Mr. Kress offers an interesting challenge this week, not only to Race to the Top, but to all of accountability-driven reform.

If I read him correctly, his thesis is simple: without consistent methods of measurement, data-driven reform cannot succeed.

I certainly won’t take issue with that. How can we make sound decisions based on measurements if we’re always changing the yardstick or measuring different things and reporting our results as though they are the same things?

Mr. Kress can always be counted on for a healthy portion of common sense. But reading his post this week, I got the meal and a healthy dessert as well when I thought about the implications of his message.

Mr. Kress’s position is correct, and his questions for the Secretary are news-conference-worthy to be sure because they illuminate something important lurking in the shadows of reform.

For the past ten years we’ve been arguing a lot in this country about carrots versus sticks, loose versus tight, and national versus local when all along the linchpin of data-driven education reform—as Mr. Kress reminds us—is standardization.

Non-standardized data is a non-starter for any statistically-guided venture. The lack of standardization in school data-gathering and analysis we’ve experienced over the last decade has left us uneasy, untrusting, and fundamentally unaware of what and how well kids are learning.

In some ways, just as Campbell’s Law predicts, our perspective on the status of and trends within our education system may be more distorted today than it was prior to testing. Note that I said "may" not "is"; the truth can not be known with certainty, and this is exactly the problem Mr. Kress is driving at. I am not criticising testing. I am criticizing the use of it for dubious reasons. Measurement is essential to understanding and increasing positive change.

Bad measurement isn't useful at all—unless one happens to be in a position to benefit from it politically, financially, or philosophically. Which is exactly why I think we have had so much bad measurement in educaiton over the years. Follow the money, follow the influence, follow the ideology; all tributaries will eventually lead back to a raging river that has little to do with improving education.

Trends on a macro level (like large gains for kids at the bottom of the scale) seem reasonable and significant; as I work in schools, I can see the benefit of NCLB in the classroom and in the data as well. But all too often, “important findings” based on “major studies” using micro-level data that is closer to individual schools, individual teachers, and individual students are found wanting. (The Hoxby study seems to me an example of this.)

It seems logical, then, that if we are going to continue with data-driven education reform, we had better begin by answering Mr. Kress’s questions—and hoping the answers are clear, unambiguous, and supportive of what we think we’ve been doing all along but obviously haven't.

National testing, national curriculum, national instructional models, national teacher preparation, and government rules regarding "generally accepted accounting principles in education statistics" would all be reasonable manifestations of good answers to tough questions about educational measurement. Yet each seems inconceivable (though of the five, testing probably has the best shot, and if we can devise and pass something as complicated and Sarbanes-Oxley, I don't see why we can't do the same for education.)

We seem to be going backwards (or, as Mr. Kress notes about the Kansas waiver, "oddwards") when it comes to consistency in how we count and compare our apples and oranges. Recently, Rick Hess and others have speculated that even the much-ballyhooed and highly popular fabric that was CCSSI is beginning to fray at the edges and is likely to be expressed, at best, as a motley set of tattered remnants.

Perhaps it’s time we agreed that education reform is stuck, that standardization is the sticking place, and that we had better screw our courage to it—now!

Think of a state with NAEP scores that, from the bottom band up, looked like this: 40, 30, 20, 10. Intense standardization competently executed could be the rising tide that lifts all boats. But those in the “South 40” will rise farther, faster, requiring less liquid to buoy their ascent. Beached already, even the slightest amount of water under the hull makes a big differencen and perhaps carries a few schools out to sea.

Over time our mythical state might end up with NAEP scores like 10, 40, 30, 20. This would be an amazing shift (and lift). But such extraordinary improvement seems more characteristic of unusually dramatic polar ice melt than predictable tidal fluctuation.

To accomplish something like this, our mythical state would have to devise ways reaching large numbers of learners with very different needs in highly optimized ways.

Considering this challenge, what immediately pops into my head is that satisfactory standardization requires highly effective differentiation.

At the same time, however, educators—and state education officials, in particular—are easily discouraged from differentiation when confronted with their own choices for standardized curriculum and standardized tests.

What I worry is that increased standardization, incorrectly applied, could actually worsen our ability to meet the needs of kids of different ability levels—especially those in the upper half of the distribution where the gains are hardest to come by, and those formally at the bottom who have floated to the middle but still have far to go.

Helping the newly-arrived "middle class" of education is a fundamentally different challenge than moving kids who have always been average one more rung up the ladder. Goups of kids that have languised generationally at the bottom have more and different issues to overcome.

Even kids in the middle have unique needs—depending on how they got there—that can only be addressed effectively through differentiation. Tracking won't get the job done becuase even when kids are tightly tracked, large ranges of difference still exist, and standardization here favors those whose lives outside of school are more advantaged.

Then there is the inevitable "flattening effect". After large leaps in early years, gains begin to decline until there are no gains at all. This might be due to test familiarity, which loses its benefit after a while. Or it might be a simple phenomenon of behavior economics in the academic economy.

In my work, I have found that schools with test results in the 20s and below are often willing to assume the risks of change simply because they have so little to lose. When they take ownership of a smart change initiative, they often succeed dramatically, ending up somewhere between the 40s and the 60s in just a few years. But now, having reached the mid-point, they are unwilling to implement the even more dramatic change that will be required to improve simply because they’re afraid of falling back to where they were.

Schools in the middle of the pack are often all too happy to standardize on anything and everything they can—especially things that lock down the status quo. It’s easy to rationalize consolidating gains when we’ve forgotten how far behind we were to begin with. For most of us, it is more comfortable to look at how far we’ve come than to face we've yet to go.

As I’ve worked behind the scenes with teachers and administrators across the country, one of the most consistent trends I’ve noticed is a decrease in differentiated instruction in favor of “all teachers being on the same page” (sometimes literally!) and every student doing the same thing, the same way, at the same time, on the same day.

These are not high-pay-off applications of standardization because they standardize the things that have to vary in order to make gains on standardized assessments.

To get all teachers on the same page, and all the kids in sync, the schools I’ve worked with have tended to adopt curricular and instructional approaches based on two factors: that which is least offensive and that to which the least effective team members can easily adapt.

Not only can this cause a regression effect, it also alienates our best teachers, and often discourages them from participating fully. Want more great teachers working in tough schools? Grant them freedom, and provide them with the resources and support they need to teach well without fear of being ostracized simply because they're good at what they do. Let's not forget that we want high quality teachers. And that the term "high-quality" is synonymous with "non-standard". Standardizing output measurse and holding a high expectations makes sense. Standardizing the means of exceeding those expectations does not.

Standardizing curriculum and instruction, in an effort to conform to state standards and tests, often leads to district- and building-level choices that are so conservative they can’t possibly produce the desired result even if they are well implemented. It's the old-fashioned "no pain, no gain" situation. Change is hard, and big change is required here. Yet none of us changes in exactly the same way at exactly the same time. This is just as true for children as it is for adults.

Standardization increases structure. But structue isn't the answer; the right structure is. Schools rushing to standardize many aspects of their operation will inevitably cut off the lifeblood of their pontential—just as lock-step teaching to standardized tests cuts off the potential of so many students and so many talented teachers as well.

It doesn’t have to be this way. Showing schools how to differentiate within a standardized assessment framework is not hard—as long as one has done it, has a model for it, and knows how the model can be replicated. I’ve enjoyed the benefits of this for many years.

Even in districts and schools running scared and making foolish decisions about standardization, I've been able to help teachers differentiate effectively, reach more kids, and get better overall results.

It would be nice if I wasn't always swimming against a rising tide of enforced conformity, but I've always enjoyed working around "school rules" for the purpose of making school better for teachers and kids. And, in any case, when the numbers come out, expressions of individuality should be forgiven of those teachers who score well, even if they don't conform to standardization as their lack of conformity may be the secret of their success.

This is where Mr. Kress's point is so important: If we keep mucking about with measurement, even people who do well will never be recognized appropriately for their success. And our nation will never be able to identify consistent patterns of success for efficient replicaiton.

At this time, however, and in this climate, very few schools—and even educational experts, policy pros, and politicians—see the possibility of "input differntiation" at all. It’s just much easier for everyone if everyone else is doing the same thing. That way, no one notices we’ve all got our heads buried in the sand.

Again, school doesn’t have to be this way. But in most of the schools where I’ve worked, this is exactly how it is—and it's getting worse, not better. Recently, in a presentation to about 100 teachers, I said, "Raise your hand if you have less freedom in your teaching today than you had five years ago." Almost all hands went up. "Now, keep your hand raised if you feel that what you have been required to do has made you a better teacher." Almost all the hands went down.

Trendy structures like Professional Learning Communities often situations like this even worse when innovative and talented teachers begin to lose their best practices to the majority votes of mediocre practitioners, and the sweeping low-bar adoptions district administrators are often attracted to in an effort to exert control in the implementation of standardized approaches. One reason innovative teachers tell me for why they won't work in tough schools is because tough schools often have the toughest structures in which to work. Great teachers want to go where they can teach, where they grow in the practice through change; they don't want to be standardized; that's anathema to their core aspirations as educators.

Macro-level output measurement standardization makes sense. Standardizing ends in thoughtful ways is reasonable; standardizing means in thoughtless ways makes reaching favorable ends impossible. In practice, the biggest gains I see come from differentiation. Having both is possible, and I know that many people like me have developed models that accomplish this. So, if we’re still hungry for change, we can have our cake and eat it, too.

Sadly, I don’t know many principals and teachers who have room for this tasty dessert. Most are simply fed up with the manipulation they’ve endured over the last fifteen years at the hands of education researchers, education officials, special interests, politicians, and powerful philanthropists.

Many of our reform policies aren’t bad at all. But by the time “the system” spits them into our schools, distortion is inevitable, and that's when things become untenable for teachers and kids. We can complain from September to June about the intolerable of the status quo. But we would do well to be mindful that many of our ideas, when implemented in certain ways, may actually make things worse, and that our zeal for standardization may lock in a new status quo of even lower status.

The responsible high-payoff opportunity, then, is to be found in a balanced three-part solution:

1. Tighten standardization of measurement through a high-quality summative assessment framework implemented with consistency on a national level. (The NAEP, SAT, and ACT tests are reasonable examples here.)

2. Loosen the dangerously tight coupling of testing, curriculum, instruction, and school culture to avoid “reforming to the test.” (The more we standardize curriculum and instruction, and spend time reproducing test conditions throughout the school year, the worse things will get for all of us.)

3. Increase differentiation at the building level to assure that all boats are being lifted as high as they can go, not just those that are so far up the beach that any rising tide will float them. (Let’s start REALLY paying attention to ALL CHILDREN whenever we catch ourselves starting any sentence with the words, “All children…”.)

As counter-intuitive as it seems, standardizing differentiation is the most direct and effective solution for education reform. The traditional factory model presupposes that the raw materials we take as inputs can be fashioned into identical outputs—with a high-quality manufacturing process being the “value add”—as long as the assembly line is correctly configured and humming along with Six Sigma precision.

This is basd on several assumptions: that the raw materials of education are identical and “ready for processing” when they enter the system, that the nature and sequence of assembly line stations matches perfectly the optimal construction method for all widgets, and that all widgets can be made in a linear, and thus highly controlled and comfortably predictable, way.

None of these assumptions, however, is true.

The raw materials of education are human beings of near-infinite variation in their intellectual, emotional, physical, and social readiness. Furthermore, the end result we need is not a nation of Stepford Children but same brand of inquisitive and inspired individualists our country has always been famous for.

Rather than standardizing, and writing into law, the 19th century “manufacturing metaphor” of education, a great opportunity exists in reforming schools with a “new product development metaphor” instead.

Each child is an important new product to be released into the world optimized for success, contribution, and well-being. But we will only succeed at standardizing this optimal output if we attend thoughtfully to the individual, non-standard needs of the non-standard children and adults who comprise our schools.

Having the advantage of writing this after the CR deal has been made, it appears that Race to the Top will get $700 million but still work through states. So no regional or direct-to-district grants, at least from what I've been able to divine.

Unfortunately, the root issue here isn't Race to the Top -- it's federal involvement in education at all. As long as Washington spends money on education it will call the shots, whether that's through competitions like Race to the Top, or rules attached to IDEA, or Title I, or student-aid programs. Washington will be the puppet master, and it will waste taxpayer money on programs that do not work.

We know this well from the four, wallet-emptying decades in which we've had major federal involvement in education. Yet we are still having the same, misguided arguments: Not whether Washington should be setting education policy at all, but what sort of policy it should set. And the two sides continue to come down, basically, to whether Washington should spend lots of money with copious "accountability," or spend oodles without as many rules.

Both are wrong. The Founders purposely gave the federal government only specific, enumerated powers, and none of them include the authority to make education policy. And as I have laid out many times, the "general welfare" clause, the "necessary and proper" clause, the taxation clause -- none of these gives Washington the authority to do anything in education. Only the 14th Amendment gives it a role in education -- to prohibit unequal provision of education by states and districts -- as well as Article I, Section 8, which gives the Feds exclusive jurisdiction over the District of Columbia.

Of course, the Founders didn't give Washington only very limited, specific powers for no reason. They did so because they knew that the strong tendency is for government to be captured by small groups and used not for the public good, but the good of those who wield governmental power. Today, that means politicians often using "the children" as pawns to win political games -- spending money to show how much they "care" -- while special interests employed by our schools lobby and politick for ever-more dough, weak accountability, and hamstrung competition. And some wonder why federal spending on education has skyrocketed while outcomes have stunk.

We can keep on bickering about Race to the Top and other proposals to twist a few screws on the dilapidated federal machine, or we can get Washington out of education altogether. Both the Constitution and educational success demand we do the latter.

I noticed that the Department recently gave a waiver to a school district in Kansas to develop and use a different assessment system than the one used around the state.

This waiver has caused a great deal of consternation among education reformers.

For one thing, if districts can use uncommon summative assessments, won't this harm, rather than help, in the implementation of common core standards? At least today states promote common implementation of standards, partly through statewide systems of summative assessments.

What is the likelihood of having fair and equal accountability in states if districts can have differing summative assessments?

What's the likelihood of having fair and comparable measures of growth for evaluating teachers across a state if its districts use differing yardsticks?

And what's the likelihood of educating disadvantaged students to the same high standards in a state if districts can have different metrics and markers?

The reason I ask these questions in response to this week's question is that I wonder what the Department's ground rules will be for local districts to compete for Race to the Top funds.

Basically, my main question is: will districts seeking to implement policies and practices out of sync with those in effect statewide be eligible for funding?

Or, put another way: is the Administration still committed to implementation of the common core standards in the states?

It would be splendid if Congress killed the funding for Race to the Top. What a monumental waste of money to promote unproven and failed "reforms." Why should the federal government push districts to add more charter schools when most of them don't get better results than regular public schools? Let Wall Street hedge fund managers pick up the tab for their hobby.

Why should the federal government prod districts to use unproven methods to evaluate teacher effectiveness? Why should it waste more money on merit pay, which consistently fails to produce results?

Future historians will look on Race to the Top as a foolish scheme cooked up by elites and foisted on the nation's public schools. It should be stopped now, before it does more damage.