when massachusetts passed its education reform law in 1993, everyone expected that the combination of greatly increased funding and high-stakes testing would produce dramatic gains in student achievement. Since then, state aid to public schools has tripled, with most of the new money going to inner-city schools, and the MCAS exams have been put in place. But schools with large numbers of low-income and minority students have made disappointingly little progress.

Massachusetts schools are as good as any in the country at educating children from well-off homes. And there’s been modest progress in the percent of high school students who eventually meet the MCAS graduation requirement. But there’s been virtually no progress in improving the basic skills of minority students in elementary grades. Last year, 65 percent of black third-graders were not proficient in reading. The figure five years ago was 64 percent. In fourth grade, 72 percent of black students fell short in reading last year, down only slightly from 76 percent five years ago; the comparable figures for math were 82 percent and 89 percent. The results for Hispanics are even worse.

The recently released 2007 MCAS results show some progress, with proficiency scores up by one percentage point for black third-graders and by six points for fourth-graders. For Hispanic third- and fourth-graders, the reading proficiency rates increased by 2 and 4 percentage points, respectively. Even with these gains, however, roughly two-thirds of black and Hispanic third- and fourth-grade students are still not proficient in English language arts.

The proposals on which the education reform law were based, which I helped put together, never addressed the question of exactly how schools would make the desired gains. Our assumption was that with substantially more funding, along with the carrot and stick of the MCAS exams, educators would have the tools to achieve those gains. It’s now apparent that there was a critical piece missing: a systematic way to help principals and teachers make the necessary changes. In effect, the reform law was based on the premise that teachers and principals knew what to do but for some reason weren’t doing it; embarrassing them through low MCAS scores, while decreasing their enrollments through school choice, would somehow get them in gear. This fundamental premise was mistaken.

Since 1993, I’ve visited scores of schools and talked with hundreds of teachers and principals. As evaluator for one of the largest school turnaround programs in the country—the Alabama Reading Initiative (ARI), which covers almost 1,000 schools —I’ve been asked to figure out why some schools have made spectacular progress with their low-income and minority students while others remain virtually unchanged. The ARI started with the assumption that all teachers needed was greater familiarity with reading pedagogy. But they quickly realized that teachers needed frequent, hands-on help from outside coaches. The ARI also found that change required active leadership from principals, who needed—and welcomed—assistance from outside coaches.

In my experience, the vast majority of teachers and principals care deeply about their students and work far harder than most people imagine. But they don’t know how to meet the needs of students (mostly minority and low-income students) who come to school with poorly developed vocabularies and language skills.

If we are to take our schools, and our students, to a higher level, we need a new model—one that recognizes that teachers and principals need, and want, specific blueprints for school change. They also need, and want, hands-on help in implementing such blueprints, from partners they like and respect.

BLUEPRINT FOR THE BAY STATE

Convinced that the most effective way to make this point politically was an actual demonstration, I joined Barbara Gardner as a co-founder of the Bay State Reading Institute. (Gardner was formerly assistant commissioner for literacy at the Department of Education and, before that, majority whip in the Massachusetts House of Representatives.) BSRI’s pedagogy closely follows the federal government’s Reading First program, which emphasizes individual student assessments. What sets us apart is that we offer our partner schools a specific blueprint for change, plus long-term coaching and hands-on training to help them get there. In so doing, we borrow heavily from the ARI, particularly on coaching for principals as well as for teachers.

We concentrate on reading because it is central to all other learning and because there’s more research on how children learn to read, how to assess their progress, and what to do when they fall behind. But success in reading requires a whole-school turnaround that includes small-group instruction, research-based curricula, hands-on leadership from the principal, and changes in the way the school day is organized. Once reading levels improve, these same ideas could be used for math and science.

The BSRI is a private nonprofit, but it receives most of its funding from the state ($3.7 million over the last three fiscal years), with additional support from the Boston Foundation. The funding is channeled through a contract with Middlesex Community College. Last year the BSRI partnered with eight schools in Brockton, Lynn, Leominster, Malden, Revere, Beverly, and Quaboag (a regional district encompassing Warren and West Brookfield). This fall we’re adding 10 more schools; our new districts are Attleboro, Randolph, Fitchburg, Orange, and Bourne.

Each of our first eight schools is a very different place today than it was a year ago. Students are assessed regularly, the assessments are used to group students and shape the instruction they receive, much more time is spent in small-group instruction, principals meet regularly with teachers to discuss the progress of individual students, and school-based coaches work actively with classroom teachers to help them improve instruction. In a recent report, our outside evaluators affirmed the critical importance of an outside partner, writing that “it is unlikely that these changes would have occurred without the direct involvement of the Bay State Reading Institute.”

Taken as a whole, during their first year with the program, these schools show fall-to-spring gains at each grade level. In the first-grade results (see chart below), 47.4 percent of students in our eight partner schools met national reading benchmarks in the fall of 2006, while 30.9 percent had dangerously low performance levels. But 65.3 percent met the spring benchmark (which is more challenging, since it reflects the gains students need to make during the school year), and only 13.2 percent still had dangerously low scores. The kindergarten level shows similar gains; the improvement in other grades is somewhat less.

DRILLING DOWN

As Massachusetts grapples with the issue of how to make good on the promise of education reform, particularly with respect to the persistent achievement gap facing low-income and minority students, it is important to appreciate in some detail the elements of an intensive reading improvement program such as the one we have developed. This is not to suggest that ours is the only approach to this challenge. Rather, it is to take readers inside the black box of school change, to lay out what is actually involved in making major change, and to underscore that our general framework—long-term partnerships with schools based on specific blueprints for change—is the best way to take schools to the next level. The elements of this framework include:

Frequent assessment of each child’s progress. This is central to any serious attempt at school turnaround. The MCAS can measure where students stand academically, but it is not well-suited for turnaround efforts with individual students. It is given only once a year, and its results are not available for several months. The assessments we use at BSRI are immediately available. A student can be tested in a couple of minutes, so a teacher can assess her struggling readers every week to make sure they are making gains.

MCAS is not enough for tracking individual students.

This type of assessment plays an enormous role in bringing teachers on board. They quickly grasp its potential for helping them see what individual students need and for making sure the interventions they’re using are working. When teachers see some students progress (and when they see their colleagues’ students making progress), most of them quickly buy in.

Small-group, differentiated instruction. We recommend an uninterrupted two-hour literacy block. No more than 20 to 30 minutes of this time are in whole-group instruction (that is, all students attending together to the teacher). The rest of the time, the class is broken into groups according to assessment results (which change over the course of the year). Each group meets with the teacher every day for instruction at its level in both phonics and comprehension. When not with the teacher, students are engaged in individual or group activities that are also differentiated to meet individual needs.

Small-group instruction along these lines is the only way to meet the needs of a class in which individual student reading ability can vary from pre-kindergarten to fourth or fifth grade. Properly administered, it keeps students’ minds constantly engaged. By contrast, when the class is in whole-group instruction and the teacher spends a couple of minutes on one student, most of the other students let their attention wander and instruction time is lost.

A comprehensive core curriculum. In recent years, there’s been enormous improvement in the materials available for literacy instruction. At BSRI, we use comprehensive curricula with research-based scope and sequence. The order in which consonants are introduced in kindergarten matters, for example, and it’s very helpful to have phonics and vocabulary development tied in with student readings. The best new curricula are geared to small-group instruction, with a common weekly theme for all groups (to facilitate whole-group read-alouds and vocabulary instruction) but with separate readings and instruction for students at each level. Readings include math, science, and social studies, so comprehension strategies and vocabulary development can be taught across the curriculum.

A good core curriculum lays out weekly learning goals.

A good core curriculum sets high expectations for student performance by laying out clearly what students are expected to learn each week. One of the most heartening developments during our first year is that teachers are finding that students are capable of progressing far faster than had been previously imagined.

Interventions and extra teaching time for strugglers. Many students—well over half in inner-city schools just starting a program along these lines—will not make adequate progress, let alone catch up if they’re behind, with only one daily session with the classroom teacher. They need extra instruction to reinforce the lessons of the day. We’ve helped our schools change their daily schedules and their use of Title 1 instructors (teachers paid with federal funds made available to schools with low-income student bodies) and special education teachers to provide regular daily interventions for struggling readers.

A key leadership role for the principal. The data meeting is a central tool for educational leadership by the principal. At least three times a year, the principal leads a discussion with each grade-level team to review student data. Each teacher’s students are discussed in turn. Praise comes first, which means celebrating progress and discussing what teachers did to help students move ahead. Then there’s a discussion—practical, not punitive—of students who are not where they need to be. The school reading coach may share excerpts from the last assessment as a way of identifying the problems of specific students. The principal asks the teacher what strategies he or she proposes, and everyone can contribute ideas. In the end, the principal and the teacher agree on approaches to try over the next few weeks. I sat in on such a meeting in one of our partner schools, and was pleased to see that the teachers were totally relaxed about the process, even though students in that grade had made only limited progress.

The data meeting is a powerful tool for moving those few teachers who don’t initially buy in to the new approach. When the principal turns and asks what strategies a teacher proposes for a particular struggling student, few will answer that they don’t plan any changes. The data meetings are most effective when combined with the principal’s frequent (daily, if possible) presence in the classroom, when he or she follows up to see if struggling students are making progress and if teachers are indeed implementing agreed-upon strategies.

A school-based reading coach. Teachers need constant help interpreting data, grouping students for instruction, learning how to do small-group instruction, differentiating instruction, and devising strategies for children who fail to make progress. BSRI requires its partner schools to hire (and pay for) a school-based reading coach whose primary job is to help classroom teachers improve their teaching.

The elements I’ve listed here all work together. One of the key lessons from the ARI is that a school must do all of these things if it’s to make major gains. That’s why an outside partner is so important. Like the rest of us, teachers and principals find it easiest to learn new techniques when someone shows them how to do it. Our reading coaches, for example, model lessons for school teachers. And our principal coaches help run data meetings until the school principals themselves are comfortable doing it.

It’s grossly unrealistic to expect principals to be at school and at evening meetings for 12 hours or more and then to go home and work out a strategy for change. But when we lay out for principals our specific model for school improvement, and make clear that our principal coach (a former principal herself) will work alongside them, we usually see an eagerness to give the effort a try.

TAKING IT TO SCALE

We work primarily with high-poverty schools that have had disappointing reading results, but we don’t force any school to work with us. Our relationship with a district starts with the superintendent; we explain our model for change, what we propose to do for the district and school, and what we expect in return. If the superintendent doesn’t share our vision, we go elsewhere. This means we’re only working where we’re wanted. It also means that there’s no conflict for the principal between working with us and meeting the superintendent’s goals.

As we succeed, we believe more schools will want to join us. To bring school improvement where it is most needed, the state needs a network of partners doing this work. Those change partners that develop a track record of success should be allowed to expand their portfolio of schools, while those that fail to deliver should be closed.

Typically, education programs are structured as grants. The Department of Education tells schools what it expects, and schools tell DOE what it wants to hear. Schools take the money and often use it to do whatever they intended to do in the first place. DOE may visit occasionally to try to keep them on the agreed-upon path, but the grant ends in a couple of years and everything returns to where it was before.

BSRI does not give money directly to our partner schools. Our coaches and the district’s central office, principal, and teachers agree jointly on the curriculum. Together they set specific goals for what they want to accomplish each year and decide on interventions and other educational materials, which we purchase. We provide training and coaching visits to the school (typically three days a month), and we reimburse schools for stipends when teachers attend our summer training. In return, we require that the district hire and pay for the school-based reading coach, and that he or she be supportive of our approach to reading instruction. We also require that the principal try to spend two or three hours each day on data meetings, class walkthroughs, and other education leadership activities.

This model works very well for us and for our partner schools. We’re all in it together, and there’s no disagreement about where we want to go.

DOE, by contrast, cannot be this kind of change agent. The adversarial nature of the MCAS exam and attendant penalties makes it hard for schools to accept DOE as a partner. Moreover, large bureaucracies cannot do this type of work well, and Massachusetts has a decades-old tradition of the department not involving itself in prescribing such activities at the school level. Thus, Massachusetts needs a network of smaller, independent change agents.

As a state, we’ll have to give careful thought as to how we bring new change agents into existence and how they’ll be funded and supervised. By their very nature, successful change agents are flexible, informal, and non-bureaucratic. The DOE has not yet demonstrated that it can work well with such organizations, so we’ll either need a major change in the way the department does business or a new home for this network—perhaps attached in some way to the state’s institutions of higher education, which have more of an entrepreneurial tradition.

Several years ago the book In Search of Excellence laid out the techniques for managing creative enterprises and maintaining the enthusiasm and loyalty of well-educated employees. BSRI draws heavily on these ideas: Praise works better than criticism; give as much authority as possible to front-line employees; manage by walking around; articulate long-term goals clearly but give all those involved as much say as possible in how best to reach them; set reasonable goals for each year; and pause to celebrate success. Sadly, these ideas have only rarely penetrated the world of elementary and secondary education.

The cost of our program (about $150,000 per school in the first year; half that thereafter, plus the cost to the school of their reading coach) is quite small in relation to the $8.5 billion or so Massachusetts spends on K-12 education and the roughly $3 million annual foundation budget for an elementary school of 400 students. Certainly, extra spending alone, absent the kind of structural change discussed here, will not lead to measurable gains in student performance. However, our approach does require reasonable class sizes, a school-based reading coach, and a corps of teachers who don’t have classrooms of their own but are instead available to provide extra instruction for struggling students.

In recent years, the state’s foundation budget—and the state funding that goes with it—has not kept pace with the actual increase in school costs. All across the Commonwealth, teachers are being laid off. Why that’s happened is a subject for another day, but, absent some combination of revenue gains and easing of cost pressures, these cutbacks will continue, and any kind of constructive change program, including ours, will soon be impossible.

Paul Reville, the new chairman of the state’s Board of Education, has said that MCAS alone is not enough; schools need help. He’s right. Our experience sheds light on the kind of help they need and provides a specific road map on how to get from here to there. This road map must be based on the frequent assessment of student progress, with continuous feedback between instruction and assessment, and it must come with hands-on coaching from an outside partner. If Massachusetts can put less emphasis on the punitive strategy embodied by MCAS and instead build a network of effective change agents, BSRI’s experience suggests that we’ll find willing and enthusiastic partners in the state’s teachers and principals.

Without such a program, the performance of our most disadvantaged youngsters will continue to stagnate. While BSRI still has a long way to go, our success in making major changes in our first year—working in schools with high percentages of low-income, minority, and non-English-speaking students—points the way to a broader strategy that can, at last, realize the vision of high academic performance by all of our students.

Edward Moscovitch, a former Massachusetts state budget director, is chairman of the Bay State Reading Institute and president of Cape Ann Economics.