerably in addressing many problems we've discussed.
The feds and the states have a role in accountability, which has obvious centralizing elements. But accountability must largely be local. Again—better administration at all levels is key.
I also believe in more targeted and well spent resources, particularly at the secondary level.
So, I'm a bit in all four camps!
Posted November 19, 2008, 11:35am
I think it would be interesting to let public school superintendents decide the answer to the private schools question.
Allow them to choose between guaranteeing real public school choice (both inter and intra) and giving the kids in crappy schools a money-back-guarantee if they don't.
Any district which opts to deny a transfer based on a lack of capacity triggers some sort of direct relief for parents of students who would prefer that they be educated someplace where they have some sliver of a chance of learning.
I also would entertain the idea of requiring any school board member from a district which denies transfer requests due to capacity to walk around on the street carrying large signs saying "Sorry, but we don't have good schools for children in our community."
The back of the sign could say either "But we're working on it" or "We have no intention of doing anything about it." We can let them decide.
And, I agree with Richard K. that financial incentives (attached to the students) is a desirable way to look at this issue.
Posted November 19, 2008, 11:27am
You read me right, Jay! Centralized, top-down accountability fails. Consumers must be empowered to leave schools they don’t like and seek out ones they do, and schools must have the freedom to specialize and innovate. It is the only way to make the schools respond to those seeking the education, not those providing it.
Posted November 19, 2008, 11:19am
Wow! Now we have everyone involved in the discussion and a lot of topics being addressed. I'd like everyone to chime in on a few basic questions:
1) Is the primary problem with NCLB that it is at the national, not state or local level?
2) Is the primary problem centralized accountability per se?
3) Is the primary problem inadequate resources and other program support (like lack of pre-K)?
4) Is the primary problem that specific provisions need to be tweaked, like switching to growth models?
My reading of what people are saying would classify Richard R as (1), Neal as (2), Andy, Joe, and Richard K as (3), and Sandy, Rick, Marty, and Elaine as (4). Am I mis-reading anybody?
Posted November 19, 2008, 11:15am
I, too, think that Title II should be restructured, like Richard’s call for more innovation in accountability we should likewise be supporting a lot more innovation around teacher quality in terms of changing how we recruit, train, induct, evaluate, and compensate teachers. That said, after the last seven years I’m increasingly leery of a national definition around teacher effectiveness. The first foray into this, the “highly qualified teacher” definition turned out to be a big paper chase with scant attention to effectiveness. Rather than try to perfect that, something I’m not sure is possible given the state of play in the field or even by moving to effectiveness as a definition, I want to see states (a) get better data systems in place so they can really analyze teacher effectiveness, and (b) invest in policy and structural innovations in this space. There are things the federal government can do to help with both in a new version of ESEA—support new ideas financially, create incentives, etc…
Posted November 19, 2008, 11:11am
Joe, I think the key to encouraging suburban districts to take in low income transfers is to provide them with hefty financial incentives (per student) for doing so. In Missouri, state money flowed to suburban districts taking in urban students as part of a desegregation order, but then when the court supervision was lifted, Republican state legislators in the suburbs argued for continuing the program voluntarily because their districts had become accustomed to the extra funds.
Posted November 19, 2008, 11:09am
When you say “by nearly any means necessary,” Joe, do you include private-school voucher or tax-credit programs? With good Catholic schools going out of business, and almost no incentive for new private schools to open and compete with “free” public schools, isn’t it time to give parents the money currently going to school systems and finally create a market? In addition to being the key to getting out of the incentives trap that constantly torpedoes well-intentioned efforts like NCLB, as well as promoting competition and innovation, isn’t school choice simple justice for the poor who have the least say of anyone in our current system?
Posted November 19, 2008, 11:06am
If readers are really interested in Kahlenberg – Dillon debate about inter-district choice, they should also check out Erin’s response to Rick’s critique. Erin’s got the goods in terms of the logistical challenges but that doesn’t mean that Rick’s idea isn’t one worth pursuing where we can, only that there are limits to what we can expect in the near-term at any scale and like any idea we should all be careful not to fall in love too fast. In other words, while I don’t dispute that there are some good examples out there the aggregate data should give pause.
And it’s that aggregate challenge that makes laws like No Child Left Behind so tricky. 50 states, 14,000 school districts, 100,000 plus schools, multiple choke points and policy irregularities as well as intense constraints on the ground—it’s a really challenging environment to make policy in and the critics will never have to worry about running out of things to say.
Posted November 19, 2008, 11:00am
I agree completely with Rick about moving to teacher effectiveness. Title II should be completely restructured to promote systems within states that measure for teacher effectiveness, support research-based assistance to teachers to become more effective, encourage differentiated pay and pay for performance, and increase the supply of highly effective teachers.
I do believe the teacher quality provisions in NCLB that were promoted by groups like EdTrust were positive and helped. There were no ideas then on the table to go where Rick and others are now proposing. This is clearly an area for improvement in the law.
Posted November 19, 2008, 10:53am
By the way, I meant "promoting" interdistrict transfers, since they are currently "allowed" but are not widely part of the solution.
I agree with Andy that getting students into better schools is crucial, by nearly any means necessary at this point.
Posted November 19, 2008, 10:48am
Glad to be joining you all and sorry to have been out of the email loop for a while.
Richard's suggestion about allowing interdistrict transfers is one worth exploring, especially since so many district leaders claimed they couldn't allow transfers to good schools BECAUSE THEY DIDN'T HAVE ANY! (And we don't exactly see many districts welcoming transfers from elsewhere with open arms.)
So how do we bypass those districts and get directly to the students we want to save?
Posted November 19, 2008, 10:46am
On testing, I would make three points. First, many of the tests we currently use are simply not that good. They do not provide adequate assessments of the range of knowledge that we want to foster. (Richard has made similar points, although he goes in a different direction than I do on this. It is possible to have broad tests that measure complex thought and still have standardized tests). Second, having a single cut-off does not make any sense. This encourages states to choose low levels of "proficient"; ignores people at the top; and leads to some bizarre school responses. Third, at least a portion of any accountability should be based on individual growth—adjusting implicitly for different starting points and the like.
These are all straightforward adjustments.
There is one mysterious element in this conversation. Nobody has brought up the issue of teacher quality. The only thing that we reliably know is that teachers are extraordinarily important. Moreover, we have "teacher quality" as perhaps the worst element of NCLB of 2001. In that legislation and the subsequent implementation of it, judgments about teacher quality have, if anything, led us in entirely the wrong direction.
At a minimum, any new version of NCLB has to scrap the old elements that emphasize credentials instead of effectiveness. At best, it would push in directions that supported truly effective teachers.
Posted November 19, 2008, 10:43am
I agree with Elaine totally about use of a growth measure that credits truly meaningful progress toward high goals. We thought about growth models when we fashioned NCLB, but virtually no state then had the data and related capacity to implement such a system. I worry that too few still do today. But we should indeed move in that direction.
Posted November 19, 2008, 10:41am
There is so much inaccuracy and bombast in what Richard R. has written I don't know where to begin.
A close examination of national education statistics shows two important facts: 1) gains in the NAEP have occurred fairly evenly through the deciles, disproving the notion that teachers have narrowed their attention to those who would help them "make the numbers," and 2) course taking across the curriculum (as measured partly by hours spent in the subjects) has remained constant over the years despite the emphasis in state policy and NCLB on math and reading.
I'd love to get into a battle with Richard R over his extravagant ideological views, but time and space push that off to another occasion.
I'll get cites for this data later, if desired, but I'm operating around meetings today on my Blackberry.
Posted November 19, 2008, 10:37am
I'm more optimistic than Andy about the logistical possibilities of creating integrated schools. In St. Louis, an interdistrict program provided opportunities for roughly a quarter of the urban student population to attend suburban schools, and the results were quite positive on academic achievement, graduation rates etc. For more background on 8 successful interdistrict public school choice programs, see a new report by Amy Stuart Wells and Jennifer Jellison Holme, available in a book The Century Foundation just published (available here and summarized here)
Erin Dillon's analysis took a good idea—mapping the distances from bad schools to good schools—but applied it in what I view as a flawed manner. She assumed much shorter driving distances than students in current interdistrict programs are willing to traverse. And she made an arbitrary assumption about how much space there is in good suburban schools. For my critique, see here.
The good question Andy raises is whether we can give low income students "better schools" that are economically segregated. I'm encouraged by examples like KIPP, but as Richard Rothstein has written, there are questions about whether this is scalable.
To my mind, NCLB provided a great philosophical breakthrough when it said that low income students stuck in bad schools have the right to transfer to better performing ones. Now I think it's time to make good on that promise by allowing more students to transfer across school district lines.
Posted November 19, 2008, 10:34am
A mixed reaction to Richard Rothstein. I have never seen anything convincing indicating that American states or districts systematically do poorly due to inadequate funding (plus, I’d note that there seems to be little correlation between high-poverty and low-spending, at least down to the district level). I think Richard R. makes an outstanding point, however, in pointing out that there is much more to education than test scores, and that all kids are unique. Only a flexible system that enables autonomous schools to tailor products to unique subsets of kids can maximize outcomes for everyone, which again points to school choice. Add to this the need for widespread competition to create real innovation (sorry Andrew, not the kind likely to come from Washington) and there are myriad, critical reasons for scrapping bad laws like NCLB beyond the very important need to hold schools accountable for math and reading outcomes.
Posted November 19, 2008, 10:30am
I’m not sure I like the new 10th Amendment model Richard Rothstein! But, I’m all for letting the states experiment more here and that’s a very good idea but two caveats.
First, that innovation has to happen within the bounds of ensuring equity for traditionally underserved populations. And while that might sound like an obvious point, a quick look at state school finance, accountability policies, and so forth gives the lie to the idea that states are the font of all that is good and wise here. And I say that as a state board of education member. So some bottom line federal accountability rules matter as a floor but within that framework states should take the lead here and federal policymakers should be willing to create policy solutions to let them innovate more.
Second, we have two experiences over the past seven years that also should raise a caution flag. The NCLB requirements are not a ceiling for accountability, they are a floor. Basically they require some percentage of students to pass state tests at the level that states have defined as proficient. Often that percentage is not especially ambitious nor are the tests and the pass points (called cut scores in the jargon). Beyond that states are free to, and encouraged to, go beyond them and do a variety of other things. They haven’t. And a look at state accountability systems, which generally operate parallel to No Child is also sobering. At the same time, when the Secretary of Education invited the states to submit their wildest blue sky ideas for how to do accountability differently the applications were very disappointing in terms of new approaches or ambitious ideas. This likely speaks to a variety of factors, capacity, political will, and so forth. But it does also point to the need for some federal pressure if we really want to see action on these points. It’s interesting that the other state board member in this conversation also pointed to the positive pressure that the law creates, despite its problems.
Posted November 19, 2008, 9:55am
I’m all for more integrated schools and the problem of increasing school segregation – in other words we’re going the wrong way on the promise of Brown – this should concern all Americans. But, efforts to better integrate schools are substantially complicated by two challenges: Housing and an inadequate supply of good schools. Our schools are increasingly segregated because our neighborhoods are and schools – especially elementary schools – tend to draw students from relatively small catchment areas. So we have a broader housing situation.
Second, there are just not a lot of good schools in too many communities. Addressing this, of course, is one of the primary impetuses behind No Child Left Behind. Earlier this year, using mapping technology and data about school performance Education Sector analyst Erin Dillon took a hard look at exactly what transfer options are available to parents and found that within a 20 or even 40 minute drive there are not a lot of options. In other words, while Rick’s idea holds great promise in theory, and we should vigorously pursue it on the ground where we can, in the near-term it’s not a scalable solution to the problem of low-performing schools in too many communities. (For the near-term Dillon offers some ideas on how to improve NCLB’s choice provisions to help with this).
The best strategy to get youngsters who currently have lousy schooling options into better schools, at any scale, is to actually give them better schools. That involves really changing existing schools and opening new public schools. The other ideas can help at the margins but will not get to the core of the problem.
For the federal government and No Child Left Behind this points to a much greater emphasis on opening new public schools and new charter schools in underserved areas and also on a much greater emphasis and investment in efforts to turn around persistently low-performing schools (pdf).
Posted November 19, 2008, 9:48am
I am grateful to Rick Hanushek for saying that “I am sure that Richard does not mean it, but some people interpret discussions like his as implying just pump more money into the system but don't try to measure or judge the effectiveness of the added funds.” I definitely do not suggest this. I am not in favor of just pumping more money in. Accountability is essential. I apologize for seeming to be self-promoting in this, but as I said before, I have just published a book (Grading Education) which lays out a detailed roadmap for how we should design an accountability system to judge the effectiveness of spending. But, to focus on the topic of this dialogue, I argue that this accountability system should be state-based, not federal, for the reasons detailed above. As Rick notes, I argue for accountability that includes standardized testing, but that goes far beyond it, assessing a broader range of knowledge and skills and building not only on our testing expertise but also on an existing accreditation system which is badly in need of reform. But the fact remains that some states do not have the fiscal capacity to support adequate schools, and only the federal government can address this. My recommendations about accountability in this book may be wildly off-base, but I urge that critics challenge them, not suggest that I am not in favor of accountability at all.
Posted November 19, 2008, 9:45am
One of the participants in this dialogue suggested earlier that only the federal government could be relied upon to defend the rights of minorities and disadvantaged children. This has been true at some points in our history (Brown v. Board of Education, civil rights laws from 1964-1974), but at other points, the roles have been reversed. In the last decade or so, the most creative education reforms have come from states in the old Confederacy. Georgia has led the way in universal pre-K. High school reform is more advanced in the SREB states than elsewhere in the country. When you think of “education governors,” they are almost all Southern. In the present period, there is no basis for believing that the federal government is a more reliable defender of the rights of disadvantaged children than the states.
In response to Neal McCluskey’s point (and a similar point by Sandy Kress) that “money isn't our problem—we spend more than almost any other industrialized nation on K-12 education but get much less bang for our buck,” a couple of points:
a) I do not suggest that money is “our problem,” but only that if money has anything to do with outcomes, it is irrational for the federal government to hold states with vastly different fiscal capacities accountable for similar outcomes.
b) Other industrialized countries have significantly less socioeconomic inequality and greater non-school social supports for children, so the “bang” our schools are expected to achieve is much greater than the “bang” in other industrialized countries.
c) Other industrialized countries do not spend less on children’s services (including schools, but not only schools) that contribute to student achievement. They spend more.
d) The states I referred to in my previous post – places like Alabama, to use Richard K.’s example – do not spend more than other industrialized countries on K-12 education; they spend much, much, less.
Posted November 19, 2008, 9:41am
NCLB is a failure, and should be scrapped, for these reasons:
a) By requiring accountability only for math and reading, it distorts the goals of education. In any institution, if principals hold agents accountable only for some of the institution’s goals, agents will distort their behavior to accomplish only those things for which they are held accountable. This is rational behavior. It is the reason the Soviet command economy was so inefficient (textile mills required to make adequate yearly progress in the production of yards of cloth produced only useless narrow widths; trucking firms required to cover more miles drove around in circles, etc.); it is also the reason that NCLB-type accountability programs in job training, welfare reform, and health care have been abandoned. Local agencies held accountable for placing workers in jobs concentrated on placing workers most likely to be placed without agency assistance, not those most in need of assistance; hospitals required to improve the survival rates of cardiac surgery patients refused to operate on the sickest patients, etc. The biggest tragedy of all this is that goal distortion has been most severe for disadvantaged children, because these are those for whom there is the biggest payoff for substituting math and reading drill for instruction in other subjects and behavioral traits.
b) Leave aside any racial or socioeconomic differences in achievement. There is sufficient variation among youth, irrespective of social background, that a single standard of proficiency cannot possibly be “challenging” (to use the NCLB term) to below-average, average, or above-average children. The problem is not whether proficiency is defined too high, or too low, or differently among states. Any standard of proficiency is irrational if applied to all students.
c) NCLB permits no adjustment for socioeconomic differences, although these are the most important determinants of student achievement, on average. (Yes, I know, some poor students excel, just as some smokers don’t get lung cancer, but on average, socioeconomic disadvantage has a big impact. As you know, I’ve written a book on this topic, but forget me as an authority - Janet Currie has estimated that 25% of the black-white school readiness gap can be explained by differences in a few measurable health characteristics of children and their mothers; our own Rick Hanushek and colleagues have estimated [for Texas] that about 14% of the racial achievement gap can be explained by differences in student mobility rates. Etc.)
d) An accountability system based on test scores alone has stimulated nationwide score inflation, as teachers and administrators have naturally learned how to game the system, both by legal means and by unlawful cheating. The nationwide improvement of state test scores, relative to NAEP scores, is evidence alone that NCLB and its required state accountability systems have resulted more in gaming than in improved instruction. We are back to the “Lake Wobegon effect” of the minimum competency movement of the late 1970s.
The above does not mean that, in principle, the federal government could not design and administer a satisfactory accountability system. But federal micromanagement of education (NCLB) has now proven to be so incompetent, for reasons stated above, that it would be foolish to rush headlong into another federal system, as poorly thought through as NCLB was. Why not let the states struggle with these very difficult challenges? Some may design better accountability systems, some may stick with NCLB-type systems. But after the NCLB fiasco, we’ve got little to lose by letting states experiment with alternative accountability systems.
Posted November 19, 2008, 9:26am
I ask everyone, how do you change politics so that the demands and needs of children are put ahead of those of politicians and the people employed in public schooling? That is the rubber-meets-the-road question, and I’ve never seen it satisfactorily answered. Until that can be done—and as NAEP, or, say, PIRLS and PISA, data strongly suggest, it hasn’t been done yet—all that federal intervention will do is spend more taxpayer dollars, offer more promises, deliver more bureaucratic burden, and not change actual outcomes in any significant way.
Now to Sandy’s point. Let me make clear, because there is a tendency to argue that if someone dislikes NCLB they must have liked the system before it, that there was nothing I found “effective operationally” about previous standards and accountability efforts. But I cannot see what is better about NCLB, and contrary to what Sandy writes, as long as consequences are attached to “proficiency,” that is what will concern state policymakers. Yes, they could aim everyone toward learning beyond that, but there is no strong incentive to do that; hence few states do. Of course, put in strong incentives, and you’ll see widespread evasion, just like under NCLB and every federal effort before it. Finally, I don’t blame NCLB per se for weak academic standards. I have been very clear that I blame the entire top-down system of public schooling in which, inherently, the people employed by the system, whose interest is to have low standards and high revenue, have the most power and parents have relatively little. That is why decoupling funding and provision of education is the only solution that can really fix American education.
Posted November 19, 2008, 9:21am
I think Andy makes an excellent point about investing big in pre-k, and I'd like to see it be universal as well, in part because if we have separate pre-k programs for low income and middle class kids, we'll lose the benefit that comes from exposing low income kids to middle class peers. [link]
I'm rarely accused of being "fashionable" (thank you, Andy) so let me advocate an idea for spending NCLB funds that is little discussed today, but very important for low income kids. Research finds that the single best thing you can do for a low income student is give her a chance to attend a middle class school—where, on average, the peers are more academically engaged, the parents more likely to volunteer in the classroom and hold schools accountable, and the teachers more likely to be high quality and have high expectations. Compulsory busing is a non-starter, but during the primaries, Sen. John Edwards proposed providing financial incentives for middle-class suburban schools to accept low-income NCLB transfer students. [link] Recognizing that movement of students should flow in both directions, he also proposed doubling magnet school funding so that urban schools could attract middle class kids from the suburbs. In my view, the continued efforts to make separate schools for rich and poor perform well will work only on rare occasions.
Posted November 19, 2008, 9:03am
Neal - I've reviewed the states you cite. All the states that have "lowered" their standards in response to NCLB have done so because their previous "higher" standards didn't pinch much and they preferred to have more achievable standards if they were to pinch. The question I'd pose to you is how effective operationally did you find the previous standards to be in actually improving teaching and learning and in "pinching" the system to change and improve when students failed to meet them? Without such a finding, it's unclear that anything but paper standards were actually lowered.
Further, as I mentioned, I'd be pleased to show you how a state could strengthen its standards without being "penalized" by NCLB. Current performance standards could be used for AYP determinations of "proficient" performance, for example. And higher standards (say, advanced) could denote "being on path to post-secondary readiness." The state would be free to set those standards where it likes, determine whether they must be met or whether a measure of growth to them is sufficient, and then establish whatever consequences it wishes.
Blaming NCLB for weak academic (and especially performance) standards is typical and inexcusable scapegoating.
Posted November 19, 2008, 8:59am
Welcome colleagues. Sounds like you have had an engaging conversation so far. I would love to jump in at this point. First, no, I don't think NCLB should be scrapped but the accountability provisions need to be revamped. As a member of the State Board of Education in Colorado and formally the Denver School Board, I have had first hand experience working with NCLB. Overall, I believe the attention on accountability has created a very healthy tension for schools and policy makers. For years, students and schools were failing and there was little attention given to reversing that trend. With all the moaning about testing required under NCLB, finally, we were forced to pay attention to the children our system has failed.
But, the way we were conducting the testing was neither fair nor terribly useful. Under the next iteration of ESEA, we, in Colorado, propose using a growth model, which would measure individual student progress toward a common destination. I would argue that a coherent accountability system must move from evaluating the status of the system to the progress of students—and, in so doing, allow an evaluation of system effectiveness for accountability purposes. This new system must:
1. measure individual student progress toward reaching mastery and remaining at that level or moving up;
2. measure the effectiveness (or productivity) of classrooms, schools, districts, and states in promoting and graduating students that reach and remain at mastery or better;
3. deliver credible and useful data to educators and the public to support learning and improvement and to evaluate effectiveness.
To sum it up, we need to hold schools accountable for all students being ready when they exit and have an assessment system that measures how effective we are at getting that done.
Posted November 19, 2008, 8:53am
Jay asked a good question about whether it’s time to double-down around the federal role or try something else. That question, just like the accountability or choice one that Rick Hanushek responded to, is too reductionist. What we need to be thinking about is how to make the federal role sharper and more strategic and how to best align it with the opportunities and challenges that exist today.
Every version of the Elementary and Secondary Education Act (No Child Left Behind is merely the most recent set of changes) has improved on the preceding version and also created or caused a host of new problems. This version of the law is no exception. So, despite fantasies from the political right and left, the law isn’t going anywhere. It’ll be changed when it’s overhauled during the next few years (hopefully) but unless we completely walk away from an emphasis on accountability, we’ll go through the same song dance with the next one and the general framework will look much the same.
So how do we make the federal role more strategic?
Well first, while it’s fashionable to say, as Rick Kahlenberg does here, that we need to “provide greater resources”, the real question facing policymakers is greater resources for what? Just putting more money through the same pipes for the same things is unlikely to dramatically change what military analysts would call the facts on the ground. In fact, when you consider those dollars on a per-student basis it’s hard to see how we’ll revolutionize American education – especially for disadvantaged students – on $1500 or even $2000 per head. So for my money we need to be thinking about how to spend scarce federal dollars to maximize their impact.
The big investment I’d make is pre-kindergarten education to help get all students – regardless of income – off to a good start in school. Anyone who has spent time in kindergarten classes with low-income students can vividly see the early gaps in achievement that happen before students even start school. The research on the efficacy of pre-K is mixed overall but the logic model is sound and good programs do seem to help students.
We should also invest in expanding support for high quality public charter schools, research and innovation efforts, and a much more reform-oriented federal role around teacher quality than we have today. President-Elect Obama’s commitment to double federal funding for charter schools and create a “grow what works” style fund are solid steps in this direction.
Finally, rather than ease up on the vital emphasis on holding schools accountable for the achievement of all students – especially traditionally underserved students – I’d couple that emphasis with much more investment in helping states and school districts intervene in low-performing schools and even close them and offer students better options. After six years of NCLB it’s clear that more needs to be done here. In fact, rather than all the sky-is-falling talk that permeates the debate about No Child, it seems like very little has actually changed in our most challenged schools.
Posted November 18, 2008, 5:51pm
Great having you with us, Marty. It sounds like you are saying there are problems but they could be fixed. What are the fixes you had in mind? (And here we are starting to transition to a discussion of solutions and not just identifying the problem).
Posted November 18, 2008, 5:45pm
Apologies for my late arrival. It is an honor to be joining such an esteemed group.
Above all else, NCLB held out the promise of greater transparency about the performance of American schools in reaching basic educational goals—something that was and remains greatly needed. Unfortunately, it is by now clear that this promise has not been fulfilled.
Earlier in the day, Richard Kahlenberg alluded to the well-known distortions created by varying state proficiency standards. Put simply, the percentage of students who are proficient or the percentage of schools making Adequate Yearly Progress (AYP) under NCLB tells us next to nothing about the relative performance of state school systems.
Not yet discussed is the fact that the information NCLB provides about the relative performance of schools within the same state is also deeply misleading. With minor exceptions, the law’s binary ratings are based solely on the level of achievement in a school rather than the progress its students are making over time. Schools making AYP are often no more effective than those that do not; even high-performing schools serving disadvantaged students are likely to be sanctioned. (See here for evidence on this point from Florida.)
Transparent information about school performance is vital to empower voters in local elections and for choice-based reforms to achieve their full potential. In the end, the question of whether we should scrap NCLB comes down to whether Congress is willing to revise the law sufficiently so that it contributes to that goal. If not, the federal government runs the risk of undermining the promise of school accountability altogether.
Posted November 18, 2008, 5:41pm
I remain stumped by the fact that the US is the second highest per spender of countries in the OECD and yet our performance is middle-of-the-pack.
Posted November 18, 2008, 5:29pm
I await Richard's evidence that the difference in spending across states leads to differences in performance. Perhaps it does not show up on NAEP because NAEP doesn't measure the range of things that Richard would like to measure or does not accurately judge the depth of knowledge. I am sure that Richard does not mean it, but some people interpret discussions like his as implying just pump more money into the system but don't try to measure or judge the effectiveness of the added funds.
Posted November 18, 2008, 5:04pm
Thanks Jay! I knew there was something to this theory, though I wasn’t even thinking about competition (surprisingly!) just the ability to know what’s going on in your district and assert pressure.
I was also reminded of something: While most people probably do not spend a lot of time getting to know local school board candidates or following the day-to-day activities of their districts, they often become very attuned to bond referenda because those have a tangible impact on them. Their votes also matter. Contrast this to the impact of voters on federal budgets—much less the education component of federal budgets—and local control in at least one respect is much more responsive to constituents than federal control.
Posted November 18, 2008, 4:46pm
This is a digression, but FYI on whether smaller districts contribute to student learning:
Caroline Hoxby’s AER article, Does Competition Among Public Schools Benefit Students and Taxpayers?, is essentially an analysis of the effect of district size on student achievement. It finds that more numerous smaller districts improve achievement. But you should be aware that Jesse Rothstein (son of our own Richard Rothstein) has disputed those results.
A number of other studies in Clive Belfield and Henry Levin’s review of competitive effects in education also examine district size and the results are generally that smaller, more numerous districts contribute to student learning, all else equal.
Posted November 18, 2008, 4:41pm
Sandy, absolutely states define “proficiency,” but the federal law demands that they do it, and threatens to punish districts if they fail to hit it. That is why there are numerous, very well documented cases of states changing cut scores for proficiency, altering the timing of tests, fiddling with sub-group statistics, and just generally setting or maintaining very low definitions of the term “proficient.” To stick with a developing theme, I review many of these evasions in my book Feds in the Classroom: How Big Government Corrupts, Cripples, and Compromises American Education, but you can go here to see the very low proficiency definitions for most states, or here to read about how Michigan responded to NCLB, or here to catch up on questionable goings-on in Maryland, or here to see how South Carolina—a state with relatively high standards—has reacted to the NCLB problems that go with such standards.
As for Richard R.’s funding point, I think there is lots of evidence that money isn’t our problem—we spend more than almost any other industrialized nation on K-12 education but get much less bang for our buck—but I don’t see how we could expect Washington to do long-term equalization even if money were the problem. The incentives for politicians are just not there because while the poor would be seeking concentrated benefits, they are a very weak interest group.
Posted November 18, 2008, 4:25pm
Thanks for joining, Richard R. I’ve also heard from Joe Williams and Marty West, who have also just got out of all-day classes/meetings. So the discussion is about to broaden considerably. Now we are cooking with gas!
Richard R., I know you’ve referred folks to your new book, but for the sake of those who haven’t had a chance to see it, I’d like to be clear on your reasons for wanting to scrap NCLB. The objections you provide here are that NCLB does not address inter-state inequities in funding and focuses too narrowly on basic math and reading skills. Could those issues be addressed without scrapping NCLB? We could pump a whole lot of new money into NCLB to address your money concerns and we could develop a broad set of new tests or other ways to measure achievement of desired goals. Why get rid of NCLB?
Also, if you want to get rid of it, what do you want instead?
Posted November 18, 2008, 4:20pm
Richard K., you could very well be right, though I would love to see socio-economic status controlled and then the effect of district size measured. It is quite possible that someone has even done that and I just haven’t seen it yet. Regardless, I’m sure SES is a much bigger factor, but I’m not prepared to assume that district size doesn’t have some effect.
Posted November 18, 2008, 4:11pm
Neal, one small point. The argument that Massachusetts and New Jersey do better because they have small districts strikes me as quite weak. As Richard Rothstein has noted in his work, the socioeconomic status of the family a child grows up in is the critical predictor, and MA and NJ are pretty wealthy. Either that, or as Pat Moynihan used to quip, proximity to Canada, explains performance.
Posted November 18, 2008, 3:50pm
Neal, remember that NCLB does not call for a federal definition of "proficient." The states make the determination. Indeed the notion that it encourages a dumbing-down is without any foundation. States have all the tools to raise standards without being "hurt" by NCLB. They can do whatever they want with their standards. I'd love to have just 5 minutes with any creative state policymaker, and we could show how this is so.
As to greater transparency, options, and competition, you'll get no argument from me.
Posted November 18, 2008, 3:43pm
Please accept my apologies for coming late to this discussion. I am today attending a conference at Teachers College, and have stepped out to weigh in on this topic of the federal role; many of the papers at this conference are relevant to our discussion, and I commend them to you.
I am in the “terminate NCLB” camp. I regard its problems as unfixable. I have described my basis for this claim in my recent book, Grading Education – Getting Accountability Right. Rather than elaborate on these reasons in my initial post, I will instead begin by discussing what the federal government should do in education.
In an earlier post, Richard Kahlenberg (quoting Albert Shanker) defends federal involvement in education policy by asking, “Should children in Alabama learn a different kind of math or science from children in New York?" But if the obvious answer to this query is “no,” then how can we defend that children in Alabama are expected to learn math or science with a fraction of the resources used for teaching math or science to the children in New York? For the last twenty years, reformers have mounted a series of lawsuits attacking the intra-state inequality of school funding. But only 1/3 of the variance in school spending is attributable to between-district differences within states. Two-thirds is attributable to inter-state differences. Low-spending districts in New York have more (real, cost-adjusted) funds to spend than high spending districts in Alabama. It is hard to see how the federal government can hold New York and Alabama to common standards until the enormous differences in fiscal capacity (not attributable to differences in effort or will) are minimized. This will be politically difficult to do, because it requires legislators from states where public spending is supported to vote to send their own tax funds to states where public spending is disfavored. But unless the federal government implements a large-scale inter-state redistribution program (block grants for children’s services to states with low fiscal capacity), it is hard to see how we can take seriously a federal attempt to enforce common educational standards, no matter how well designed or intentioned.
Second, a most serious flaw in the accountability provisions of NCLB has proven to be the goal distortion that has resulted. Holding schools accountable only for easily-tested basic math and reading skills has led to the inevitable de-emphasis, especially for the most disadvantaged children, of other educational goals that we can all agree are important—history, science, character development, citizenship, behavior, health education, physical education, the arts, and critical thinking in all of these. It is hard to see how an accountability system at the state or federal level can overcome this flaw if we don’t have information on student outcomes in these other domains. Forty years ago, NAEP actually did assess these many domains. NAEP’s focus was narrowed to basic academic skills under budgetary pressure in the late 1970s. Restoring this focus, and generating state level data on, for example, whether students know how to solve problems cooperatively, will be an enormous undertaking. Only the federal government can undertake this. Until it does so, no accountability system, whether at the federal or state level, can avoid goal distortion. (Again, I’ve discussed this in greater detail in Grading Education.)
Posted November 18, 2008, 3:30pm
Jay, you've asked a great question. My own thought is that a constructive tension between the feds and the states is what we should seek. I've heard from numerous state leaders that a real positive of NCLB is that it has given them "cover" to strengthen their own systems. It's also interesting to see states like Iowa and Nebraska go to stronger state roles. (Remember that one under-recognized feature of IASA and NCLB has been to boost the role of the state vis-à-vis the locals.) States have also become more responsible for more solid policy on such issues as graduation rates and educating the disabled. All of this speaks to the desirability of federal pressure. But it's not either/or. There must be state regimes. And the feds need to listen more and show greater flexibility in the tough areas, including proper assessment of the disabled and the English language learner, rigorous growth measures, etc. Nuances in consequences for low performing schools also must be recognized and permit for variation in state policy.
NCLB was a positive, but by no means, final step. When we get to perfecting federal law and policy, we can discuss the details. But, for now, I want to argue for federal/state partnership in which both well structured funds/support AND expectations/accountability flow to the states, and the states then fundamentally set up their distinctive regimes. In the absence of a national system of education (which is not viable as far out as I can see), this is, in my view, the best and right approach.
Posted November 18, 2008, 3:10pm
Let me start with Jay’s invocation of Federalist #10. Publius argued that a republic would dilute the power of factions, hopefully overcoming them. But Federalist #10 relied largely on a geographically extensive republic, and while the physical size of the United States hasn’t shrunk since Publius’s time, its virtual size has. Telephones, Blackberries, the Internet, all have enabled factions to communicate at the speed of light, and with that ability the best place for them to concentrate power has been at the federal level. It’s also no mistake that almost every education special-interest group has its headquarters in DC.
How about local control? Unlike federal control it dilutes the power of factions overall, and while, yes, people are often ignorant of what their school districts are doing and don’t participate in school board elections, how many do you think read federal education laws, or federal education regulations, or follow education testimony in front of congressional committees? I would guess almost none. Heck, when he signed NCLB, President Bush himself said that he hadn’t read it! And it is quite possible that many people pay little attention to local education authorities because “local” has lost so much meaning, with massive district consolidation over the last several decades and increasing state and federal control. Interestingly, though—and I’ve not seen any research into the effect of this, so it’s just something to ponder—it seems that many of the states that have better academic outcomes, like Massachusetts and New Jersey, have small rather than county-sized districts. Perhaps more meaningful local control makes a substantive difference.
As for Eric’s points, I agree that choice is critical to accountability, but disagree that somehow action by Washington is the best way to get parents the information they need to make wise choices. As I and many others have noted about NCLB, for instance, the information parents get from it is largely deceptive, being told their kids are “proficient” but not how little learning “proficient” actually represents. And why is Washington the ideal place to set standards? That was tried in the 1990s and was a failure because no one can agree on what the standards should be. Moreover, why should we expect that tough federal standards attached to real sanctions won’t be dumbed-down, or if they’re not attached to sanctions they won’t just be ignored? Those are the most likely combinations given the incentive structure in top-down reforms. In contrast, in free markets we have good, meaningful standards and information providers that align the incentives of consumers, producers, and referees, whether we’re talking about Auto Week or UL listings. Each one is a strong check on the other, and all three profit most when each one does its job.
Posted November 18, 2008, 2:02pm
I personally do not think the issue should be posed as accountability or vouchers. Indeed, it seems to me that you need both improved accountability and real choice in schools. Choice does not work well without good information. And accountability does not work well if you simply tell those schools that have not succeeded to do better. Moreover, without effective choice, the system itself will tend to eat away at accountability to the point where it does not have much impact.
On the role of the federal government, I think that there is a clear rationale for federal involvement. First, people are very mobile, and the education in one state has direct impacts on what goes on in other states. Second, there is, in my opinion, a national imperative to improve our schools. Our future well-being as a nation depends on improving our development of human capital. In this, the federal government can help in developing educational standards, in developing assessment instruments, and in prescribing performance around the country. It is less clear that the federal government can tell states or localities how to do it.
It seems that there is some confusion in NCLB about the best way for the federal government to interact with states and localities. But there is little doubt in my mind that the federal government should continue to be involved in accountability issues.
Posted November 18, 2008, 1:53pm
Let’s assume, for the purposes of this current discussion, that large-scale school choice is not a possibility (we can change that assumption when we move on to solutions). If choice is off the table, then we are back to Sandy’s concern about how we motivate schools to improve. And while Neal makes a fine theoretical argument about why localities should be more effective, their track record has been very bad.
In fact, one could make a theoretical argument that localities are less likely to be effective and more likely to be dominated by organized interests. The decisions of school boards and school officials are given less attention by the media and subject to significantly less public scrutiny. Local school board elections on off-election days have very low turnout, often in the single digits. Given the obscurity of local school politics, it’s easier for the employees and their organized interests to dominate school politics. They’re just about the only ones following what is going on and voting in those elections. At the national level there is at least enough information, scrutiny, and organized advocates on all sides, that it isn’t so easy for anyone to dominate the debate (see Federalist 10).
So, if choice is off the table, we are left with having no or largely weak accountability regimes, an organized national accountability regime (like NCLB), or a bunch of state accountability regimes. Among these choices, what do folks prefer?
Posted November 18, 2008, 1:21pm
Jay, we might expect to see delayed 12th grade improvements. However, Sandy said we have seen improvements for younger kids since the early 1990s, but 12th grade scores in the late 1990s and early 2000s have dropped in reading and simply not moved in math. There’s no sign that improvement has worked through the system, and lots of reason to conclude that federal policy has created a lot more dodging than learning.
As for the state and local levels, as I wrote earlier, I agree that they, too, have often performed horribly, largely because they have the same top-down, powerless-consumer structure as federal interventions. So they have not inspired confidence in me – they are just slightly less bad.
Ultimately, the solution is to decouple funding and provision of schooling, giving parents control of education money through school choice, letting schools be autonomous, and forcing providers to satisfy consumers. States and localities can compete with one another, and they have to answer to voters more directly on education than do the feds, but in the end letting parents be real consumers is the key effective reform.
Posted November 18, 2008, 12:57pm
Well, NCLB is targeted mostly at elementary and middle school students, so we might expect to see less and a more delayed effect in 12th grade results. And while it’s not possible to make any causal conclusions about NCLB or federal policy in general just by looking at NAEP average results, it should tell us something.
But let me also press Neal to address Sandy’s earlier argument. Neal, what about state and local action has inspired more confidence in you?
Posted November 18, 2008, 12:43pm
I’m not sure how Sandy makes the leap that since some scores have gone up—but all gains seem to disappear by 12th grade—federal reforms are working. If anything, we see the same old story that the longer kids stay in our school system, the worse they do. And even if scores were to improve across the board, how can that in any near-conclusive way be attributed to federal interventions? Why not school choice, which took off in the 1990s both in terms of real programs and, perhaps even more importantly, as a legitimate threat to the status quo everywhere? Or couldn’t it be that parents of young children are increasingly concerned about mathematics and pushing improvement in spite of a system that is loaded against them?
Given what we know about NCLB and IASA evasion, as well as toothless federal-education policy for decades before those versions of the ESEA, there is little reason to think federal policy is helping. Indeed, the rhetoric behind those laws reflects increasing demand for good schools, but their execution reveals mainly avoidance of that demand. Factor in the ballooning price tag, and I still see no reason — besides prohibiting discrimination — to keep the feds in education.
Posted November 18, 2008, 12:37pm
Oddly we have agreement on the observation that educational progress has been inadequate and federal intervention has largely been weak and mostly ineffectual. The debate seems to be whether that means we should do more at the federal level or do less. Sandy says we need to do more because he’s convinced that localities are often unmotivated to improve. Neal says do less because he can’t see why larger intervention would work when modest intervention largely hasn’t.
What does everyone else think? Do we go all-in or save our chips for something else?
Posted November 18, 2008, 11:52am
I agree with Neal that we do not yet see gains at the 12th grade on NAEP. We've seen flat scores on the SAT and ACT as well. Yet, 4th and 8th grade NAEP in math are way up, and 4th grade reading has begun to budge.
Having agreed to that, I'll really provoke you with this thought: let's emphasize the S in ESEA by getting serious FOR THE FIRST TIME about standards, assessments, resources, and accountability for the secondary level! Let's have policies that promote changes at the state and local level, particularly in strengthening high schools. Let's be smart about doing it, too, and evaluate the impact of such initiatives as well as others on getting students to graduation ready for college/career. Then perhaps we'll see gains for older students.
Posted November 18, 2008, 11:48am
I agree that the federal government has an important role to play in ensuring that states do not discriminate in provision of public education. Brown v. Board, the Civil Rights Act of 1964, the Office of Civil Rights—all are constitutionally legitimate, and Washington has done laudable work—though often with regrettable excesses—in ending de jure discrimination. Where it inevitably fails is in redistributing dollars effectively and accountably. I have discussed why it fails at accountability—the incentives are all wrong—and the interesting thing about Robert Kennedy’s demands for Title I is that they have never really been met. States and localities get more and more money, but still avoid almost any meaningful accountability with tacit federal approval, whether it’s NCLB’s toothless school-choice provisions or how states define “proficiency.” And federal compensatory aid, like accountability, has followed political incentives, not educational goals, over the years being distributed to wider and wider swaths of children (read: their voting parents and communities) rather than being targeted strictly to poor kids.
Making Washington the honest broker in education is a tempting idea, but doesn’t seem to comport with reality.
Posted November 18, 2008, 11:40am
Let me jump in on this issue of which level of government is best to implement accountability. My view is that there is no simple answer. I served as president of the Dallas school board. We rarely felt pressure to raise standards or hold educators accountable to delivering to those standards. On the contrary, the pressure was typically to protect the mediocre. I don't want to pick on Dallas, so I would ask this question: where's the evidence, broadly, in our big districts, of "local control" shutting down bureaucracy, closing achievement gaps, giving parents real and serious choice, and educating to high standards?
Look, I agree parents and teachers are key. A commitment locally to excellence is essential. And the states must make a positive contribution. But, federal law does not materially disrupt that control. States set the standards. States make the tests. States set the performance standards and specific consequences for performance. States and districts determine curricula and personnel policies.
The feds' role is rather limited. They apply pressure to fix schools where subgroups are doing poorly against state standards; they put pressure against sham graduation standards; they put pressure to give parents choices out of poor-performing schools.
Indeed the best complaint against NLCB, even better than its flaws in execution, is that it's weak! It doesn't combat standards that fail to educate students to college/career readiness; it doesn't require rigorous enough assessments and performance standards; it doesn't go far enough in insisting upon teacher effectiveness; and it doesn't enforce well enough parental choice.
My question to those who would scrap NCLB because it gets in the way of the locals is: are you broadly and generally satisfied with the quality of public education in our schools? And, if not, how, other than "voting the bums out" (which doesn't seem to happen), can the forces at play locally be changed to bring about improvement?
Posted November 18, 2008, 11:31am
I think before we give NCLB—or federal reforms before it—credit for improving outcomes, we ought to look at some data. Since the early 1990s long-term NAEP data does not show marked improvement in reading for any age group, save maybe 4th graders, nor does it show any improvement in mathematics for 12th graders, the final products of the K-12 system. NAEP subject tests do show decent improvement in mathematics, but almost none in reading, and subject-test data only goes back to the early 90s. That is hardly proof that federal reforms have worked. Indeed, the 12th grade scores show that whatever gains are made largely wash out before kids graduate.
Of course, it is very difficult to tease out the effects of federal laws versus the myriad other variables that affect academic outcomes, but I see no way to look at NAEP data and conclude that the federal government has been successful at improving education. What improvements there were could be attributed to increasing school choice, more state-level accountability, better nutrition, more-motivated kids and parents — lots of things. Then I see that federal K-12 spending has grown from inflation-adjusted $18.04 billion to $37.93 billion since 1990, and can’t help but conclude that, no matter what the cause of our small improvements, the federal investment has probably been a dud.
Posted November 18, 2008, 11:25am
Thanks, Sandy. I think the discussion is beginning to reach critical mass. Right now we are focusing on whether the federal level is the right place for accountability regimes, if we are going to have accountability regimes. Neal says no, because Washington is further away from parents, children, and schools, harder for those parents to hold accountable electorally, and more easily captured by special interests. Sandy and Richard K. emphasize advantages of action at the national level—1) it prevents laggard states or localities from failing to fulfill their responsibilities; 2) the national government is better suited for the “redistributive” effort of providing educational opportunities to disadvantaged students; and 3) increasing federal dollars require increasing federal accountability for the effective use of those dollars.
Does each side in this argument agree about the benefits and drawbacks of national action, but just weight those benefits and drawbacks differently? Or do folks disagree with the claims of benefits and drawbacks made by the other side?
Posted November 18, 2008, 10:53am
Jay, I think there should be a strong federal role in helping to establish a system of standards and accountability, and here I'll disagree with Neal. Although education is mostly run at the state and local level, the federal government has had a special role in promoting equal opportunity in education, which the states had failed to provide on their own. The federal government was right to establish the Elementary and Secondary Education Act (ESEA) in 1965, in order to provide Title I funding for low income schools and it was right to put pressure, through the 1964 Civil Rights Act, on school districts to desegregate. So too, given the flow of federal funds to states today, the federal government is right to insist on accountability, something that Robert Kennedy sought with passage of the original ESEA.
In fact, I think there is a strong argument for some form of national standards. I recently wrote a biography of Albert Shanker, who asked, "Is an understanding of the Constitution or the way to write a decent paragraph more important for students in some communities than others? Should children in Alabama learn a different kind of math or science from children in New York?"
Posted November 18, 2008, 10:45am
Thank you, Jay, for moderating and all the good work you do.
It will come as no surprise that I strongly favor an accountability push from the federal government. I liked the IASA, and I liked NCLB even more. Further, I think re-authorization of the ESEA can both fix problems that have arisen and also push education reform nationally to an even deeper level.
Fundamentally, as Bobby Kennedy argued forcefully in the mid-60s, the federal government should have a stake in improving education for disadvantaged students and being sure states and local districts are accountable to that end as well as in the use of increasing federal funds.
This involvement has been messy to be sure, but NAEP results in 4th and 8th grade math and at least recently in 4th grade reading show significant gains and gap-closing since federal involvement through standards based reforms began in the early 90s.
Groups as diverse as the Council of Great City Schools, the CEP, and a whole variety of civil rights groups (who rose in protest last summer over a proposal in Congress to strip NCLB's accountability provisions) agree that these policies have improved education in our country, particularly for disadvantaged students.
I'll close this entry with one final thought: many states and districts are, and have been, using the strategies of standards-based reform on their own and to the benefit of their students. Jay, you've shown this in your studies of certain states. Massachusetts, Virginia, Maryland, and Texas are good examples. But, in my view, federal policy has promoted, furthered, and helped spread the movement—sometimes painfully—throughout the nation. This is a good thing.
Posted November 18, 2008, 10:31am
Jay, you are correct that my primary concerns about NCLB are applicable to top-down education control at all levels—federal, state, and local. That’s why I agree with NCLB defenders when they rightly point out that leaving control at state and local levels is hardly satisfactory. That said, the worst place to locate education authority is in Washington, which is furthest from the parents and children the schools are supposed to serve and, hence, most likely to be controlled by special interests. It’s hard enough to get parents to regularly head to school board meetings; how are they all going to get to Capitol Hill? And consider voting, the one check parents supposedly have on top-down public schooling. At least with a school board voters elect candidates based principally on their education stands. Representatives in Washington, in contrast, are elected based on positions on foreign policy, the economy, abortion, health care, defense, energy, Internet policy, and on and on. There is no way for parents to put direct pressure to change education on federal lawmakers through voting, but voting is the only concrete power they have. Finally, there is federalism. The Constitution gives the federal government specific, enumerated powers because the Founders knew that there were just a few, inherently national, things the federal government had to do, and that it should do nothing else. In large part, this was done to disperse power, and the benefits of that are clear. In addition to avoiding tyranny, it allows individual states to experiment with lots of different things without bringing all the states down if they fail. In contrast, put all authority in Washington, and when things go sour there’s nowhere to run.
So, to answer your question, Jay, I think putting power in the hands of parents is the key to high standards and accountability anywhere, but it’s better to keep NCLB-like reforms at the state rather than federal level.
Posted November 18, 2008, 10:00am
Thanks, Richard K., for jumping right in as well. Your comment slipped in just as I responded to Neal’s, but let me ask you the same basic question. Could the goals you see NCLB addressing, such as guiding teachers about what to teach and motivating people in the educational system to improve, be done by state or local accountability systems? Why did this have to be done at the national level? In other words, is your argument for an accountability regime or specifically for a federal accountability regime?
Also, we’ll have time to get to the solutions you and Neal suggest later in the discussion. For now let’s focus on identifying what our basic goals should be. Thanks!
Posted November 18, 2008, 9:55am
Thanks, Neal, for getting the ball rolling. Let me press you on a few things. It seems like your concerns are not primarily with a federal accountability regime, but with any centralized accountability regime. That is, in your list of reasons for NCLB’s “extinction” you include “bureaucratic costs, narrowing the curriculum, stifling innovation, the Constitution” and most of all “the inherent incompatibility of political and educational incentives.” Other than the Constitutional objection, it seems like all of your concerns would apply to state or district programs as well. Would you be any happier if an NCLB-like program were done at the state level? If so, why?
Posted November 18, 2008, 9:51am
I’m in the “mend it, don’t end it” camp on NCLB. While the law has very serious problems, the core underpinnings are sound: We should set standards for what students should know and be able to do; use tests to see whether students are learning; and hold educators and students accountable for failure.
Before standards-based reform, there was a big hole at the center of American education: a lack of agreement on what skills and knowledge students should master. Teachers had textbooks but no real guidance on what to prioritize, so they were essentially asked to create their own curricula. Teachers ended up choosing very different topics to pursue, based on personal interests, which created confusion and incoherence. And there was very little outside pressure for anyone in the system to work very hard.
NCLB was right to require states to set content standards; right to authorize additional money to help students meet performance standards; right to require states to separate out the performance of low income and minority students to see whether they’re doing well; and right to provide students the opportunity to transfer out of bad schools.
Having said all that, NCLB needs a major overhaul: to provide greater resources necessary to help students meet standards; to fix the standards, testing and accountability provisions; and to allow low income students the genuine right to transfer to better-performing schools in the suburbs. We also need to supplement NCLB with efforts to improve housing, health care, and pre-k programs, given the longstanding evidence that the #1 predictor of academic achievement is the socioeconomic status of the family a student comes from.
Posted November 18, 2008, 9:37am
First off, I’d like to thank NewTalk for hosting this discussion. I was critical when I thought the first NewTalk NCLB chat put scrapping NCLB off limits, and I think it’s terrific that we now get to explore the idea.
Here’s the quick-and-dirty on why I think the law—and almost all federal education involvement—should be ended. While I think NCLB was well-intended, good intentions filtered through politics more often than not end up bad, mainly because it is very hard to properly align incentives. The strongest incentives for most of the people employed in public schooling are to maximize revenue and minimize accountability, just like I would be most happy if I could get paid $1 million to do whatever I want. That’s not to say that educators and I don’t care about standards, but we are all also constantly tempted to take the path of least resistance, and when push comes to shove it is easier not to have tough bars to leap, especially set by someone else. Politicians, for their part, find that the path of least resistance is to bend to the will of the most powerful interest groups—in this case, the people employed in education that they would supposedly hold accountable—or to do whatever has the best sheen for voters, and that has meant evading and destroying strict accountability. Finally, you get to parents and taxpayers. Their incentives really are to get the best education for the least money, but due to concentrated costs and diffuse benefits they are politically outgunned and, hence, almost never get what they want.
There are a lot of other good reasons to call for the extinction of NCLB—bureaucratic costs, narrowing the curriculum, stifling innovation, the Constitution—but the inherent incompatibility of political and educational incentives is the biggest of them all.
Thankfully, there is a solution: Universal school choice implemented at the state and local levels. Make parents instead of politicians the locus of education power, and the incentives will align.
Posted November 18, 2008, 9:00am
Good morning and welcome to the NewTalk forum on whether we should "scrap" No Child Left Behind (NCLB), the federal accountability regime for K-12 schools. A past forum on NewTalk considered ways to revise NCLB. The focus of this forum is on more fundamental questions: Was NCLB a good idea at all? What were the basic goals of NCLB and can they be achieved through a federal accountability system? What is the proper and useful role of the federal government in education? If we are going to revise NCLB, we should first address these basic questions. It's not possible to fix something unless we are convinced that it is worthy of being fixed and can be fixed.
Some people in this discussion are likely to suggest that NCLB's flaws are eminently fixable. If so, they should articulate the defense for the federal accountability approach and make the case for how any shortcomings in the current law could be addressed. We should avoid speaking in generalities about proposed reforms, so that the case for those specific revisions can be considered fully.
Other people in this discussion are likely to suggest that NCLB was either flawed in concept or embodies an approach that is too impractical to fix. If so, they should propose specific alternatives. One cannot effectively be against something without being in favor of something else.
Today, let's focus on defining what exactly is so problematic about NCLB, and then we'll move toward solutions and the path forward. I look forward to an interesting and productive exchange over the next few days.