State education aid is up $2 billion from what it was in 1993, the year the state’s education reform law was enacted. While there’s little doubt that school programs are richer in countless ways and that quality education is now at the top of everyone’s radar screen, by any honest reckoning we haven’t got $2 billion worth of improvement.

As a state we have failed to understand what’s needed to improve our schools. The state has in place many of the key elements of what could become a successful strategy for turning schools around. But there is to date no full-scale program for school improvement. Until we have one–one that is based on what we know makes a difference in schools–progress in educational performance will continue to be disappointing.

And progress has been disappointing. Public attention has been focused on the 10th-grade MCAS results, since next year’s seniors and subsequent classes must pass this exam to graduate. But consider a slightly higher standard of performance–rates of proficiency. In the charts below, “proficient” means students whose scores are in the “proficient” or “advanced” categories under the MCAS scoring system–the same definition the state is likely to adopt under the new federal education law. Students who score in the “needs improvement” category have passed the MCAS and will receive high school diplomas, but they are not “proficient” –a standard of academic adequacy set by the state Board of Education with considerable input from educators. I use this higher standard because education reform was about educational excellence, not barely passing. In any case, the trends over time are similar, whichever standard of performance you track.

We haven’t got $2 billion worth of improvement.

Confronted in 2001 with a test that really mattered for the first time, 10th-grade students had a higher proficiency rate than in previous years. Perhaps this was a result of improved instruction, at least in part; it’s more likely that students simply took the exams more seriously than before. Readers can judge for themselves by looking at the first chart, which breaks down results along racial and ethnic lines. For all categories of students, 10th-grade scores were essentially level for three years and then jumped suddenly in 2001. Had there been underlying progress in student learning, it seems fair to expect that we’d have seen scores rising over time.

A better measure of progress over time is offered by the three MCAS exams for which we have four years of results, and for which there has been no change in consequences (unlike 10th grade exams) or in scoring (unlike fourth-grade language arts). These are the fourth-grade math exam and the eighth-grade exams in math and language arts. Back in 1998, 40 percent of white fourth-graders were proficient (or better) in math; by 2001 this had risen by only two points–virtually no improvement at all. Only 9 percent of blacks and Hispanics were proficient in 1998; three years later this had risen to 10 percent for blacks and 11 percent for Hispanics.

Ten years ago, all of us working on the reform law–and most educators as well–believed that if, by some miracle, the state were to adopt our funding targets and stick to them (as it now has), we’d see major progress. That only 42 percent of fourth-grade white students statewide were proficient in math eight years after the funding increases began is deeply disappointing–particularly since this rate is barely higher than when we started.

If hardly budging scores for white students is disappointing, the results for minority students are truly tragic. Most of the new money has gone, as it should, to inner city schools. But after eight years of additional funding only 11 percent of minority students in the fourth grade are proficient in math. This is not simply a problem for inner city schools. The proficiency rate in math for fourth-grade black students in Newton and Brookline–26 percent and 28 percent, respectively–is barely better than in Chelsea (25 percent). The 101 black fourth-graders in Randolph had the same proficiency rate–14 percent –as their 389 colleagues in nearby Brockton. The highest proficiency rate for fourth-grade black students anywhere in the state–32 percent, in Lexington, Wareham, and Plymouth–is well below the average statewide score for white students.

Scores in 2001 on the eighth-grade math test are broadly similar to fourth grade–about 10 percent of minority students are proficient. The best results are on the eighth-grade language arts exam, where there’s been steady, marked improvement for all students. At 74 percent, proficiency rates for white students are close to what we’d like to see. At 36 percent and 31 percent, proficiency rates for blacks and Hispanics, while far from acceptable, are up markedly over a three-year period. One of the strengths of the MCAS tests is their emphasis on student writing. This has had a positive impact on schools across the state. The heartening improvement in eighth-grade language arts results may well reflect the fact that schools are having their students do a lot more writing.

It’s clear that the challenge of education reform is to find a way to educate low-income and minority students. (Student performance is correlated as much with poverty as with ethnicity. In 2001, the proficiency rate in fourth-grade math for students eligible for free and reduced cost lunch–14 percent–was essentially the same as the rate for blacks and Hispanics.) After nine years of education reform, we have to ask why schools haven’t been able to do the job.

A faulty assumption

The fundamental premise behind the 1993 reform law was that if only they had clear goals and adequate funding, educators would be able to improve schools. It now appears that assumption was faulty. As Harvard professor Richard Elmore, who’s written extensively on education reform, observes in a recent issue of the magazine Education Next, “Low-performing schools, and the people who work in them, don’t know what to do. If they did, they would be doing it already.”

I don’t want this observation to be misconstrued. In the course of my work as an education consultant, I’ve met literally hundreds of teachers and scores of principals and superintendents. In the main, I like and admire them; I find them to be hard-working, caring people who genuinely want what’s best for children and who went into education because they want to contribute. As in every profession, there are bad apples, but the share is probably a bit smaller than in the population as a whole. Nonetheless, many–most–educators do not know how to turn schools around.

This came home to me a year ago when I had the opportunity to visit schools in Lynn and Everett–communities where the superintendents had clear strategies for improving schools–and where there was evidence, both from test data and direct observation, of student improvement. Once you’ve seen such strategies in action, the absence of them elsewhere becomes clear.

This absence is no accident. In a paper written for the Albert Shanker Institute, Elmore offered the following explanation: Traditionally, public schools were staffed mainly by relatively low-status female teachers, generally working in isolation from one another, and supervised by (mostly male) administrators whose authority was more administrative than pedagogical. As a result, educators and the public came to see the core of teaching–what and how students should be taught, how much they should know, and how they should be evaluated–as more art than science, and best left to the individual teacher. The profession tends to believe that people are good teachers because of personal traits (“she’s a born teacher” ) and not because they’ve mastered a body of professional knowledge or because they work in an organization that expects excellence and helps its employees achieve it. Administrators, meanwhile, are there to buffer and protect the teacher from outside pressures and interference, not to provide educational leadership.

It’s not that educators actively resist constructive change, it’s that they’re not oriented toward the use of data and systematic analysis about education itself. When I ask educators about the 30 years of research on how children learn to read (and how to teach reading effectively) done by the National Institute of Child Health and Human Development, fewer than 15 percent indicate they’re familiar with this work. Imagine where Gillette would be if only 15 percent of its engineers followed the research on the cutting properties of steel. Or Raytheon if only 15 percent of its college-educated workers kept up with the research on rocket propulsion.

There’s no other profession less oriented to research and to data than teaching. We need to build a strategy that takes this orientation into account, working with teachers to help them systematically improve their teaching and with administrators to teach them effective reform strategies.

That’s not the way we’ve gone about it so far. Our strategy has been to provide more money, set high standards, administer MCAS exams, and publicize the results, relying on pressure from parents and community leaders to force educators to make changes. Under former Board of Education chairman John Silber, this approach was carried out in a particularly aggressive way that alienated most educators.

In fact, teachers are not motivated by public embarrassment or incentive pay, according to Richard King, a University of Northern Colorado professor who’s studied state efforts to improve teacher performance. What does motivate them is demonstrated student success. I can attest to this from my own visits to highly successful schools. I am absolutely convinced that when teachers see proven techniques that help their students they will adopt them, and do so enthusiastically.

Testing for improvement

We’ve got the graduation requirement about right–a comprehensive 10th-grade exam looking to identify advanced and proficient youngsters as well as those in trouble, followed by a series of “make-up” tests with questions only at the basic level. As currently configured, however, MCAS cannot do effectively one of its most important jobs–determining which schools are making reasonable progress and which are not. Comparing the performance of this year’s fourth-graders with last year’s is not a reliable measure of change, since each class is different. Consider the small Lynn school that typically had 25 to 28 students in its only fourth-grade class–and usually with a large number of students with difficult behavior. One year they had only 18, and all were orderly and well-motivated. Test results soared–but not because it had become a better school.

Similarly, the tendency to compare scores between communities is misplaced. That average scores are lower in Lowell than in Wellesley has more to do with the impact of poverty on student readiness than it does with the quality of teaching. (I suspect that if the two systems traded faculties there’d be surprisingly little change in either town’s results.) But it’s not unfair to hold Lowell teachers accountable for how they play the hand they’re dealt. At the very least, teachers should be accountable for each student making a year’s worth of progress each year –ideally, more than a year’s progress, so that by fifth or sixth grade they’re reading at grade level.

We will have the capability to do this from this point forward. The new federal education law requires annual testing in English and math from third to eighth grades, and the state Department of Education has put in place a database that can follow individual student progress from year to year. It is this progress that matters–and for which educators should be held accountable.

Furthermore, this “value-added” approach is fair. It can check on progress of struggling students anywhere in the state–whether in Lowell or in Lexington. I’ve talked to literally hundreds of teachers about testing, and they say this is an approach they can accept. One bonus of “value-added” accountability, then, would be the buy-in by educators that is a critical element of constructive change.

It’s not unfair to hold teachers accountable.

The question, though, still remains: What to do about schools that aren’t making progress–and don’t seem to know how to? In this, we run into a major roadblock. The Massachusetts tradition of local control holds that education decisions should be left as much as possible to local school districts. This attitude is reflected in the Legislature’s reluctance to provide adequate funding for a beefed-up Department of Education. But our experience over the last nine years–and the experience of many other reform-minded states as well–suggests that you can’t run an effective statewide reform program without a strong, professional, well-respected state education agency.

DOE currently has in place a program to intervene in failing schools, but its effort to do so is small. The department does preliminary assessments in roughly 12 schools a year and has done in-depth intervention in only four cases to date. Given the lack of progress in educating our neediest students, the department should have interven-tion/assistance teams in 20 to 30 districts each year. DOE should be authorized to withhold aid increases from struggling districts until, working co-operatively with the intervention team, they’ve developed turnaround plans that rise to the occasion.

To truly have an impact on failing schools, however, the department needs more than resources; it needs a point of view. Intervention/assistance teams need to go into these schools armed with a game plan for effective education. They should have a strong orientation toward the use of data, analysis, and research to guide instruction. They should look not simply to investing in individual teacher training, but in programs that help groups of teachers work together to improve their skills. Their goal should be more than simply to provide schools and districts with specific suggestions for improvement, but also to move teachers and administrators from the view of education as art to a more rigorous approach. Specifically, their intervention should be based on what we know about how to turn schools around–and, above all, how to teach students to read.

Turning schools around

Over the past year, I’ve had occasion to visit some truly exceptional schools, not only here in Massachusetts but in Pennsylvania and Alabama as well. I’ve found that out-standing schools have the following elements in common:

  • A principal who cares deeply about education improvement and knows how to be an educational leader;
  • Frequent assessment of individual students and small-group instruction that allows teachers to work with students on their individual problems;
  • A commitment to use research-based instruction in reading;
  • At least one teacher whose job it is to help other teachers improve their teaching;
  • Extensive investment in teacher professional development, with priority given to training experiences shared by several faculty members at the school;
  • A commitment by teachers at each grade level or in each subject to take collective responsibility for educating their students; and
  • Frequent contact with outside advisors who can keep the school on target, praising its successes and pointing out its weaknesses.

The Department of Education should incorporate these principles into a formalized prescription for education improvement–one that it is willing to impose on school districts if necessary. This will represent a break from the Massachusetts tradition of unfettered local control, but these principles still leave room for substantial local flexibility in how to implement them.

The leadership of a good principal is perhaps the single most important element in school turnaround. Unfortunately, nothing in the training most principals receive prepares them for this task. I believe the best way to help principals is by letting them meet outstanding principals, learn from their experiences, and in the process develop contacts with like-minded colleagues across the state.

The state of Alabama is following my recommendation to recognize 20 “Gold Star” schools, based in part on outstanding leadership demonstrated by the principal, with a reward of $100,000 a year in extra state funds. The schools can use this windfall to hire an additional reading specialist or assistant principal, with funds left over for buying more books or hiring substitutes to cover for teachers who visit nearby schools. In return, these schools will serve as training sites for principals from across the state who will come for one- or two-week visits to learn leadership from these masters.

Reading lessons

If we can’t teach children to read, we might as well give up on education reform altogether. After nine years of reform, 30 percent of white third-graders–and a devastating 66 percent of black students and 72 percent of Hispanic students–are not proficient readers. The results of the fourth-grade language arts test are even worse: 42 percent of white fourth-graders, three-quarters of black students, and four-fifths of Hispanic students are not proficient in language arts.

Roughly half our children will learn to read tolerably well no matter what approach their school takes to literacy instruction. Most (but not all) children from literate homes, whose parents read to them from the time they are a year or two old, fall into this category. But most children from low-income homes, and some from more comfortable backgrounds for whom language does not come naturally, need systematic instruction in phonics and language structure to achieve full literacy.

We know what that systematic instruction should be. After 30 years of research, it has become clear that teaching children to read consists of five key elements: phonemic awareness (recognition that words are made up of separate sounds), phonics (attaching sounds to letters and understanding the structure of language, including long and short vowels, syllables, plurals, etc.), fluency, vocabulary, and comprehension. All children benefit from at least some instruction in all five elements of reading, but the ideal mix will vary somewhat from one student to another.

We’ve all heard of the “language wars” between phonics and whole language proponents. But this oversimplifies a complex debate. The goal of any reading program is comprehension–students being able to read and understand text. By stressing techniques for improving comprehension and by encouraging schools to buy appealing books and foster a love of reading, the whole language movement has made an important contribution. I know of no one who disagrees with this.

However, there is not the same agreement on the desirability of systematic instruction in phonics, phonemic awareness, and language structure. Almost everyone pays lip service to phonics. But whole language purists teach it incidentally–that is, as an adjunct to reading, and not as a subject in its own right. Indeed, in the last month alone, two Massachusetts teachers in two separate schools have told me that their principals have ordered them not to teach phonics–in one case because the principal thought that such instruction would jeopardize a state grant. In contrast, I’ve spoken with scores of teach-ers, both in Massachusetts and all across Alabama, who see major gains when they introduce phonemic awareness and systematic phonics to low-income students and to those for whom reading does not come easily.

After nine years of reform and several billon dollars, 75 percent of children from (all too often low-income) minority homes in Massachusetts are not proficient readers. To do something about that, the state Department of Education is going to have to adopt and promote an unambiguous approach to reading instruction that is consistent with the research.

Almost everyone pays lip service to phonics.

The place to begin is DOE’s Bay State Readers program. Following a model developed in Alabama and brought to Massachusetts after my reports on it and visits by key lawmakers, Bay State Readers runs two-week summer institutes to train entire school faculties (17 in the summer of 2001) in the five key elements of reading instruction; it also provides funding for a reading specialist in each participating school. To date, the program has been hampered by staff that did not agree on these principles and by the department’s traditional reluctance to articulate a point of view on successful instruction. Staff turnover and the new federal Reading First grants have given the department the chance to address these five problems. But Bay State Readers will also need “training the trainer” programs to address the shortage of reading specialists who have hands-on experience with (and believe in) the five basic elements of reading instruction. Indeed, new federal law requires all kindergarten-through-third-grade teachers in the state to be trained in these principles.

No school or district should be required to participate in Bay State Readers. If schools can teach reading successfully, using their own methods, so much the better. But if we are to solve the literacy crisis for schoolchildren in Massachusetts, the DOE is going to have to make Bay State Readers one of its flagship programs. In four years, the Alabama initiative on which Bay State Readers is modeled has involved 424 schools and trained 17,000 teachers; we should move toward a program of similar size. Such an effort would cost perhaps $50 million a year –a small amount in relation to the extra $2 billion a year we’re spending on education reform. In many ways, such an effort will pay for itself. Successful Alabama schools have seen reductions of 50 to 70 percent in special education enrollment and discipline referrals (as well as increases of 200 to 300 percent in library circulation). Much of this can be covered with federal funds; the new Reading First program earmarks some $15 million per year for Massachusetts literacy programs based on the five elements of reading instruction.

The new federal education law requires states to measure and report on the progress of low-income, black, and Hispanic students in each school district. Barring major changes in the way we teach these youngsters, most schools in Massachusetts (and all across the country) will soon run afoul of the law’s requirements to raise the educational performance of disadvantaged students. Our failure to educate disadvantaged youngsters will become painfully clear.

To date, we’ve had an arm’s-length debate on education in Massachusetts. We discuss testing, incentives, consequences, “staying the course” –but not how to make schools better. And we’ve done so in a way that has alienated teachers and principals and made them less likely to accept outside help. We need to recognize that this approach hasn’t worked–and that it can’t work. Let’s now discuss how good schools work and what it takes to make every school a good school. Let’s invest in our teachers and our principals. And let’s understand that we’re not going to get constructive change unless we treat our educators as valued partners.

Edward Moscovitch is president of Cape Ann Economics, in Gloucester. He worked closely with the Massachusetts Business Alliance for Education in drafting the 1993 Education Reform Act and participated last spring in a Massachusetts Education Reform Review Commission study of the reform process. He has served as the outside evaluator for the Alabama Reading Initiative.