Manhattan Institute for Policy Research.
Subscribe   Subscribe   MI on Facebook Find us on Twitter Find us on Instagram      

Civic Bulletin
No. 49 September 2007

The Science of Reading Instruction and No Child Left Behind

This bulletin is adapted from a transcript of a Manhattan Institute forum held in New York City on May 22, 2007.

Reading First, part of the 2001 No Child Left Behind Act, seeks to lift reading achievement by encouraging the use of reading programs that have been scientifically proven to work. Reading First, which accounts for just 2 percent of federal education spending, is helping many districts to achieve promising results – including Richmond, Virginia, where test scores have risen dramatically. But negative publicity stemming from recent reports by the Department of Education’s Office of Inspector General could put Reading First under a cloud when NCLB comes up for Congressional reauthorization. That would be tragic for millions of children at risk for reading failure. To see clearly what’s at stake in the reauthorization battle, the Manhattan Institute gathered a panel of experts on the science of reading instruction. As our luncheon speaker, U.S. Secretary of Education Margaret Spellings then discussed the prospects for reauthorization of No Child Left Behind.

* * *

Panel Discussion: Reading First and Reading science

G. Reid Lyon, Former Chief, Child Development and Behavior Branch,
National Institute of Child Health and Human Development, National Institutes of Health

Diane Ravitch, Education Historian; Research Professor, New York University; and author of The Language Police: How Pressure Groups Restrict What Students Learn (Alfred Knopf, 2003) among other books

Rick Nelson, Former President, Fairfax County Federation of Teachers, Fairfax County Public Schools, Fairfax, Virginia
Maria Casby Allen, Parent Activist, Fairfax County Public Schools, Fairfax, Virginia

Moderator: Sol Stern, Senior Fellow, Manhattan Institute

Luncheon Address

The Honorable Margaret Spellings, Secretary, U.S. Department of Education

* * *

I’m a Manhattan Institute senior fellow and a contributing editor of City Journal. On behalf of the Institute, I want to welcome you to this conference on reading science, the Reading First program, and No Child Left Behind.

This morning’s panel is really about an American tragedy. After a century and a half of universal public education, and despite the highest per-pupil expenditure on elementary and secondary education in the world, 40 percent of U.S. fourth-graders can’t read proficiently. That’s according to the gold-standard NAEP (National Assessment of Educational Progress) tests. For minority students in inner-city schools, the reading failure rate is a catastrophic 65 percent. The consequences of this education failure are devastating. Children who don’t read by fourth grade almost always fall behind in other subjects, often end up in costly special-education programs, and are more likely to drop out of school.

But this is an entirely self-inflicted wound. American scientists have figured out an answer to the reading-failure problem. For the past several decades, the National Institute of Child Health and Human Development—a wing of the National Institutes of Health—has, under the direction of our first panelist, Reid Lyon, sponsored reading research by scientists in the field of cognitive neuroscience, pediatrics, and educational psychology. We now have hundreds of peer-reviewed studies that describe not just how children learn to read but also why so many fall behind, and how schools and teachers can keep this from happening.

But here’s the scandal: in the education schools that train our future teachers, science is disdained. What’s worse, education professors have convinced many school districts to choose reading programs for the classroom that satisfy the professor’s philosophical beliefs about children but have no scientific support. When this happened in California in the 1980s, reading scores plummeted to the bottom in the nation.

In New York City, our schools chancellor dismissed the federal Reading First program, which is based on this science. He took the federal money—$2 million—but still asked, “Where’s the science?” Perhaps the panel we have assembled today can help the chancellor and his staff find the science. Doing so would point the way to better reading scores. Coincidentally, the state is releasing its 2007 report on reading scores in a couple of hours. I’ve been told that the report shows that New York City could use a lot of help.

Our panelists are, each in his or her own way, heroes of American education. Our first speaker, Reid Lyon, was the chief reading scientist at the National Institutes of Health for twenty years. Reid will give you a glimpse into the power of the reading science that has been incorporated into the Reading First program, and he will tell you how it can improve classroom instruction and narrow the shameful racial gap in academic performance.

Diane Ravitch is our nation’s leading historian of education and is the author of numerous highly acclaimed books, including the definitive history of the New York City schools. She’s an eloquent champion of the idea that true education reform begins in the classroom with a rich curriculum, high standards, and instructional approaches that are backed by evidence. She will provide you with historical background on the reading wars and explain why most of the education establishment still hasn’t come to terms with the scientific revolution.

Our final speakers are Rick Nelson, a former president of the Fairfax County Federation of Teachers in Virginia; and Maria Casby Allen, a parent activist from Fairfax. They did exactly what we hope all parents and teachers will do as a part of our democratic public school system: they educated themselves about the science of reading, looked at the data for their own district schools, and lobbied district authorities to do the right thing for the children, based on the evidence.

I am currently working in Dallas, Texas, at Higher Ed Holdings. I have left the government. One thing I’ve learned is that Washington, D.C., is 30 square miles of unreality surrounded by reality. And it’s nice to be in some of that reality. What I’m doing in Dallas is building a school called the American College of Education, so that I can put my money where my mouth has been. We’re trying to develop colleges of education that do straightforward things. We try to help teachers understand what the best scientifically based instruction is as well as what the most current and accurate content is. We teach the teachers in the public schools so that they can immediately apply the concepts we’re teaching them. And we monitor teachers’ progress just as we monitor the progress of youngsters. The data will tell us whether we’re doing it right or not. And if we’re not doing it right, we need to fix it.

Today, I’m going to talk about the research we’ve done over many, many years. But I want to read something to you first.

“The history of the profession has never been a particularly attractive subject … and one reason for this is that it is so deplorable of a story. For century after century, the profession got along by sheer guesswork and the crudest sort of empiricism. It is hard to conceive of a less scientific enterprise among human endeavors. Virtually anything that could be thought of for treatment was tried out at one time or another. It was the most frivolous and irresponsible kind of human experimentation based on nothing but trial and error and resulting in precisely that sequence.”

That passage was originally presented by Dr. Lewis Thomas, president of Sloan-Kettering. He was talking about the medical profession—a profession that took over a century to move out of its treatment armamentarium: leeches, apricot pits, cutting, slicing, bleeding. It wasn’t until the public stood up and said, “You are killing people” that the profession began to monitor itself and changes were made. Indeed, when the polio vaccine was first developed, it wasn’t seen as productive or effective because doctors didn’t do the right science: they didn’t do randomized clinical trials. Once they did, they saved a lot of children’s lives.

Over my career, I’ve looked at some of the gaps that we have in our educational practice, and I’ve always tried to use common sense foremost. Why, with children in front of us, would we ever present to them or provide them with something that might not work, or something for which we might not understand whether it works? It has never made sense to me. Yet, as Sol pointed out, we continue to stretch kids through a wide variety of malpractice. And they continue to fail.

I’ll tell you later how all of that prompted Reading First. I don’t want to bore you with certain parts of the research. But when I talk about most scientific research, all I’m talking about is common sense. If you want to understand something, you have to ask a good question. And at the NIH, I asked the same three questions for three decades. First, how do kids learn to read—that is, what skills and abilities do they need? Second, what goes wrong when kids don’t learn—what’s getting in the way? And third, what do you do about it?

Three questions, that’s it. Year after year, we tried to map out how kids from every walk of life learn to read. Does the learning process get messed up for some kids because of genetics and neurobiology? Yes, for a very small number of kids. Does it get messed up because teachers aren’t prepared? Yes. The biggest impediment to kids’ learning to read is not biological or genetic: it’s instructional. Instructional casualties account for the majority of that 50–60 percent of our poor kids who can’t read. It does not have to be that way. It’s as simple as that.

But we just sit back and watch as enormous travesties are placed on our nation’s children. And we don’t get it. It’s more complicated than a program or a method. But nobody wants to deal with complexity, either. They have a hard time, particularly in education, recognizing common sense. It’s either/or: everything is this or that. Nothing in life that I know of is either this or that.

By 1990—maybe even by 1985—we had figured out what goes into reading, how the process works. We’d also figured out what gets in the way when the process doesn’t work. What we hadn’t figured out was what to do about it. And to figure that out, I asked: For which kids are certain instructional strategies or approaches most beneficial, at which stages of development, in which settings, for how long, and by which teachers? I definitely did not ask: Does whole language work better than phonics? That’s a dumb dichotomy. That’s political. Science asks: For which kids are which instructional strategies most beneficial, at which developmental phases, in what classroom, and by what teachers?

We’ve pretty much answered those questions. But will anybody use our answers? No. The press can’t understand these issues. Every time they talk to me, they ask, “Does phonics work better than whole language? In early childhood development, is it social, emotional, or cognitive? If you’re a researcher, is it quantitative versus qualitative research?” The dichotomization of complex concepts is a proxy for IQ; we seem to be operating at an IQ of 75 or lower when we talk about reading. But I know that we have to do research to tell us what works, why it works, when it doesn’t work, and under which conditions it works. Businesspeople understand this sort of thing. If there are any physicians in this room, they know about dosage effects and combinations of medications. We have to do that a lot in education. That is why teachers have to be prepared with the best information, because they will know that not every program is equally beneficial for each child. If a kid doesn’t get it, the teachers will have to modify instruction, even with the most well-supported, strongly based programs.

There is no script that teachers can follow for every kid. But they can’t be teaching kids without some facts in front of them. Most important: reading is complex. You’ve heard it before. Reading requires the development of a sound structure, called phonemic awareness. It requires the ability to place sound on top of letters—the F-word in our society, phonics. It requires the speedy application of those phonemes or sounds onto the letters; speedy so that kids can read words and sentences quickly, without getting bogged down and bored. But you can’t read and understand anything unless you have the vocabulary to bank new against known. I can read an astrophysics text all day, quickly and using good decoding strategies, but it’s going to bank up against a limited vocabulary. And I won’t understand it. You’ve got to be able to comprehend.

To make a long story short, learning how to read is a multidimensional, complex process that requires the development and integration of many equally complex subparts. Our teachers usually don’t get that, because they teach what they have been taught, as Sol pointed out.

What gets in the way when kids have trouble learning to read? Usually, there’s a missing link, or links, among the components I just described. Some kids don’t have good phonemic awareness. Some kids don’t have good vocabulary. Some kids don’t read quickly. Some kids suffer from all of the above, and they bomb. Do they have to? Absolutely not.

In 1996, as I begin to tell this story publicly, I got a call from Bill Goodling, who was the chairman of the Education and Workforce Committee in the House of Representatives. He said, “I hear you guys at NIH are doing some science on reading,” even though it was his committee that had given us a lot of money for it. And he said, “Look, we are about to fund a program called America Reads. It’s a program where we are going to pay for people to read to children. And we are hoping that it will really help the kids out.” I said to him, “Reading to children is absolutely necessary. But to spend that amount of money on having grandparents and adults read to kids who don’t often get read to, while laudable, will not teach them to read efficiently.”

Kids don’t pick up reading naturally. It’s like golf or piano playing. The whole idea that reading is natural has been disproved over and over again. If people hang onto it, they do so at kids’ risk. Be that as it may, Goodling asked me to take a look at the data. He was a former high school principal, and his wife was a kindergarten teacher. He understood education. When I finished talking to him in detail over a day and a half, he said, “You mean to tell me that we’ve got 60 percent of poor kids not reading? And the studies that you have in Washington, Houston, and Tallahassee in schools with real teachers are showing that you can reduce that to 10 percent? What do we do about that?” A fellow I worked with, Bob Sweet, said that we should enact some legislation, and what came out of that was the Reading Excellence Act—the first time you ever saw scientifically based reading research (SSBR) placed into legislation.

It didn’t work. There was no accountability to it. People wrote grants and said that they would adhere to the criteria for SBRR, but then they’d get the money and do exactly what they had always done. When we moved down the road, the science kept accumulating; we knew how reading developed, why it went wrong, and what you could do about it. So along came Reading First, which contained one crucial difference. With Reading First, we said: federal money will be provided for programs that have been shown to be effective, with well-defined groups of children, and under well-defined circumstances and conditions. That was the general message.

That would have allowed about three programs in the entire country to receive funds—that’s it. I’m a scientist, and I wrote that language in there because I knew, even from a policy point of view, that if we weren’t specific, people would take advantage of the system. We had to be specific, or otherwise everybody would say, “Well, I’m research-based,” and they would change the language in their materials to say that. That’s just human nature. The lobbying efforts were tremendous, and in the end Congress changed the language and said that federal money would be provided for programs based on SBRR, not merely for effective programs. People tried to game the system all over the place.

Chris Doherty, the Reading First program director in the Department of Education, was one of my heroes. He said, “This state has said that it will adhere to programs based on SBRR.” Yet it gets the money and pays for programs that are like apricot pits and leeches. But we wrote in the law, thank goodness, that they couldn’t do that, and Bill Goodling told the states that he would need to pull their funding if they didn’t get it right.

The policymakers on the Hill forgot about something. They passed Reading First, but they didn’t think about how it might bump up against local control. What people are saying now is that you cannot mess with how local districts use that money. But in Reading First, we said that you cannot use federal money for programs not based upon SBRR. People started saying that they would adhere to SBRR, but then they didn’t because of local control. Does your tax money go to people who say they will do one thing, only to do another at the detriment of the kids because it’s a local control issue? I’m afraid so. These are self-inflicted wounds.

Nevertheless, with all of the policy developments that we’re talking about, it’s the first time that I’ve ever seen research driving policy in Reading First and No Child Left Behind. These developments weren’t driven by the typical kinds of policy engines. They certainly weren’t driven by consensus conferences or policy workshops. For example, there was the National Reading Council report on beginning reading—a great report. But it was a consensus report, and it went nowhere. It was not read in terms of changing the legislation.

What had to happen was for someone with decision-making power—like a Bill Goodling, Thad Cochran, Ann Northup, or Ted Kennedy—to ask, “You mean to tell me there’s this gap between all these kids learning how to read, and that we know how to reduce that rate all the way down to 10 percent or less in the worst schools with the kids with the most difficulties? You mean to tell me we can do that and we’re not doing it? And you’re studying these kids in real schools? This isn’t laboratory stuff?” Yes, sir—1,000 schools, 2,500 teachers, 44,000 kids, all studied for at least five years longitudinally and an average of nine years longitudinally. The kids are now thirty-something; they started when they were five. All those kids contributed to what we know, but their legacy has been lost on the nonsense perpetuated by this dichotomization that I talked about earlier and by the failure for adults to use common sense.

If there’s anything I can leave you with, it’s that we do have the science that can reduce these numbers dramatically if we do it right. A word of caution: no one program, as I mentioned already, is equally beneficial for all kids. But the process does have to be direct and systematic. It cannot be happenstance. It cannot be everything-and-anything-goes. Kids do not learn to read naturally. When you think that they do, they come from middle- or upper-middle-class homes where the kids have been read to since in utero. They come into school with vocabulary, sound structure, and everything else. Poor kids don’t typically have that advantage, and people expect them to read naturally by looking at good literature, which might help kids who already have the foundational skills. Poor kids don’t know what’s going on. It’s like anything else; they’ve got to be taught directly and systematically. We cannot overturn that scientifically. We cannot falsify that hypothesis. You just can’t let them flounder.

People say, “Well, if you teach them too directly, they’ll never love reading.” I’ve never met a kid who loves something he cannot do. You have to have great programs and comprehensive, direct, and systematic instruction. But you also have to have a great teacher. And there isn’t going to be a chance for those things to work if you don’t have good building-level leadership providing a context where they have the time and can collaborate and talk with one another. That way, they can solve problems using a common language, based upon scientific knowledge of how reading develops, what goes wrong when it doesn’t, and what you can do about it.

As the designated historian, I’m going to give you the background to what Reid Lyon was talking about. Reading wars are nothing new. Education experts have been arguing about how to teach reading for nearly 200 years.

When we first began as colonies in this country, the popular method was called the alphabet method. There were hornbooks that listed the alphabet. People used to memorize those letters and somehow figure out how to read by memorizing the letters.

The alphabet method was succeeded by the phonetic method, whereby children learned the sounds of letters and combinations of letters. This was the method used in Noah Webster’s famous Blue Back Speller, which taught millions of people to read. The Webster spellers were succeeded by the even more famous McGuffey Readers, which sold more than 120 million copies from the time they were initiated in 1836 until about 1920. At one point in our nineteenth-century history, the McGuffey Readers were used in half the classrooms of the United States. By the way, they have wonderful literature—it wasn’t just phonics. People look at them and say, “Oh, this was just drudgery.” But no: they had great literature for kids.

The third method that became popular was the word method, in which children learned to recognize words and to read sentences without necessarily learning the alphabet or phonics. They memorized the look of the word. One of the early champions of the word method was Horace Mann, who is better known as the father of the American public school system. Mann believed that learning the alphabet was pure drudgery and torture. He spoke of the alphabet as the twenty-six bloodless, skeleton-like figures that torture American children.

When I was in graduate school, I learned that Horace Mann had a famous debate with the Boston schoolmasters, but nobody ever said what the debate was about. So I dug up some information about these old debates. The Boston schoolmasters, angry with Mann, said, “When children are taught to read without learning those letters, they don’t ever learn to read very well. And they spell very poorly. And by the way, Mr. Mann, the words that you love so much are made up of those twenty-six bloodless, skeleton-like figures.”

A later variation on the word method was what we now know as whole language, in which children learn to read because they’re motivated to do so by their interest, with no analysis of letters or sounds. Phonetic methods had dominated the teaching of reading in the nineteenth century, but with the rise of progressive education, a new philosophy came along that was much more in tune with what Horace Mann had been arguing. The progressive philosophy was articulated by men such as Francis Parker in Massachusetts, G. Stanley Hall at Clark University in Massachusetts, and John Dewey.

One hallmark of this philosophy was that children should not be taught to read until they’re at least eight years old. Early reading is bad for the child’s nervous system, this philosophy claimed, and teaching the alphabet discourages children from learning to read because it lowers their motivation and their interest. The most important book about teaching reading in the early twentieth century was Edmund Burke Huey’s The Psychology and Pedagogy of Reading. It was a must-read in every college of education in the country. Huey echoed Dewey, Parker, and Hall in saying that children should never learn to read until they’re at least eight years old, preferably even later. They should be taught to read at home, not at school. Reading should be taught like pictographs, with no phonetic analysis of any kind. And the best method for teaching reading is no method at all. The best readers, Huey said, were compilations—not textbooks, but compilations of students’ own work.

Researchers in the 1920s and 1930s endorsed all these themes. Among the other findings that resulted from the research of that era was that reading out loud is harmful to children; it slows them down. They would learn to get information through the ear, not the eye, and this would be very bad. According to this research, they should read silently. Putting the emphasis on silent reading obviously took the emphasis away from any kind of phonetics, because now the sounds of letters didn’t matter at all.

Another part of the research synthesis of that era was that any linguistic analysis of sounds or letters was a bad idea. It was unnecessary and not helpful in learning to read. The best way to learn to read was to form a mental image of the word instead of recognizing letters or sounds.

The Dick and Jane reading books, which first appeared in 1930, were whole-word look-say books. They were intended to be the anti-phonetic books. They had a controlled vocabulary of short, simple words that were repeated again and again. They dominated the world of reading instruction, along with other textbooks of that era—which were modeled on the Dick and Jane approach—until the mid-1950s, when Rudolph Flesch blasted this approach in his best-selling Why Johnny Can’t Read. That book was on the best-seller list nationally for over thirty weeks, which, for a book about education, is very impressive. Flesch argued that there was a national crisis in literacy because of the systematic neglect of phonics. He said that children were memorizing one-syllable words but that they were unable to read unfamiliar words.

In the early 1960s, with the great debate that was going on across the country about phonics, about Dick and Jane, and about bad readers and good readers, the Carnegie Corporation of New York commissioned Harvard University researcher Jeanne Chall to review the research about reading. A former kindergarten teacher in the New York City public schools, Chall had gone on to become the nation’s preeminent reading researcher. The book that resulted from this Carnegie Corporation study was her Learning to Read: The Great Debate. It appeared in 1967, and it is still, from my point of view, the definitive book on the reading debate. Chall had never heard of Reid Lyon or the research that he has done, but her book still provides the best overview of the research and the arguments. Her book should have ended the great debate.

Chall said that the debate about reading methods sounded more like religion and politics than science and learning. She said that there was not just one successful way to teach beginning readers. No method had completely eliminated the problem of reading failure. Some methods were better than others, but none is a panacea. She found it difficult to compare methods of teaching reading, because each approach contained some elements of the other. “Every school that introduces a new method still retains a good deal of the old one,” she wrote. Teachers tended to stick with whatever method they knew, regardless of what the administrators said that they were doing.

Chall said that there are two primary approaches to teaching reading: one stresses the importance of breaking the code of language; and the other stresses the meaning of language. Phonics programs had a code emphasis, and look-say programs had a meaning emphasis. The research, Chall said, unequivocally supported the use of a code emphasis for beginning readers—and she stressed “beginning readers.”

She found that the first step in learning to read in one’s native language is essentially learning a printed code for the speech we possess. The code emphasis was especially important for children of lower socioeconomic status, she said, because they were not likely to live in homes surrounded with books or with adults who could help them learn to read. Knowing the names of the letters and the sounds of the letters before learning to read, Chall said, helps children in the beginning stages regardless of which method is used. She concluded that for a beginning reader, knowledge of letters and sounds had even more influence on their reading achievement than the child’s tested IQ did.

Chall warned that the schools should not go overboard in teaching phonics. She warned that if phonics was overemphasized to the exclusion of comprehension, there would be a reaction and the pendulum would swing back to the whole-word method.

Jeanne Chall, a very wise woman, was exactly right. By 1980, the whole-language movement had emerged, and it swept the field for at least the following generation. The leaders of the whole-language movement insisted that children should read for meaning and pleasure and should not study the mechanics of language. They ridiculed phonics and any other kind of linguistic analysis. They insisted that children would learn to read without any instruction in the names or the sounds of the letters. Because whole-language theory dovetailed with progressive education theory, stressing the joy of learning as opposed to the drudgery of instruction, it proved immensely popular in the schools of education. State departments of education, too, championed it in their credentialing policies. Textbook publishers rushed to become part of this exciting new movement.

The whole-language movement, however, ran into a wall when California, the state that had most enthusiastically embraced it and had really launched it, received a terrible ranking on the NAEP reading test in 1994. California scored almost at the bottom nationally, ahead only of Louisiana and Guam. Students in every racial and ethnic group—even children of college graduates—did very poorly. State leaders immediately concluded that whole language was responsible for California’s terrible showing.

In 1995, the California state legislature mandated phonics instruction, and the state board of education adopted a new phonics-based reading curriculum. Nationally, several events that followed suggested that there might be an end to the reading wars. In 1997, a report sponsored by the National Research Council, called “Preventing Reading Difficulties in Young Children,” emphasized the importance of phonics in teaching beginning readers, echoing Chall’s finding. Additionally, the report of the congressionally mandated National Reading Panel in 2000 confirmed the importance of phonemic awareness for beginning readers.

Congress included the Reading First program as part of the No Child Left Behind legislation of 2002. The Reading First program—and many people are surprised to hear this—is not a mandate on districts; it is not a mandate on anyone. It’s a competitive grant program. No state or district is required by No Child Left Behind to use any particular reading method. If they ask for Reading First funding, they’re asking for money to introduce programs of reading instruction that have some evidentiary basis.

Reid Lyon spoke about the Reading First program, and the other speakers will do so as well. What I want to emphasize is what Jeanne Chall wrote in her classic book Learning to Read: The Great Debate: there is no single best method of teaching reading. There is no surefire, failsafe method.

Efforts to dedicate the schools solely to skill-building are narrow and unrewarding. And efforts to eliminate phonetic instruction in the alphabet are misguided. What we should have learned from the history of the reading wars is that reading is not natural, and it’s not easy. If it were natural and easy, we wouldn’t need teachers or schools. We wouldn’t have any illiteracy, and we wouldn’t have any problem readers. The literacy rate would be equally high around the nation and around the globe, regardless of the quality of teachers or schools, if there were any, which there wouldn’t be.

What we do know is that good teachers and good teaching make a huge difference. Children need teachers who know how to teach them decoding skills and how to teach them good literature. Most children will not learn to read unless they are taught to read. The only way ever to break free of the pendulum that moves from one extreme to another is to insist upon solid evidence before adopting any reading program on a broad scale.

It is imperative for educators and parents to ask: “Has this program been tried before? Where was it tried? For how many years, with what results, and with what kinds of students? Are the students similar to ours? Is there a research base behind the program?” Unless there is a consistent and predictable demand for evidence, we shall continue to be players in a drama that has gone on for far too long, arguing about methods based on ideology and sentiment, instead of insisting on methods that are known to be successful.

I’m a parent of two boys who attend public school in Fairfax County, Virginia. Fairfax County is a very large, mostly suburban school division. It’s the twelfth-largest school division in the country, with more than 140,000 students. It has an excellent reputation as an outstanding school system with high achievement. Children in Fairfax County are generally fairly easy to teach. The median household income in Fairfax County is about $90,000 per year.

At a school board meeting in October 2004, almost three years ago, I stood in front of the school board with this graph (see Appendix, Fig. A), which compares the level of achievement of black children in Fairfax with that of black children in Richmond, Virginia. Both school districts have about the same number of black children—between 15,000 and 20,000—although in Richmond, black children represent 90 percent of the schoolchildren in that city. In Fairfax, they represent 10 percent.

Across the chart, you’ll notice that the data are pretty consistent: reading, math, science, and social studies in third grade, and the same in fifth grade—all the standardized tests they take in Virginia.

What may surprise you, as I told the school board and the administrators at the time, is that the blue bars represent the city of Richmond, and the red bars represent Fairfax County. On every state test given to elementary school students in Virginia that spring, black children in Richmond significantly outperformed black children in Fairfax County.

The data, I told the school board, get even more interesting. When one adds in the averages for the state, you see the same sorts of results (see Appendix, Fig. B). The state is the black line in the middle, and Fairfax County is the red line on the bottom. Richmond is on the top, again, in every single subject that’s tested in elementary school in third and fifth grade.
So black children in Richmond outperformed not only Fairfax children but their counterparts across the state as well, while children who attend our highly regarded schools in Fairfax are seriously lagging behind not only Richmond but the whole state. The data came as a surprise to school administrators, who were convinced that Fairfax County, because of its high reputation, simply was untouchable.

However, I saw things a bit differently. In my town, Reston, which is a planned community, children from subsidized low-income housing and children from million-dollar homes walk the same pathways to school. But it seemed that in our schools, only some of the children were learning. Children who came to school without the basic skills in reading, writing, and math left school pretty much the same way, or at least many of them did. The home, the parents, and poverty all took the blame. But as a frequent school volunteer, I observed that the blame lay elsewhere. The children were simply not being taught.

Education would determine the fate of these children. So I hoped that somebody somewhere was doing something different, and I hoped that Fairfax County could learn from whoever that was. If not, the future for these kids and for Reston looked bleak.

The first data that I came across, as I started to look for something different, came from George Mason Elementary School in Richmond—or at least, those were the first data that surprised me. George Mason Elementary School is 99.6 percent black, and it’s also in a high-poverty section of Richmond. I discovered that in 2004, despite this poverty, their scores were as high as the scores in the wealthiest part of Fairfax, which was 99 percent white.

Until two years previously, academic achievement at George Mason had been exactly what one would expect for such a school: rock-bottom, about 30 percent pass rates for both third grade and fifth grade in every subject. And then, all of a sudden, something happened. Scores shot up in third grade to between 90 and 100 percent; and in fifth grade, the same thing. Within two years, they were at the top (see Appendix, Fig. C).

I wanted to know what it was they were doing there, so I telephoned the principal to ask him some questions. With great enthusiasm, he and several other Richmond principals talked about their successes, which, it turned out, came from the same things I would later learn are at the center of No Child Left Behind. George Mason had to change, the principal explained. All the children were failing. Yet there was nothing radical in what they were doing, he told me. They were simply making every moment of the school day count by teaching in ways that were proven to get results. The hardest part, he said, was changing the mind-set of teachers and staff. But once that was done, everything else was just plain common sense.

What had to come first, he said, was that they had to stop blaming others and making excuses for failure; instead, they had to take responsibility themselves for teaching their students. He said, “We have no expectations of the home. We understand that we can’t count on anyone else to teach our children. It’s our job to do it here in the school. And it’s not easy. So every minute of our school day is precious.”

“You’re out of excuses,” said another principal. “And you know what you have to accomplish in so little time, so you look carefully at how you’re teaching. You make sure that what you’re doing and that the way you’re teaching really work so that you’re getting the most out of every precious minute.”

How did Fairfax respond to the data? Initially, the Fairfax administration said that Richmond was known to inflate its scores, so it was probably cheating. Others dismissed the success in Richmond—and still do as, a matter of fact—by saying that Richmond was simply teaching to the test, whatever exactly that means. The data that I presented did not create a sense of urgency; in fact, the data seemed too quickly forgotten. So, not much later, I returned to share more bad news with the Fairfax County School Board: other school districts in Virginia that were doing good things were also starting to get good results.

I compared scores of the ten school divisions in Virginia, which were urban as well as suburban, with the largest black student populations (see Appendix, Fig. D). Fairfax County came out rock-bottom, ten out of ten—or at least, tied for the bottom position—on every single state test taken in elementary school, right across the board. Richmond schools were consistently some of the top performers.
Where are things now, a few years later? Frankly, not much has changed. The staff in Fairfax County went to observe Richmond schools, but the school board never asked them to report on what they found. If they thought that Richmond had anything to offer, they were careful not to say so publicly. Fairfax felt some heat when the Washington Post picked up the story on its front page, so they intensified remedial efforts for those at the greatest risk of failing, and they picked up a few percentage points.

Six months ago, the instructional staff in their annual report to the school board gave no hint of any sort of change in what they were doing, even though they acknowledged that the reading-ability gap between children who are and who are not living in poverty in Fairfax County had actually widened slightly. They seemed to be saying in this report that what they were doing wasn’t working—but that they would continue to do it. As usual, the term “phonics” or “phonemic awareness” never came up. You’re not supposed to talk about things like that in Fairfax County. But a great deal of time was devoted to discussions with elaborate charts showing the effects of poverty and home life and the education of the parents and everything else on reading scores, although the staff said many times that they weren’t bringing up these things as an excuse. Why, then, were they bringing them up? They can’t change poverty in Fairfax County.

A comment by the head of instruction in Fairfax County to the Washington Post was quite revealing. She said that Richmond’s progress had little relevance for Fairfax County because in Fairfax County, the vast majority of students were passing. School officials, she said, didn’t want to give up the creativity that comes with current teaching methods. She feared that many in our community would say, “This is not what I want for my child.”

The majority of children in Fairfax County are wealthy and white. They can get by with poor instruction. And majority rules, I guess. The disadvantaged children in Fairfax County—well, I guess they’re simply out of luck.

From 1972 to 2004, I worked as a teacher and as an elected union representative in Fairfax County. For the last ten years of that career, I tried, without much success, to get the school system to change its reading and math policies. After I went off to fight a different battle, Maria came along and did her amazing studies, which took hours of work eyeballing data presented by the state in a most unfriendly fashion. Over the years, I collected some data that may help to explain Maria’s findings. I suspect that what is true in Virginia is true in many other states.

Virginia’s Reading First plan invited the state’s 200 lowest Title I schools to apply competitively for seventy-two grants. Those grants totaled $1 million per school over a five-year period—serious money. The schools were given a choice of about five or six curricula to choose from, all of which were science-based. But they weren’t too limited in their choices.

Despite the flexibility and the money, many schools in the state did not apply for the lottery, in which the chance of winning $1 million was one in three. The seventy-two schools in Virginia represent only 10 percent of our Title I schools, and they represent only 3 percent of the schools in the state, so Reading First is really quite limited. It’s a demonstration program, and, frankly, not many teachers are directly seeing the benefits of Reading First unless they’re at those schools. The teachers at those schools love it, though.

Most Virginia schools, like schools in most of the country, got their Reading First money at the school level after August 2003. And one thing you don’t do, no matter how strong a principal you are, is to ask your reading teachers and your classroom teachers to change the reading program after the year has started. So most schools used the 2003–04 school year for training and for acquisition of materials, and they started Reading First in the classrooms in September 2004. Because of that, we only have two years of data on how Reading First is doing in Virginia, and those two years represent children who spent kindergarten and first grade in the lowest-scoring Title I schools in the state with programs that weren’t working, then spent second and third grade in Reading First. That’s not a fair test of Reading First, but nevertheless, the initial data are encouraging.

These data are compiled by Chris Braunlich at the Thomas Jefferson Institute in Virginia. I want you to construct this same chart for your state and your local districts and then add the scores that we get this spring and try to use the data to drive reform that helps children. This is not difficult to do. All these data are up on the Web in every state, thanks to No Child Left Behind (see Appendix, Fig. E).

Virginia does have one school district with five years of data on Reading First and science-based instruction, and that’s Richmond. What Richmond did was to start Reading First three years before everybody else. I should say that Richmond has 25,000 students in fifty-one schools—a big district for the South—90 percent of whom are African-American, and 75 percent of whom attend Title I schools. In 2001, Richmond got a new superintendent, Deborah Jewell-Sherman. Her first initiative required that every school adopt some form of science-based reading instruction over a two-year period. Most, but not all, of them chose Voyager, a specific core reading and professional development program. Unlike California, Richmond didn’t do just phonics. Richmond incorporated all five elements of the science-based reading research, which Dr. Lyon talks about.

What kind of results did they get? Richmond went from the bottom 5 percent in Virginia to the top 40 percent in four years. Does anybody else know of an urban school district, 75 percent Title I, 90 percent black, that has reading scores in the top 40 percent of its state for third grade or above?

I think Richmond’s success is unprecedented, and it’s one demonstration of what science-based instruction can do. This was an average for forty schools across the district. Some individual Richmond schools, as Maria noted, did better. Some did worse. Some of the highest-poverty schools in Richmond scored higher than any school in Fairfax County in third- and fifth-grade reading. Those schools attributed their success to three things: science-based reading research; training their teachers on how to use science-based reading research; and training their teachers to test the students frequently to make sure that they were keeping up, and to adjust instruction accordingly.

In Fairfax County, they were doing something very different. Fairfax County is a suburban system. It’s the wealthiest county in America in terms of median family income this year, according to the Census Bureau, with the highest rate of parental education in the nation for the 1990 census. About 10 percent of the students are black, and 10 percent are Hispanic. There are more than 700 non-school-based administrators in Fairfax County. It’s very bureaucratic.

In 1987, the central office curriculum administrators persuaded the school board to make whole language the official policy of the entire county. Teachers were told to stop teaching reading and to let kids learn to read naturally, the way they learn to speak. For twelve years, we had no adopted readers. I had a phone call in 1993 from a teacher who said, “Rick, I just got told that if I got caught one more time with a spelling book on my desk, I’m going to be fired.”

So they went whole hog for whole language. In 1999, after the reading scores had plummeted and an election was coming up for the school board, they made a slight change. They made the schools buy a reading book, and the schools were given the choice of two books: one was an unstructured reader with a lot of stories in it; and the other was from Open Court Resources, which at the time was the only thing out there that had any structure in it, I think.

Out of those from 134 schools, how many principals bought Open Court? Two. Why? Well, if you’re a principal, what you quickly learn is not to irritate the people at the central office, because there are many ways that they can get even with you. Even though all school districts will tell you that our principals are empowered to run the school, the truth is that they’re all careful not to irritate the people at the central office who clearly supported whole language.

Today, in 2007, 85 percent of the schools still have not purchased any class sets of textbooks with any systematic phonics, much less the elements that Dr. Lyon talked about. They now call it “balanced literacy.” But how balanced is it? There are very few phonics books and very few vocabulary books.

This reality has had an impact on the budget. This year, in Fairfax County, 12 percent of our school budget is spent on what they call the “problems of children with learning difficulties.” Most of them are reading difficulties. Then we have some with difficulties in math.

In Fairfax County, one out of every ten teachers is a learning-disabilities teacher. Most of their students are not learning-disabled; at least half of them are curriculum-disabled. They are curriculum casualties, whom we are spending lots of money to remediate unsuccessfully, and many of them drop out.

In Fairfax County, we had seven schools that qualified for Reading First funding. Only one principal applied for the $1 million, and within a year, that principal had been replaced by a new principal who then pulled the school out of Reading First. If you’re a principal, you don’t do that unless you have very strong support at the central office level. Normally, that would kill a career.

The problem is that if you bring in an outside program to a school district where the administrators have decided the curriculum for years, and the outside program works better, then the administrators who are your bosses get angry because you’ve proved that they were the problem. Bureaucratic structures don’t like to be proved to be the problem. They’re going to get even.

A lot of the bizarre educational practices you see in Fairfax County are not bizarre at all if you understand that the schools are run in the interests of the people who run the schools—and those people are not the teachers and the principals. They’re the people who have the authority and the responsibility, either at the state level in California to set the curriculum, or at the district level in Virginia to tell the teachers what textbooks that they can use. They’re the people who put pressure on principals to buy stuff that’s favored ideologically. This is exactly what you see in Fairfax.

In Richmond, they cut the black-white achievement gap in reading from 30 percent to 13 percent in two years—they cut it in half. In 2005, the gap in Fairfax County was 26 percent. Every school district can cut its black-white achievement gap in half if it does what Richmond did: all it did was use common sense. It didn’t do what the wing nuts were proposing—vouchers, and a whole lot more money for the schools, and better teacher salaries (though I’ll admit that I would love to have those). They didn’t propose reconstitution. All Richmond did was what Dr. Deming taught the Japanese to do in the cases of Toyota, Sony, and Honda: give the workforce improved skills and improved tools and train them to use them. What Richmond got was Toyota-like results, and we all know how good that is. That’s all you have to do to improve achievement dramatically.

Let me summarize the lessons of Richmond versus Fairfax. First: poverty does matter, but curriculum matters more. That’s good, because a curriculum is cheap to fix. You could raise scores by cutting poverty in half, doubling teacher salaries, and cutting class sizes in half. But that’s tough to do—it’s not going to happen tomorrow. This is a cheap fix. Richmond bought stuff that worked instead of stuff that didn’t.

Second: science-based instruction works, and it works especially well. It’s not phonics. It’s not whole language. It’s beyond that. As you can see, in Richmond it got amazing results for children.

Third: I’m concerned that the penalties of No Child Left Behind fall heavily on the schools when, in fact, the people in the schools don’t decide what the curriculum is. The curriculum decides what teachers do in classrooms, and it decides whether or not children learn.

I hope that the people who are the stewards of federal law will admit that the law has some good parts as well as some bad parts that aren’t working. I hope that they will work to correct the problems. I don’t want to go back to the situation that we had with the Reading Excellence Act, where the law had plenty of good language that Congress could not enforce. If we do that, none of these school districts is going to change.

Finally, there was a time when attending the schools in this city and attending the public universities put you on the inside track for the Nobel Prize. I believe that if we can get rid of the curriculum bureaucracy and go back to a system in which the teachers are in charge of the curriculum and are informed by the science, and in which they’re required to do what the science says—as all professionals should be—maybe we can return to those days when going to the schools of New York put you on the inside track for that Nobel Prize. Thank you.

Questions for our panelists?

I’m from New York. My question has to do with testing. It’s clear that testing children to find out what they’ve achieved is an important thing, but I haven’t heard any of the speakers use the term “diagnostic testing.” And if we understand, as the first speaker said, that there are various needs and deficiencies that children have, how do we find out about them if we don’t test them diagnostically?

That’s a great question, and I’m sorry that I left that out. If, in fact, we know how kids learn to read, we also have to develop assessment instruments that are easy to use and that quickly give the teacher an idea of whether the youngster has mastered those skills. Those instruments should be used in the progress-monitoring phases of instruction, and they are diagnostic.

If you have a youngster who has been presented with evidence-based instruction, and you see gains in certain areas—let’s say that the student is getting better in vocabulary but not in comprehension—the assessment will tell you that immediately, and you can reframe your instruction. But the diagnosis is not geared toward whatever it is biologically, or genetically, that may place the youngster at risk. The diagnostic instruments are simply used to determine exactly how the student is responding to instruction.

Scientifically based reading research says that continuous diagnostic assessment has to be an intimate and explicit part of all instructional programs. It has to be carried out in Reading First schools because of the data that have to be reported. I can’t answer for schools that are not Reading First schools, apart from saying that our data, from the time when Reading First was implemented several years ago, indicated that most schools did not use progress monitoring.

Richmond uses diagnostic testing, and they say it’s one of the keys to their success.

I’m tremendously impressed with Dr. Lyon’s presentation, but I don’t like his answer to this question, at least in this respect. You said, “The thing to do is to test the skills that the kids have learned, and that’s the answer.” And I’m sure that is part of the answer.

I remember when Chicago went through a big business of reteaching skills. They tested the students’ skills and were very intent—I think they had developed 375 reading skills that the kids had to learn, and they drilled them on those skills. Then they tested the skills and found that the students did very well.

But then they found—and this is a big exaggeration, I’m sure—that by sixth grade, most kids had never read a book. So if we are going to use the scientific method and if we’re going to base everything on assessments—particularly if we’re putting decisions in the hands of teachers or even of schools—and if we’re going to test results, don’t we need to develop an assessment system that accounts not only for the skills that the kids are learning but for whether they do indeed learn to love reading, and whether they comprehend much better than the skills would indicate? Do the kids really learn to read, and not just learn a whole lot of individual little skills?

In the longitudinal studies that we did each year at NIH, we consistently asked whether the time spent reading increased in school and at home and whether the number of books that kids read increased as well. The relationship between learning how to read successfully and reading widely and frequently is substantial. I don’t know about all the kids who didn’t read a book in Chicago. But if we look at the 44,000 youngsters whom we studied over a period of five to thirty years, the relationship is extraordinary. Again, as I tried to point out, it’s hard to love doing something you cannot do.

I am deputy chancellor of the New York City Department of Education. How do these results carry over into the eighth grade and the higher grades as well? Even if you use graduation rates as a proxy, these are third- and fifth-grade scores.

Clearly, a lot of great things are going on in Richmond, but they can’t possibly all be attributed to Reading First because in every single subject—reading, math, science, social studies—there’s a positive differential. I’ve heard nothing in any part of the presentation here today that does anything but attribute all that to Reading First.

If you can’t read the test, you’re not going to do well on the test. In 2005, Richmond scored higher than the other nine big districts. They were number one in the state in fifth grade on three different tests. They beat the six suburban districts in the region, and they beat the other three urban districts. They beat everybody. Fairfax County, on the other hand, lagged on every test.

One critical issue you’re bringing up is whether, if we bolster reading capabilities at foundational levels with Reading First, those capabilities will be able to translate into results later on. Will they continue as the kids move forward?

The youngsters who are the beneficiaries of Reading First come into the system with as much difficulty in vocabulary and background knowledge as in word-level skills. Those reading programs that intensively, directly, and explicitly make sure that vocabulary is heightened tend to serve the youngsters fairly well as they move forward. What are the conditions?

For the kids most at risk, the conditions are that the program has to be comprehensive: teach phonemic awareness, phonics, fluency, vocabulary, and comprehension. Those capabilities have to be taught directly and systematically. Kids without background knowledge do not acquire them easily unless they are taught intentionally. When that happens, and when the kids spend enough time on those activities, coupled with teachers who use diagnostic assessments, you get generalization. In our studies, eighth-grade scores look pretty good when those conditions are in place. Scores look pretty good in the twelfth grade, too.

You’re asking a question about kids who don’t get into Reading First until the second or third grade. How well will they do? We’ll just have to follow their progress. Many people want to see Reading First fail, which is amazing. But I’m not saying that you want it to fail; Rick and Maria looked at some of those conditions that have increased capabilities at certain ages. If students don’t read in the eighth grade, we haven’t gone anywhere. So let’s figure out how to extend what we’ve done fairly well in the early grades.

The best way to get the eighth-grade reading scores up is to adopt some of the ideas and principles of Core Knowledge, as articulated by E. D. Hirsch. I would suggest that the DOE give it all the support it can get and then monitor and compare the results of those schools with those of the other schools.

When I put together these charts, I addressed all the subject areas. I wasn’t looking for reading in particular. However, schools that looked for what works in reading also taught well and looked for the best approaches to every kind of instruction, especially math instruction. They tried to make the best use of every minute of their day, and that affected all scores in the same way. It wasn’t just that the schools directly affected reading.

It’s also hard to do math if you can’t read.

I am a seventh- and eighth-grade English teacher in New York City. I’ve been very frustrated by the Teachers College curriculum. Many other teachers I know are intimidated into teaching the Teachers College curriculum. This curriculum expects children to sit in little groups and read as many little books as they can to themselves, the theory being that the more they read to themselves, the more they’ll understand, and the more they’ll know without any instruction from the teacher. I am not a proponent of this way of thinking, obviously. I’d like to hear your comments about the Teachers College curriculum.

Thank you for your comments. I’m not a fan of the workshop-model approach, because I’m a strong believer in a content-based curriculum. I think that kids should—starting in kindergarten and even in pre-kindergarten—be learning about the world. They should be getting history, civics, geography, science, and literature. And the literature should not be from the box of books that came into the classroom or from a library that has absolutely no literature of any quality. It should be carefully selected to introduce children to the central myths and tales and biographies of our society and our culture and the cultures of the world. That’s called the Core Knowledge curriculum.

As I spoke before a Core Knowledge group recently and was describing what they try to do, I thought, “We used to just call this good education.” It didn’t have a name, Core Knowledge. It was called good education. But now it’s been supplanted all over the country by all this process orientation, which robs children of knowledge and vocabulary—and then we’re surprised when the children don’t turn out to be good readers. So I think that you’re right.

I’m a former superintendent in the South Bronx. I was a building principal in District 7 for twenty years. Teaching students, especially Hispanic and black children, to read is my passion and my devotion. I know that they must learn to read or else they will not succeed.

What do you think of the DIBELS (Dynamic Indicators of Basic Literacy Skills) program? Do you think that it is the type of program that should be in our schools?

The DIBELS program, as a measure—if you put aside the issues relating to its implementation—is well constructed. All it is designed to do is to measure, on a frequent basis, the ability of kids to read rapidly. Why does it focus on fluency? Because fluency is a proxy for all the word-level skills. You can’t read fast if you don’t have phonemic awareness and phonics. So if you’re reading fluently, it’ll cover those. If you’re not reading fluently, the teacher then knows to go to the lower word-level skills and identify specifics. Fluency is an extraordinarily strong predictor of comprehension and vocabulary. As a means of giving teachers immediate information so that they can modify instruction, DIBELS does a good job.

I am executive director of the New York Civil Rights Coalition. I want to follow up on Gail Badillo’s question. My question addresses the so-called at-risk teacher, because teachers aren’t what they used to be. I know teachers who believe in such things as Ebonics, which they actually teach.

So how do we know that teachers know how to read? How do we know that teachers can teach reading skills? What are the specific assessment instruments for teachers?

We have studied teachers for a long time. Teachers ought to be good readers. In reality, it depends on whether you’re looking at inner-city districts. I don’t want to paint with too broad a brush here, but on average, teachers in inner-city districts are much less fluent. We do have teachers who are reading between the sixth- and the twelfth-grade levels. When we measure their phonemic awareness—that is, how they understand the sounds within the language—my colleague Louisa Motes and her team have found that phonemic awareness is a difficulty. Can adults learn to do all of that? Absolutely. But remember, all of us are products of where we’ve come from. It’s not the teachers’ faults. It’s the system that was in place where they were taught.

To answer your question, teaching capabilities are among the strongest predictors of student learning and achievement. If we don’t ensure that teachers have the proper capabilities in the field that they’re teaching, we’re going to see that in the results.

Everybody hates tight, scripted reading programs, but a lot of them are very good, for two reasons: they’ve been through experimental trials and come out looking pretty good; and they train the teachers as they’re teaching the kids. The programs are so scripted in requiring the teachers to present the information that, through teaching, they actually learn a good deal.

Let me just add a teacher’s perspective on that. The teachers I know who have been in very good reading programs have complained that they should have learned this stuff in college and in the education schools. In Virginia, we now have a test to see if the education schools are teaching science-based reading instruction. The test is good, but it’s been slow to be implemented, partly because under No Child Left Behind, the education schools face no penalties if they don’t teach reading. They keep getting hundreds of millions of dollars in federal aid. There ought to be some accountability there.

- Luncheon Address -

Margaret Spellings was sworn in as secretary of education on January 20, 2005. Prior to that, she was an assistant to President Bush on domestic policy, including areas such as education and immigration. Prior to that, she worked for George Bush when he was governor of Texas. She also worked on education reform committees throughout the state.

But since we’re talking about No Child Left Behind, the important thing today is that Margaret Spellings was the prime mover behind the No Child Left Behind Act and has been responsible for implementing its provisions. When I was in Congress in the 1970s, Congress did not want to get involved with education because it felt that education was a local matter. It was difficult to get anything approved that would regulate the states and localities in any way. Therefore, No Child Left Behind is a tremendous achievement, because it forces the states and the localities to do things that they haven’t done before.

For example, today, the New York City Department of Education says that the graduation rate from high school is 60 percent. But two weeks ago, it said that the graduation rate for Latinos was 45 percent. That is tragic. The No Child Left Behind Act compels states and localities to break down the results not just by general category, but by ethnic groups as well.

It’s very important that No Child Left Behind and Secretary Spellings be supported, because I believe that education is going to be the most important domestic policy for the next generation. Take the Latinos—we have a population that is now 15 percent of the country, the largest minority ethnic group, going up to 25 percent. And the kids are not performing. That is a national disaster.

No Child Left Behind can help to change that. And that’s why I’m delighted to present Secretary of Education Margaret Spellings.

Thank you very much, Herman. I appreciate that great endorsement. Not much has changed in Congress since the 1970s; lots of people still feel that way, I’m sorry to say. But with your help, we will get No Child Left Behind reauthorized this year. I very much appreciate the opportunity to talk with you about something near and dear to all our hearts.

Sol Stern, senior fellow, thank you for your participation today. And my friends Reid Lyon and Diane Ravitch: I’m sure your panel was spectacular. Thank you not only for your participation today but for all the good work that you’re doing with regard to reading. I believe that if we get reading right, we are on our way; and if we don’t, we ought to just close shop. The good news is that we have a lot of great results to talk about in reading, and that’s in no small part due to you and your good work.

We continue to hear critics of No Child Left Behind say that our focus on these basic skills of reading and math distracts from teaching other subjects—that we’re narrowing the curriculum. But how are students going to master history or science without being able to read or to decipher? This is as obvious and commonsensical as anything. We all know that reading and math are the subjects that are the gateways to every other area of learning. That’s why I was so pleased, and I’m sure you all were, that the new NAEP data show that our young students are making very good strides in both history and civics.

The report on U.S. history that was released last week shows increased scores across the board and a narrowing achievement gap among our fourth-graders. Similarly, in civics, fourth-grade students showed improved scores and a narrowing gap between white and Hispanic students. These reports confirm what we all know: if you can’t read, you can’t read the history test, the history curriculum, or the history book. The reports also affirm to me that this “narrowing and teaching to the test” stuff is a lot of baloney as well.

As you all know, I’ve been working with President Bush on education issues for a long time. A lot has changed, of course, since our days in Texas, when we were one of the incubators for some of the policies of No Child Left Behind. But he and I continue to be guided by the same principles that were at work then.

One of the president’s first stops on the campaign trail was here at the Manhattan Institute: in 1999, he came to talk about his core philosophy and how, as a different kind of Republican, he was going to talk a lot about education. He sometimes reminisces about the days that Herman talked about, when lots of people were talking about abolishing the Department of Education, and he often says that people just hear “abolish education.” That’s not where we want to be.

So I’m proud and pleased that the president has worked to change the national conversation on education. Sometimes we get bogged down in mythology, which I want to confront, but I do think that it’s a very different discussion today, in no small part because of No Child Left Behind and the president’s ability to frame the issue of education.

The first thing I want to talk about is some of the core beliefs that have informed our policies. We know for sure, as you do, if you heard the panel that preceded me, that there is such a thing as scientific research—or data-driven decision making, as we now call it—when it comes to education policy. We have to use that research to inform our policies and our investments, just as we do in every other endeavor. We also know that the federal government is only about a 9 percent investor in K-12 education. But our experience has shown that it’s a very important 9 percent, and we need research to focus our policies and resources where they will be used to maximum effect.

Reid Lyon has done great work to transform scientific insights from the laboratory—the things we’ve learned about the brain—into practical tools for our reading teachers. The Reading First program that my department runs grew out of twenty years of research that now is helping more than 2 million schoolchildren make gains in fluency and comprehension. We have proof in Reading First. So where research shows what works, let’s do it, as we do in medicine and other fields.

The second thing is that parents do know what is best for their children—I know that this is sometimes belied—especially now that they are armed with data and information about their schools. I’m saying this not only as the secretary of education but as a mother of schoolchildren. We believe that the wisdom of parents and families must be brought to bear on education reform. In particular, it’s helpful when we look at school choice options.

Thanks to the president, if I may say so modestly, we have done more to expand choice and opportunity in education than any other administration has. Exhibit A is the first ever federally funded Opportunity Scholarship Program in the District of Columbia, a program that is now helping 1,800 students from economically disadvantaged backgrounds attend fifty-eight private schools.

We’ve also been huge supporters of the charter movement, providing significant resources for school facilities and for start-up funds. The charter movement, which just last week celebrated its fifteenth anniversary, is helping to dispel the myth that some children can’t learn. Charter schools act as great laboratories for some of the best practices. I met some of the people who are involved in KIPP (Knowledge Is Power Program) here in New York City. Alternative public schools are great examples of this, and we’re all learning a lot about these innovation laboratories.

The third thing is that we need high standards, and people at the state and local levels are in the best place to set them. Since we’re a 9 percent investor, it’s right and righteous for those who are paying the bills to set the standards. We don’t need to establish federal standards; No Child Left Behind doesn’t call for that. But I do think that we have some genius in the policy, with the NAEP data being made more widely available. Meaningful accountability, of course, must include deadlines and consequences—just as we have with No Child Left Behind and the year 2014—along with the flexibility to achieve those goals.

Thanks to No Child Left Behind, the NAEP has become more accurate and more informative, because now every single state is required to participate in that national report card. It’s a much better report card than it was when we had states coming in and out of it annually. It is the only national assessment that tells us with accuracy how we’re doing. Now that local policymakers have this information, they can act on it; and I can tell you, having served at the state level, that people look at it closely. They don’t want to be dead last but still paying the bills. I do think we can do a better job, in reauthorizing No Child Left Behind, of ensuring that parents have the NAEP data so that they’ll be armed with even more information than they have now.

The fourth thing that we know for sure is that teachers make the single biggest difference in enhancing student achievement. So we have to do everything we can to get our best teachers in our most challenging educational settings. We need qualified teachers to deliver a rigorous curriculum that challenges students. No Child Left Behind is a floor, not a ceiling. It is necessary, but not sufficient.

The old solution to education challenges, of course, was to spend money and cross our fingers, or simply hope for the best. Now we can actually find out what works, because we measure with regularity. One of my mottos is, “What gets measured gets done.” You all see that in your work; it guides every other endeavor in American life. Certainly, it’s a welcome principle in education.

The basic premise behind No Child Left Behind is that we expect results from our federal investments. That’s a smart and wise thing to do as taxpayers, not to mention that it’s good for kids. For a long time, children—especially our neediest kids—were shuffled through the system, as the president says, and were left to just move on through, without the necessary skills, until they either dropped out or were given a diploma that didn’t mean a lot. I have yet to meet a parent who says, “Count my child out—I don’t want my kid on grade level by 2014.” Mostly, they say, “I’d like my kid on grade level today.” I’m pretty sure that the parents in this room feel this way, irrespective of their neighborhood, color, or income level.

We are already seeing some very promising results from No Child Left Behind, among our young readers. We have made more progress with our young readers over five years than in the previous twenty-eight years of our national education report card. It tells us that this policy recipe is working. But it also shows us where we need to continue to work.

Similarly, we have seen little progress among our high school students over that same period of time, nearly thirty years. We know that we have to be smarter about targeting resources and strategies in our chronically underperforming schools so that we can know specifically where we are. Those are some of the key issues that will be before us as we renew and reauthorize No Child Left Behind.

In addressing our lowest performers, those chronic underachievers who for more than five years have missed No Child Left Behind targets—that’s about 2,000 of 95,000 schools across our country—we simply have to bring more vigorous tools to bear. We have to give superintendents the opportunity to staff those schools with our best teachers. We have to give local officials the opportunity to charter those schools, notwithstanding arbitrary caps that might be set. And we have to give parents, who have waited too long for options, the opportunity to get additional help or scholarships to send their children to private schools or to obtain additional enriched supplemental services. That will be a key issue in the No Child Left Behind reauthorization.

Another key frontier will be, of course, strengthening our high schools. We have to change the fact that, as Herman Badillo noted, about half of our Hispanic kids get out of high school on time. It is a national crisis. Here in New York City, the graduation rate at about ninety high schools is worse than 50 percent. Our high schools, I think, are often failing to prepare our kids adequately, not only for the workplace, but for college as well. We simply have to do more to expand access to rigorous course work, such as advanced placement classes, and train more good teachers to teach those classes.

We also need to do more to link high school standards with the expectations of the workplace and of higher education. We have to be honest about dropout rates. And as we continue the fight to empower parents and promote choice and turn around failing schools, we must stay focused on this great goal of getting every child on grade level by 2014.

As I’m sure you all are aware, recently some conservative members of Congress have suggested overhauling No Child Left Behind by basically reverting to the old days of sending resources without demanding accountability. We cannot fix education, or pick up the pace, without accountability and without the deadline of 2014. Flexibility without accountability is an absolute recipe for failure. We cannot afford to go back to the ways of the past. We tried that, and we know that it doesn’t work.

So if you are committed to turning around our chronically underperforming schools and to making sure that this country remains the world’s innovator, then No Child Left Behind must be reauthorized this year. If you are committed to more flexibility, and to preserving momentum for school choice and local control, No Child Left Behind has to be reauthorized this year.
We have a moral responsibility to give every single student a chance for success. Only education builds the skills, the habits of mind, and the knowledge for our children as well as for our country. This idea goes back to what we Americans believe in. It goes back to our founding. And it is the key to the American dream.

I look forward to working with you this year and for your continued strong support for this very important law. Thank you.

Secretary Spellings, how can we possibly meet the goal of a qualified teacher in every classroom by 2014 if the federal government is completely agnostic about the education schools that are supposedly going to train these qualified teachers? How can we continue to certify education schools that don’t even teach the science of reading?

The law requires that we have highly qualified teachers long before 2014; in fact, we’re supposed to have them now. One of the conversations that we’re having in Washington is about how we can go from an input-driven system to one that talks about efficacy and highly qualified and effective teachers, where we start to use data. I’m talking about student achievement results, to more accurately reflect who our best teachers are.

I think we’re changing the conversation from inputs, training, course hours, and so forth to talking about who’s doing the best job in the classroom. I think that change will clearly be debated in Washington, but it certainly makes a lot of sense to me.

You spoke about the imperative of 2014, and I certainly agree that we need accountability and results. But we all know that eliminating the achievement gap—reaching proficiency for all by 2014—is an impossible goal. I also know that in 2001, when you and the Democrats worked out the coalition for No Child Left Behind, there was a feeling that we needed this 100 percent proficiency as a motivator and that it would make people pay attention.
I think that has worked. But now that we’re five years down the road, isn’t it time to say that we have to come up with a more realistic output measure? If we go down this road, we’re calibrating Annual Yearly Progress (AYP) standards to an impossible standard. More and more schools are going to be listed as needing improvement. And it seems to me that this whole effort is going to implode upon itself.

No, the effort won’t implode upon itself, and here’s why. On any given day, in any given state, there are kids outside of the accountability system. We’ve provided the flexibility to have it that way. One percent of the student population is profoundly handicapped, and, as such, these students are obviously not part of the accountability system.

We published the new rules following the reauthorization of the Individuals with Disabilities Education Act (IDEA), which said that there were an additional 2 percent of students outside of the accountability system. Remember, special-education numbers are growing; this is a large number of kids. These students are going to take more time, different strategies, and different kinds of assessments. Now we’re up to 3 percent. We have limited-English students who are allowed to be exempted from the system for one full academic year—or more, if they’ve come in the middle of an academic year.

Also, the states decide the sample size necessary to have a valid and reliable group of students to be counted.

Accommodations have been made to reflect the reality of this organic system called public education, and I think that’s appropriate. But I also think that we have established an achievable goal. What parents want to say that their kids aren’t going to be on grade level?

Madame Secretary, my home county and yours kept all their eligible schools out of Reading First. As a teacher, I know that many things in my classroom happen because of decisions made at the central office level. Yet in No Child Left Behind, almost all the penalties fall at the school level and on the school staff, for decisions made by the central office that tie our hands.

Governor Spitzer in New York has proposed that, as part of the state accountability system, school boards and superintendents be held accountable. That’s something that’s missing in No Child Left Behind. Can we add it for the reauthorization?

That’s an interesting question. As you mentioned, I’m also a resident of that school district—Fairfax County, Virginia. We’ve had some issues about the assessment of limited-English students there. Nationally, two-thirds of our limited-English students are United States citizens, 80 percent of whom have been here for five years or longer. So as we read these stories that say, “So-and-so got here six months ago, and now that mean old secretary is expecting him to be proficient in English,” let’s be mindful that lots of these kids were born here. Can we make some distinctions about arrivals? Sure.

With respect to accountability for school boards, superintendents, and others who are making decisions, one of the most profound things about this law is the information about results that enables people like you to say, “Well, how come there’s $1 billion of federal aid flowing toward needy readers and we’re not getting any?” I think that the power of sunshine is an important dimension. Whether there will be specific tools for me to enforce requirements on school boards and superintendents, I don’t know.

I’m from the Women’s City Club. I think that you put up a straw man when you spoke about children who have been here five years or longer. I come from Queens, which has the largest number of new immigrants in the whole city. Every one of our schools is going to be considered a failing school because your law says that if a student is here for one year or longer, he has to show the same results as other students on the test.
This is ludicrous. We do not have failing schools; we have wonderful schools. They are going to be called failing because of your law.

Let me clarify on the two-thirds, 80 percent calculation. That, of course, is a national calculation. There will be community anomalies in any given place. I’m not asserting that two-thirds of the students in your schools are United States citizens. But nationally, that is the figure.

With respect to the transition issues, clearly, that’s what I’m speaking about when I talk about the need to build nuances into the accountability system. No Child Left Behind, as you know, is largely a pass-fail system. We need to start making distinctions between schools that are within range and those that are chronic underperformers, with five or more years of not meeting AYP. And I think we’ll certainly do that as part of No Child Left Behind.

But I think there is a right balance between intensity of effort and throwing our hands up and saying, “You know, we don’t think that this majority-minority population in Texas or California can read on grade level.” That balance is what we’ll discuss in Congress.

Since the 1960s, we’ve been throwing hundreds of millions of dollars at Title I, Head Start, and all those things to boost the bottom, and we’re still not up to par.

Meanwhile, money for gifted programs has virtually dried up. As you well know, the Javits Program, which I think was the only federally funded program for the gifted, was canceled this year. Even so, there was not much in it to begin with; it was a few million dollars. Most of that, actually, was directed toward kids at risk, so it wasn’t truly a program for the gifted. In many states, money that was normally going to the gifted has now been pushed over to satisfying the demands of No Child Left Behind.

It’s my opinion that our gifted program in the United States has become the H1B Visa Program. Every year, we import perhaps 100,000 gifted people, maybe more, to fill the positions that we cannot supply ourselves.

What is the Department of Education doing for gifted children—the people who are the future Bill Gateses and Larry Ellisons—aside from pursuing a strategy that has proved ineffective for forty-five years or longer, namely, pouring money into the bottom?

Let me start with the Javits comment. It’s a $1–2 million program, and Congress has made the judgment that it’s hard to have a nationally scalable program for fifty states, plus the territories, for $1 or 2 million. You probably agree with that.

Just as a point of evidence—and if you have more, please share it with me—we are not seeing that bringing up the bottom means that we are keeping the top down. It just doesn’t bear out in the data. We do have more of a rising-tide-lifts-all-boats phenomenon. We’re seeing gains across the board. That’s the second point I would make.

Third, I would say that what is different about No Child Left Behind is that it is a game-changer away from the thesis that you first laid out. We are about results in exchange for resources. This is the first time we’ve done that. It’s not true, frankly, of Head Start. That program is not run by the Department of Education; it’s at Health and Human Services.

This is the first time we’ve said that we have some expectations for a particular goal. There’s more accountability here for federal tax dollars than ever in the history of the forty-year commitment to education.

What do you do about districts, particularly urban districts, that are living with No Child Left Behind and yet somehow life goes on just the way it did before the statute was passed?

I would just say that we passed the best law we could five years ago, when about half the states did annual measurement and we didn’t know a lot about certain things. Have we learned things in the last five years that we ought to be mindful of and be guided by in the reauthorization? Of course we have. That’s why the president’s reauthorization proposals speak about some of the things that you’re talking about—the need for a growth model, for instance. That would help on your issue of being able to chart progress more accurately over time.

But when we were trying to take a snapshot of the accountability system in half the places around the country, half the states waited until the 2005–06 school year to do annual assessment for the first time.

Can we be smarter and more precise about doing that now? Yes, we can. I’ve given five states waivers—and I’m actually going to do a couple more this week— for this growth model notion. Are there things that we can do to fix and be watchful of the unintended consequences? Yes; that’s why we have reauthorizations in Washington.

Madame Secretary, fifty-three years after Brown v. Board of Education, which outlawed racial segregation in the public schools, there are school districts in this nation that are trying to show that segregation by race will work to raise the academic achievement level of black male students, despite Title VI regulations.

What are the Office of Civil Rights and the Department of Education doing to counteract this racial idiocy?

If you have the specifics of that particular assertion, I would obviously like for my Office of Civil Rights to look at those. Obviously, that assertion is not in keeping with the law that I took an oath of office to uphold.

What I see and talk about now—and this is why looking at our high schools is so important—is a rationing of opportunity, if you will. Forty percent of our high schools have no advanced placement classes. I use the example of our neighborhood. In Fairfax County, Virginia, Langley High School has twenty-eight advanced placement classes. Ballou High School in inner-city Washington, D.C., has maybe three or four.

We all know that our most experienced, most degreed, and often most effective teachers are at Cream Puff High, while our least supported, brand-new, unmentored, lowest-paid teachers are in our most challenging educational settings. We have to do something to reverse that, such as rewarding teachers through pay systems for doing the hard and challenging work. Clearly, I worry about the issue that you raise, but I also worry about what undergirds that, which is the rationing of rigor and personnel that we often see.

Coming from the sciences, I’m always astonished that discussions like this happen on a national level, because in the sciences everything is transnational and relatively border-free. With that as a context, what is your response to the OECD (Organisation for Economic Co-operation and Development) Programme for International Student Assessment (PISA) studies of international comparisons of national educational systems?

I think that it shows us that we have work to do in focusing on math and science. You are all scholars, and you know all the issues surrounding the use of PISA data to rate our schools or to make comparisons. But there certainly are some takeaways there, among them the need to focus on our high schools and on math and science more intensively than we have.

A great deal of emphasis earlier today was on Reading First and on the virtues of phonics programs as compared with whole language. These things are supposed to have been scientifically proved. Yet we are told that they are not required by the Department of Education in funding local programs. What is your view on that? If phonics is better, as is widely asserted, shouldn’t the Department of Education reflect that policy judgment in giving out all this money?

The law says that there are certain criteria—research-based principles that ought to exist to teach youngsters how to read. There are a myriad of products, programs, plans, and approaches that meet those criteria. Can we do a better and more appropriate job of overseeing the program? Yes. That’s why I adopted every recommendation that the inspector general said that I should to improve the oversight and stewardship of that program.

We can take this core set of principles, however, and ask states to develop plans that meet those criteria while still allowing them to employ many different approaches, strategies, or products. It’s a combination of national core principles with flexibility and local control. These things are not in disharmony; and that’s how we ought to do it.

I’m from With respect to Reading First, your inspector general said basically that the New York State Education Department shouldn’t have received a grant. Your inspector general’s subsequent look-see has shown that the state Department of Education gave out money to districts under Reading First in accordance with some agreements or arrangements that really had nothing to do with Reading First. We’ve also had similar audits with respect to Title I.

I find it hard to support these federal programs when your agency is not requiring that strict requirements be met in this state. So I’d like to know what you’re going to do about the New York State Education Department and its handling of Reading First and Title I.

Clearly, Congress is going to provide additional guidance with respect to conflict of interest provisions and such things. I’m not going to get into the specifics of particular state grant approvals. But as I said, the good stewardship of very large grant programs and the integrity of those programs are of paramount importance to me, particularly when they’re proving to get such great results for kids.

I’m from the Cahn Fellows Program for New York City distinguished principals. You mentioned that No Child Left Behind is a floor, not a ceiling. You also said that what gets measured gets done, typically. If No Child Left Behind is how schools are measured, I’m assuming that it’s what the schools are going to shoot for. Many schools, we know, do fall short. But what’s the incentive for reaching the ceiling—or even the light fixtures—in striving for success on the AP exams, the SATs, the ACTs, true graduation rates, and other things that would promote innovation through education?

The federal commitment to education has been directed to our nation’s neediest students—poor kids, special-education kids. That is how we’ve engaged for the last forty years, pre–No Child Left Behind.

This is why the NAEP is important; when you have this kind of information, the federal government has a role to play. And that role is to see that every child is performing at grade level by 2014. But I can tell you, having worked for two governors, that it’s also incumbent upon governors. There are certainly no state impediments. If I were still working for the governor of Texas, I’d say, “Let’s have an accountability system that expands the subjects that are taught. Let’s measure social studies. Let’s measure history. Let’s measure these other dimensions. And let’s ask ourselves how well we are doing across the spectrum.”

Now we have the highway in place. We have the infrastructure. We can ask ourselves how many students are on grade level. We can also ask ourselves how many students are at the top. I would recommend that those of you who are in this arena talk to your state legislators and your governors about filling out your accountability system now that No Child Left Behind has brought this infrastructure to bear. These are knowable values that state and local policymakers ought to be looking at.

(Figrues supplied by Maria Casby Allen and Rick Nelson, Fairfax County, Virginia.)

Figure A

Figure B

Figure C

Figure D

Figure E

Center for Civic Innovation.



Reading First, part of the 2001 No Child Left Behind Act, seeks to lift reading achievement by encouraging the use of reading programs that have been scientifically proven to work. Reading First, which accounts for just 2 percent of federal education spending, is helping many districts to achieve promising results—including Richmond, Virginia, where test scores have risen dramatically. But negative publicity stemming from recent reports by the Department of Educationís Office of Inspector General could put Reading First under a cloud when NCLB comes up for Congressional authorization. That would be tragic for millions of children at risk for reading failure. To see clearly what's at stake in the reauthorization battle, the Manhattan Institute gathered a panel of experts on the science of reading instruction. As our luncheon speaker, U.S. Secretary of Education Margaret Spellings then discussed the prospects for reauthorization of No Child Left Behind.



Home | About MI | Scholars | Publications | Books | Links | Contact MI
City Journal | CAU | CCI | CEPE | CLP | CMP | CRD | ECNY
Thank you for visiting us.
To receive a General Information Packet, please email
and include your name and address in your e-mail message.
Copyright © 2015 Manhattan Institute for Policy Research, Inc. All rights reserved.
52 Vanderbilt Avenue, New York, N.Y. 10017
phone (212) 599-7000 / fax (212) 599-3494