publisher colophon
George Farkas and L. Shane Hall - Can Title I Attain Its Goal? - Brookings Papers on Education Policy 2000 Brookings Papers on Education Policy 2000 (2000) 59-103

Can Title I Attain Its Goal?

George Farkas and L. Shane Hall

[Comment by Chester E. Finn Jr.]
[Comment by Douglas Carnine and Hans Meeder]

In a statement released on April 1, 1965, President Lyndon B. Johnson used the following words to argue for passage of the Elementary and Secondary Education Act (ESEA) of 1965:

This bill has a simple purpose: To improve the education of young Americans.... How many young lives have been wasted; how many families now live in misery; how much talent has the Nation lost; because we have failed to give all our people a chance to learn.... This bill represents a national determination that this shall no longer be true. Poverty will no longer be a bar to learning, and learning shall offer an escape from poverty.... For this truly is the key which can unlock the door to a great society.

More than thirty years and $118 billion later, two national evaluation studies have concluded that these goals have not been met. 1 The skill gap in reading, writing, and mathematics has not been closed between, on the one hand, children from low-income households--often African American or Hispanic and attending central city or rural school systems--and, on the other hand, middle-class children--often Anglo and attending suburban school systems.

This comes as no surprise. The federal government's more than thirty-year attempt to solve the problems of poverty--in particular, the diminished life chances of children from low-income households--has been largely unsuccessful. Title I of the ESEA is, and has been, the most heavily funded program in this area. At $8.3 billion for 1999 alone, this program is funded at approximately twice the level of the better-known Head Start program. The failure to win the War on Poverty is largely attributable to the failure of these two programs to achieve their goals: the school success of low-income children. [End Page 59]

Our focus here is Title I--federal aid to schools with large numbers or percentages of poor children, historically targeted on boosting the performance of these children in grades one through three and usually working most intensely on reading. The ineffectiveness of this program occurs for many of the same reasons that cause the more general ineffectiveness of low-income schools and districts. The funding creates a separate bureaucracy within the larger central district bureaucracy, and the education officials do not have the same focus on getting the children reading at grade level as do the teachers in the schools.

The teachers themselves are poorly supported. Colleges of education teach them little of relevance to assisting low-income students to read. Once they begin teaching, they are typically isolated with a classroom of children who are cognitively and behaviorally unprepared to do the work prescribed by the curriculum. Training and supervision from administrators are often minimal.

Title I has functioned as a small add-on to this general situation. Deeply embedded in the existing culture of schools, it has funded additional teachers and aides who work with the lowest performing children in small groups or one-to-one. These teachers and aides often have little special training, or when they do have such training, it is based on the whole language model, which has been popular in education schools for many years. Thus, they are typically untrained in those research-based (phonics) models that are most effective, particularly for low-income children.

Unfortunately, the 1994 reauthorization of the ESEA introduced new problems and exacerbated old ones. It diluted what was already a weak intervention for the lowest performing, low-income children. What was once a program directed almost exclusively toward grades one through three, when it is still possible to successfully assist low performers, was expanded to cover grades four through twelve, so that now only 37 percent of participants are in grades one through three. What was once a program providing personalized, add-on services to the lowest performers in low-income schools was altered to enormously increase the number of schoolwide programs. The result has been to decrease add-on services provided to the lowest performers and to encourage school administrators to use the funds to displace local funds; that is, to spend the Title I money on goods and services (computers, copying machines, furniture, classroom teacher salaries) that would have been paid for with local funds. [End Page 60] As a result, the population of Title I participants has shifted, from an almost exclusive focus on students receiving additional, personalized assistance to include every student enrolled in more than sixteen thousand schools with schoolwide Title I programs. Consequently, the Department of Education in 1999 counts 11 million program participants, where before reauthorization in 1994 there were only 6 million. And a program that for many years has been criticized as being merely a funding stream, with too little impact on the neediest children, has become even more incoherent and minimal in impact.

This is a dire situation. To even begin to correct it, the damage done by the 1994 reauthorization must be undone, while the more long-standing problems that have plagued the program for the past thirty-three years must be corrected. Accordingly, we have adopted the following strategies: First, see the problem in its most fundamental terms. That is, begin with the nature and determinants of the cognitive skill gap between low- and middle-income children as they enter and age through their schooling years. Next, focus on the most cost-effective practices for bringing these children up to a middle-class performance level. Be fearless in describing the current state of instruction under Title I and the extent to which it differs from these practices. Finally, be bold in suggesting how the program might be improved.

The Problem of School Achievement for Low-Income Children

In all societies at all times, the children of lower class parents (parents with the lowest levels of education and income) have begun life at a disadvantage. Central to this is their home experience during the preschool years: imitation of nonstandard speech patterns, too little conversation with adults, weak vocabulary development, too little experience with books, too little practice using language to express complex ideas, little or no instruction and practice with phonological awareness and other prereading skills such as learning the names of the letters and the sounds they make. Perhaps most damaging is the immaturity that many low-income children bring to first grade. They often come to school unready to sit still, pay attention to the teacher and the lesson, and do their own work. A first-grade teacher with a class of twenty or more low-income children, more than half of whom are at this level of immaturity, typically [End Page 61] finds it difficult-to-impossible to get them to concentrate on the assigned work and put forth the effort and time-on-task necessary to master the curriculum. This difficulty is further compounded by two instructional problems. First, the children are already behind in their skills. And second, the teacher is not typically using the most effective, research-based curriculum and instructional techniques.

As a consequence, children from low-income households reach the end of the first grade seriously behind in their schoolwork and already well into a cycle of failure in which their skills are below the level demanded by the curriculum, and their self-esteem, willingness to try, and time-on-task are inadequate to succeed at the assigned tasks. 2 They then fall even further behind during the summer. 3

By the beginning of second grade, the situation of children who are receiving negative feedback from this process, and are not progressing at grade level, worsens dramatically. The second-grade curriculum assumes that students can read acceptably, and the reading level expected increases substantially during this year. The teacher has little choice in moving the class on to higher reading levels, because many other and related skills must be mastered: spelling and capitalization, punctuation, composing and writing essays and stories, and (sometimes) cursive writing. In addition, the students must be moving forward in their mathematics, social studies, science, music, art, and other subject matter instruction. All of these require substantial time-on-task, and many are themselves dependent upon the student's ability to read and write. Inevitably, children who begin second grade below the level expected by the curriculum, who are already discouraged about their inability to master the required material, and who receive less assistance from their parents than do middle-class children fall further behind during second grade. And then, low-income children fall even further behind middle-class children during the summer break.

The pattern repeats again in third grade. And the third-grade curriculum is the last to include substantial amounts of basic skills instruction. By now, the assigned reading is demanding in terms of sophisticated vocabulary and reading comprehension. Reading is not to be done simply line by line. Instead, students are expected to keep the main theme of the piece in mind while observing the development of subthemes. Reading must be fluent, as must the student's ability to compose and write essays and stories. Cursive writing must also be mastered. As noted by Nancy Madden and others, "Disadvantaged third graders who have failed [End Page 62] a grade or who are reading significantly below grade level are very unlikely to graduate from high school ... and will experience difficulties throughout their school careers." 4

By fourth grade, the curriculum and teacher focus are no longer on basic skills, such as learning to read. Instead, students are expected to be reading to learn detailed subject matter in science, social studies, and so forth. Language arts and mathematics assume the more basic skills and advance rapidly through more sophisticated material. Yet, the majority of low-income fourth graders read below grade level. 5 At this point basic skills instruction is reduced, yet the curricular shift to higher order skills is a further obstacle for low-income children whose home life and school experience have provided weak vocabulary development and little experience in extended conversations using abstract concepts. 6 As a consequence, these students become essentially lost to the system. Whether or not they are placed in special education or other remedial programs, they never catch up.

To summarize this process, low-income and minority-group children begin first grade approximately one instructional year below middle-class children. As measured by instructional time (the time required to teach the curricular material that must be made up), the distribution of low-income childrens' skills shifts leftward relative to the national norms, so that by twelfth grade it is at the eighth-grade level. 7 Narrowing and ultimately eliminating this gap has long been the Title I goal.

What Could Be Done under Ideal Circumstances?

At least half the twelfth-grade gap could be eliminated if the first-grade gap were eliminated. 8 Thus a natural demarcation seems to exist between actions for children up to kindergarten versus those for first through twelfth grade.

Preschool through Kindergarten

Because too many low-income children enter first grade lacking the cognitive and behavioral skills and habits to do the assigned work, this problem must be attacked directly in the preschool and kindergarten years.

First, the oral English language skills and early alphabetic and phonemic awareness skills of these children must be increased during their preschool years. Perhaps most easily implemented and most valuable [End Page 63] would be to improve the instruction provided by Head Start and similar programs. At present, Head Start has no curriculum whatever. Meanwhile, this federal program--with no issues of local control--is nonetheless totally decentralized and sets no standards for learning. Preschool and Head Start staff too often use the excuse that the children are not developmentally ready to avoid teaching them the very skills that middle-class parents provide as a matter of course. Yet research shows that children aged three to six are able to learn these skills and must be taught these skills if they are to succeed in school later. 9 Appropriately providing this instruction to low-income children requires strong skills on the instructor's part and is absolutely essential if these children are to be ready for elementary school.

Other desirable interventions include programmatic efforts to assist parents to provide more learning experiences for their preschoolers. Low-income parents should be taught to converse more with their children, to read to them, to work with them on phonemic awareness, and to teach them their letters and sounds. 10 Such instruction would be enormously valuable but has rarely been successfully implemented on a large scale.

Second, it is crucial that all children attend a full-day kindergarten, that kindergarten teachers work hard to get their students ready to read by the beginning of first grade, and that they use research-based instruction as the basis for their efforts. Far too often the attitude is, "It's only kindergarten, we can't expect teacher and student to get that much done." Yet, as noted by Grover Whitehurst, "For low income children, first grade is bearing down on them like a freight train." 11 And children who are unused to sitting still and doing academic work, who are without practice in phonemic awareness (particularly if they speak a language other than English, black English, or another strong dialect), and who do not know their letters and sounds by the beginning of first grade are unlikely to end the year reading at grade level. First-grade teachers in low-income schools often complain that kindergarten teachers send them children who are not ready to learn to read. The low-income schools that have solved this problem typically have a strong phonics-based instructional program operating in kindergarten.

Finally, a strong, skills-based program during the summer between kindergarten and first grade could provide low-income children a crucial head start toward successful reading in first grade. But this will be the case only if the program focuses on those skills--phonemic awareness, letter and sound correspondences, sounding out skills with decodable [End Page 64] text, concentration in class, and good work habits--that research has shown to translate directly into reading success in first grade.

First through Twelfth Grade

Test-score data, classroom observation, and conversations with teachers lead to the same conclusion: Teachers in grades one through twelve in low-income schools are daily failing to adequately educate hundreds of thousands of children. The reality in each such classroom is a teacher with too little appropriate training and too little administrative support, attempting to work with twenty to thirty children, many of whom are unready to concentrate on schoolwork, unsupported at home, and unprepared intellectually to learn the assigned curriculum. If teachers were doctors and this many children had physical injuries that were not being successfully treated, the situation would be considered a national scandal, and aggressive efforts would be made to set matters right. And yet this instructional situation is even more damaging to these children because they are blamed for their failure. What can be done?

First, adopt a medical model. That is, allow zero tolerance for losing children instructionally. Every child could be approached on an individual basis with the goal of doing whatever is necessary to have that child performing at the level demanded by the curriculum. This would include assessment of phonemic awareness, letter and sound knowledge, word decoding skill, and the ability to sit still and concentrate at the beginning of first grade. Something like the case-management file maintained on the child's physical health by the child's doctor would be kept for each child in school. This file would focus on the child's healthy cognitive development and would be monitored to ensure that the child was progressing adequately through the curriculum.

Second, institute a schoolwide behavioral management system to assist teachers in gaining and retaining the discipline and order necessary for effective instruction. 12 This may be the most important prerequisite to the effective instruction of low-income children. Because these children have been inadequately socialized to the classroom and are cognitively unprepared for the curriculum, their teacher often spends so much time and effort maintaining order that time and effort on instruction are significantly reduced. The best of these teachers develop their own techniques for maintaining classroom discipline. Most involve some sort of [End Page 65] public record of the number of each child's good and bad behaviors for that day ("smily" and "frowny" faces on the corner of the blackboard next to each child's name, for example), combined with rewards and punishments at the end of the day. Nevertheless, even the most effective use of these classroom systems is enormously enhanced by a schoolwide program in which appropriate student behavior within a positive learning environment is promoted by such devices as a creed recited by the children each day; daily assemblies with a positive message; administrators who are constantly present, are supportive, and know all children by name; and a disciplinary system that directly links the classroom with the front office. Further, the absence of this schoolwide support makes it much more difficult, if not impossible, for teachers to maintain effective behavioral management within their classrooms.

Third, teacher training and the classroom curriculum must be built around research-based instructional methods. This includes the use of decodable text for beginning readers and an early instructional focus on the most effective techniques, including phonemic awareness and explicit phonics instruction. 13

Finally, appropriate and powerful interventions should be used early to bring low-performing children up to grade level. For children who are unable to concentrate, a medical workup and both behavioral and chemical therapy may be indicated. 14 For children who are falling behind the classroom instructional pace, and thus cannot participate successfully, intensive tutoring should be introduced to bring them back up to the classroom instructional level. This tutoring should use the most effective possible curriculum. 15 It should occur during the school year to catch children up to grade level, and during the summer to keep them from falling behind. It must then be available in future grades as an added boost for those who are once again beginning to fall behind. (Unfortunately, because low-income children benefit from weaker home inputs than do middle-class children, such falling behind by low-income children in higher grades is inevitable, given a national standard of curricular pace geared to middle-class student support levels.)

Why Has This Agenda Not Been Implemented?

This agenda would go a long way toward solving the schooling problems of low-income children. Certainly any serious attempt to solve [End Page 66] these problems must include many of these elements. Why, then, has this agenda not been implemented? Answering this question will provide two benefits to our analysis of Title I policies. First, the culture and institutions of schooling that have made Title I what it is will be explained. Second, the likely system responses to change in Title I policies will be discussed. Such understanding is a necessary basis for meaningful strategizing about the likely benefits and costs of alternative policy changes.

Schools Have Weak Technical Environments

Organizational theorists John W. Meyer, W. Richard Scott, and their associates characterized the public schools as operating within strong institutional but weak technical environments. Consequently, according to these researchers, "organizational attention is directed toward maintaining conformity with the socially standardized categories of the educational system, while little effort is expended in the control and coordination of instructional activities." 16 Specifically regarding curriculum and the delivery of instruction, detailed standards of instructional content or procedures are lacking; instructional innovations are poorly supported and come and go unsystematically, whereas innovations in noninstructional areas such as school health and cafeteria services are better supported and are more stable; teachers receive little useful training or colleague support to assist them with their instructional duties; and little attention is paid to coordinating instruction, so that, for example, individual teachers and classes are permitted and even encouraged to pursue unrelated or contradictory programs and students are often promoted to the next grade without having mastered the present grade's material. As for monitoring the results of instruction, the job performance of teachers and principals is rarely observed systematically or monitored in a meaningful way, and although teachers constantly test their students, opposition exists within districts to using standardized test data to evaluate the performance of individual teachers, schools, or districts. 17

Control by School District Administrative Elites

Analyzing the same behaviors from a rational choice perspective, Terry M. Moe agreed that the schools are too little concerned with the control and coordination of instructional activities but ascribed this to interest [End Page 67] group politics within and beyond school districts. He stated, "The structure of education ... has to do with who has power, with what their interests are, and with what kinds of structures they demand, design, and impose to see those interests pursued." 18 As presented in a longer work by John E. Chubb and Terry M. Moe, this rational choice and interest group politics view sees individual school actors pursuing power, self-interest, and rents via all the available techniques of interest group politics. 19

Institutionalist researchers in the same tradition as Meyer and Scott have begun to adopt a similar perspective on organizational environments that fail to adequately control and coordinate their technical functions. As stated by Walter W. Powell, "How are practices and structures perpetuated over time, particularly in circumstances where utilitarian calculations would suggest they are dysfunctional? Why are practices reproduced when superior options are available? Why are less-than-optimal arrangements sustained, even in the face of opposition?" The answer, he said, lies in the exercise of power. "Practices and structures often endure through the active efforts of those who benefit from them.... It is clear that elite intervention may play a critical role in institutional formation. And once established and in place, practices and programs are supported and promulgated by those organizations that benefit from prevailing conventions. In this way, elites may be both the architects and products of the rules and expectations they have helped devise." 20

This emphasis upon school district and education school elites and their use of power in the pursuit of self-interest via all means available, including "preservation of patterns of values[,] ... the selection of new recruits, the socialization of successors, and control over the conditions of incumbency," provides a necessary background for understanding the implementation of Title I in the nation's school districts. 21 The districts we have observed display intensely networked management structures, supporting almost constant strategic behavior by individuals and groups. Classroom teacher is the lowest status among professional staff. Advancement out of this status typically requires the support of the school's principal and assistant principal, but professional specialty groups (for example, the group of reading curriculum specialists, Title I teachers, special education teachers, bilingual education teachers, and so on) and ethnic or other affiliation groups (for example, the Hispanic Teachers Association, the African-American Teachers Association, and their community affiliates) are also a resource. 22 The higher one seeks to rise, the [End Page 68] more important are network connections. Every principal was once some other's assistant principal. And the real jump in power, prestige, and compensation is out of the schools and into the central administration, a step requiring patronage by individuals already there.

In addition to this vertical hierarchical structure is a horizontal structure of functional specialization, typically tied to funding streams. Not only does, say, the Curriculum Division have its own budget and management structure, but compensatory programs such as Title I, special education, and bilingual education account for even larger budgets, whose source is outside the district but whose expenditure is controlled by divisional managers within the district. When these vertical, horizontal, and other (such as ethnic group) structures are overlaid, the resulting grid provides almost unlimited potential for networked alliances and conflicts. For example, an assistant principal may be tapped for promotion into the central administration Reading Division, where his or her activities involve programs paid for with Title I funds but deployed and implemented by the staff of individual schools. Success requires the acquiescence and support of a wide variety of power centers. And each of these power centers has its own beliefs and culture, sustained by a network of personal affiliations, both within and outside the district. Only by understanding these can the deployment of resources and instructional practices for low-income children in the schools make sense.

The Reading Wars

In our visits to schools and districts over the past ten years, we have found the same conditions reported by Meyer and Scott at the beginning of the 1980s. 23 In particular, with regard to curriculum and instruction issues for low-income children, we have seen few serious and even fewer successful attempts to cope with the fact that low-income children enter school with nonstandard speech patterns, less mature behavior, and weaker oral vocabularies and knowledge of letters and sounds than do the middle-class children at whom the curriculum and teacher training are aimed.

Do most school districts begin working more intensively in kindergarten to help these students make up their deficits? No. Kindergarten is often optional or only half-day. And despite research to the contrary, many kindergarten teachers and district instructional specialists believe that the [End Page 69] children are "not developmentally ready." 24 In this case, do districts modify at least their first-grade curriculum to be less demanding, so as to better match the skills brought to school by these students? No. As one administrator said, "It is politically impossible to officially 'dumb-down' the curriculum for inner-city children." Instead, low-income children are presented with a curriculum they are unprepared for, taught by teachers who are untrained to cope with the students' lack of preparation.

Are the most effective instructional methods being used? No. In spring 1997 one of us attended a meeting at which the reading director of one of the nation's largest school districts told some two hundred of the district's elementary school principals that "because whole language instruction has won out over phonics in the education schools for the past 10 years, few to none of your teachers know how to properly teach reading." The principals nodded their agreement. This raises the question: Where are the schools today in the phonics versus whole language reading war?

Efforts have been made to impose a truce, based on a National Research Council (NRC) report arguing for "a balanced approach." 25 In practice, this means implementing the results of research demonstrating that explicit phonemic awareness and word decoding skill training are the most effective instructional techniques for beginning readers. The strength of these findings is overwhelming. 26 This research also demonstrates that, while some children do pick up the necessary skills without explicit phonics instruction, those suffering most from the absence of this instruction are low-income, second-language, and ethnic minority children. And yet, the NRC recommendation for explicit (not "embedded") phonics instruction for beginning readers is far from being fully implemented.

A networked group of researchers has sought to solidify and disseminate the findings in support of phonics instruction. Reid Lyon, the chief of reading research at the National Institute of Child Health and Human Development (NICHD) and a reading disabilities specialist, has led this effort. Coordinating work by more than one hundred researchers across fourteen research centers, he verified the results and then undertook a public campaign of dissemination. 27 At first, the greatest success occurred among the top administrators of a few state education systems. Bill Honig, the California superintendent of instruction from 1983 to 1993 who presided over the takeover of the state's reading instruction by whole language, noted that [End Page 70] since then state reading scores on the National Assessment of Educational Progress (NAEP) declined to next-to-last in the country. Honig thus declared himself a convert to phonics instruction for beginning readers. The California Department of Education moved to reduce whole language instruction and reintroduce phonics. Similar actions were taken by the Texas Department of Education, beginning in 1996.

Each side in the war has a social-movement-like commitment to a set of values and practices. On the phonics instruction side typically are scientifically oriented reading researchers--often with doctorates in cognitive psychology and appointments in psychology departments or medical schools and often affiliated with the NICHD research group--joined by some practitioners, including decisionmakers in some state education offices and local school districts, particularly in those states such as California and Texas that have had Republican governors. 28

On the whole language side are advocates more typically based in the teacher training and community outreach departments of education schools, as well as in administrative positions within the instructional and compensatory education program divisions of local school districts. Title I funds controlled by these district administrators and shared with their allies in local university-based schools of education have played a central role as the funding base for the dissemination and maintenance of whole language instruction in a great many school districts.

Whole language is a rhetorical stance that resonates deeply with the core values of many education professionals. That is, it assists them in finding a sense of human agency and meaning in their professional work. 29 Building on these feelings, much of the organizational and political work of disseminating the whole language approach has been undertaken by a compensatory education program called Reading Recovery. Invented in New Zealand and first disseminated in the United States via the Ohio State University School of Education, Reading Recovery advocates have aggressively promoted their program and whole language philosophy to the point where both flourish in thousands of school districts nationwide. To take one example, in 1993-94 Reading Recovery operated in more than five hundred districts in California alone. 30 Its success at least partially stems from its structure as a networked interest group with a social-movement-like mentality. 31

Reading Recovery is a compensatory education program that trains teachers to tutor at-risk first graders, one-to-one. It is based on a whole [End Page 71] language philosophy that has traditionally resisted explicit instruction in phonemic awareness and sounding-out skills, preferring instead to teach a set of word guessing strategies (based on pictures in the text and the initial consonant of words) that research has shown to be used only by the weakest readers. 32 Despite these counterproductive instructional strategies, the program has an elite reputation in school districts and completely controls reading curriculum, teacher training in reading, and the expenditure of Title I funds in thousands of school districts. How has this occurred?

The answer lies in a remarkable example of grass-roots organizational growth across the national field of school districts, utilizing Title I funds to build mutually advantageous ties between education school faculty and school district administrators. Each district selects teachers to be trained at the university, on a fee-for-service basis. Typically the recruits are among the best teachers, which maintains the program's elite status and inclines it toward success (to the extent that these teachers are not rendered less effective by the whole language practices they are taught). After their university training in Reading Recovery, the teachers return to the district as reading specialists, receive higher salaries, are removed from the classroom to work with low-performing children one-to-one, and many rise to become district administrators with responsibility for reading curriculum and compensatory education programs. Other teachers, after receiving advanced degrees, stay on the education school faculty to administer the program there. These transactions are typically paid for by Title I program funds under the control of district administrators. Because these administrators have often themselves been trained in Reading Recovery and have close personal ties to other district staff who were so trained, they typically ignore the high unit costs of the program and the questionable validity of its research results. 33

Despite research findings in opposition to its claims, whole language has come to dominate many districts' reading curricula. This has occurred partly because education school faculty who train reading teachers have become enamored of the whole language philosophy. But it has also been materially advanced by the financial base provided by Title I funds under the control of Reading Recovery-trained administrators and their allies. Even when phonics supporters in state education offices have moved to force local school districts to support "research-based instructional techniques," district administrators have used every trick, from simple lying [End Page 72] to refusing to apply for earmarked funds, to resist. 34 This attitude is typified by what one superintendent said while refusing to continue funding the much lower (less than one-tenth) cost, paraprofessional (trained, managed, and compensated college student) tutoring program we had successfully demonstrated in her district: "We have $1.5 million of Title I funds, and they will always be completely dedicated to funding our Reading Recovery teachers." 35

The Teaching Profession: A Comparison with Medicine

Because the daily implementation of Title I is strongly driven by the culture and structure of the teaching profession, any meaningful discussion of program reform requires an understanding of the profession itself. Using the comparative method, we contrast the structure and culture of teaching with that of medicine--the preeminent profession, which has demonstrated the ability to deliver high-quality services to the low-income community.

Superficially, at least, many similarities exist between school teaching and medicine. Both professions have government agencies and government funding directed toward them. Both have national lobbying organizations representing their interests. Each has its own professional school within many universities, where a faculty of experts conducts research to advance knowledge while training the next generation of researchers and practitioners. These faculty members belong to professional associations and contribute to a professional literature. Beginning doctors and teachers learn their trade by practicing it within the bureaucratic service delivery structures of hospitals and school districts, respectively, often in low-income areas. Both sets of practitioners must perform under difficult circumstances within these bureaucracies. As they establish themselves within their profession, each set of practitioners often migrates from low-income urban service delivery to a higher income clientele in the suburbs.

Given these structural similarities, what is it about these professions that leads to such dissimilar outcomes for the low-income individuals who are the recipients of their services? In particular, why is it that gunshot-wound victims receive excellent medical care at inner-city hospitals, while children with reading difficulties rarely show improvement in inner-city schools? [End Page 73]

At least part of the answer lies in the dissimilar culture of the two professions. Medicine is grounded in an outcomes-oriented, scientific worldview, where evidence of effectiveness is central. Its practitioners are trained to use state-of-the-art techniques to achieve the best possible result for each patient. Doctors-in-training are constantly presented with specific cases and asked what they would do. Best practice is demonstrated, and beginning doctors' efforts are critiqued. An unexpected negative outcome for any patient--whatever the experience level of the doctor in charge--is the immediate subject of discussion at a weekly meeting attended by practitioners.

By contrast, the education profession is not solidly based on a scientific outlook. Regular reading of the Educational Researcher (published by the American Educational Research Association) demonstrates that many education school faculty believe that the field, by its intrinsic nature, cannot and should not attempt to be built on a scientific basis.

Thus, most teachers, not surprisingly, report that their education school training taught them little that usefully applies to classroom teaching. As noted by Lyon: "Most report that they received little formal instruction in reading development and disorders during either undergraduate and/or graduate studies, with the average teacher completing only two reading courses. Surveys of teachers taking these courses indicate: (A) teachers rarely have the opportunity to observe professors demonstrate instructional reading methods with children; (B) coursework is superficial and typically unrelated to teaching practice; and (C) the supervision of student teaching and practicum experiences is fragmentary and inconsistent." 36

When a student completes medical school, he or she undertakes a closely monitored internship and residency. At each stage the focus is on optimum outcomes for the patient. But a beginning teacher in a low-income school district is typically placed in the classroom and left to fend for himself or herself. He or she receives little or no assistance or monitoring. It is expected that, like the experienced teachers, he or she will concentrate on controlling the behavior of a class where most of the children begin and end the year well below grade-level performance.

Title I Implementation

A great deal is known about the actions that should be taken to bring low-income children up to grade level. Also clear is the structure and [End Page 74] culture of the education profession, and the reasons that these actions have not been implemented. What sort of program has Title I legislation aimed to produce?

A Coherent Program or Merely a Funding Stream?

Title I has always been about money--the desire to provide extra financial assistance to schools and school districts with high concentrations of low-income children. However, as critics have noted, a revenue-sharing funding stream does not necessarily constitute a coherent program. The question arises: Has it been the intent of the legislation to create a coherent program?

A reading of the legislation shows that, from the beginning, the focus has been on providing a funding stream rather than a coherent program. Thus the original 1965 legislation provided funds to local educational agencies (LEAs) to hire staff, build facilities, purchase equipment, and cover "all appropriate costs" for any and all programs with the broad purpose of meeting the educational needs of disadvantaged children. In particular, the Senate report on the bill states that it is not the legislation's intention to prescribe specific programs or projects, believing that such decisions are better made at the local level. This has been the case for much of Title I's history, leading to comments such as those by James Berke and Michael Kirst, calling Title I an enormous success "as a fiscal device." 37 That is, for most of its history Title I has been a revenue-sharing program, in which the federal government passes funds to local school districts to spend on assisting the education of low-income children. This spending must occur within federal guidelines, which are loose enough to permit wide variation in program implementation. The result has been a program shaped, not by designers in Washington, but by the practices and culture of the education profession as it exists at the grass roots. Further, even though nationwide Title I expenditures are large in absolute terms, they constitute less than 10 percent of the expenditures of local school districts. Accordingly, districts have treated these funds as little more than a small add-on to whatever they were doing already.

Which Children Have Been Served?

Title I funds are allocated by formula grants through state education agencies to local education agencies. At least some funding has typically [End Page 75] been provided to almost all counties, to more than 90 percent of school districts, and to more than 70 percent of public elementary schools in the nation. 38 The U.S. Department of Education allocates funds to the county level, based primarily on census statistics regarding the numbers of poor children aged five to seventeen in those areas. State education agencies then suballocate the funds to school districts based on the number of children from low-income families in each district. More than 90 percent of the funds are allocated through basic grants (funds in proportion to the county's share of the nation's children in poverty), with approximately 10 percent coming through concentration grants focused on counties and school districts with either sixty-five hundred or more students from low-income households, or more than 15 percent of students from low-income households.

Each school district then allocates funds to individual schools based on the number of low-income students in each school. Districts may rank schools separately by grade span (there is a long tradition of targeting elementary schools), but they must serve schools with poverty rates of 75 percent or more before serving lower poverty schools, regardless of grade span.

Schools with very high poverty rates may also qualify for the status of "schoolwide program," which permits wide latitude regarding how the funds are spent. In particular, schoolwide programs are released from the requirement of providing add-on (usually pullout) services to the lowest performing children and are instead able to serve all children by using Title I funds to pay for teacher salaries or to purchase computers, books, copying machines, furniture--whatever will benefit the school as a whole. This raises the possibility of displacement; that is, using Title I funds to pay for goods and services that the district would have paid for with its own funds in the absence of Title I, thereby decreasing the add-on services provided by Title I to the lowest performing children. This is one reason that, before the 1994 reauthorization, the threshold for attaining this status was set high--a school had to have 75 percent or more of its students in poverty (on free or reduced-price lunch)--and there were consequently very few schoolwide programs in the nation.

However, using the logic that pullout programs are bad because they are stigmatizing and the student misses regular classroom instruction while they are out of the classroom, the 1994 reauthorization changed the regulations so as to facilitate and encourage the use of schoolwide [End Page 76] programs. For the 1995-96 school year, the threshold to qualify to be a schoolwide program decreased to 60 percent of students in poverty, and for the 1996-97 and subsequent school years the threshold decreased further to only 50 percent in poverty. The response has been dramatic. In 1988-89 only two hundred schools were in schoolwide status. By 1995 approximately five thousand were in this status, and by 1999 the number has risen to approximately sixteen thousand. 39 All other Title I schools are designated "targeted assistance" and must direct their Title I activities to the lowest performing children, who are provided add-on (typically pullout) services. Priority for service is usually determined by testing the students and giving highest priority to those with the lowest scores.

The 1994 reauthorization instituted a second dramatic shift in the program's target population. Districts were encouraged to direct their funding beyond its usual focus on grades one through three, so as to also serve grades four through twelve. The impact of this policy is indicated by the fact that a program that once served children in grades one through three almost exclusively now has only 37 percent of its participants in these grades. 40

What was for more than twenty-five years a program providing personal add-on instructional assistance to the lowest performing first to third graders in low-income schools has, in the past five years, dramatically reduced its services to these children. Further, these changes have made problematic the meaning of the term "program participant." This is because, with few schoolwide programs before the 1994 reauthorization, Title I program participation for most children referred to the actual receipt of personalized, small group or one-to-one, add-on instruction. Now, however, with the proliferation of schoolwide programs as well as extensive expenditures in grades above grade three, "program participants" include all students in schools with the schoolwide program designation. That is, the definition of a Title I program participant as a student receiving personal, customized add-on instructional assistance has effectively been eliminated. This is why the Department of Education reported annual program participation of 6 million students in the early 1990s, but 11 million students in 1999. 41 If the funds were spread thinly before, imagine the present situation.

Before reauthorization of Title I in 1994, most participants received personal add-on services. It is therefore this definition of program participation that applies to the data presented below from that time period. [End Page 77] Detailed information for subsequent time periods is not yet fully available. However, where more recent information is available (such as the Department of Education report Promising Results, Continuing Challenges, released in March 1999), we report that information also. However, note that for the post-reauthorization time period, the definition of "program participant" has become increasingly problematic.

Using NAEP data from 1994, we have calculated the effect of students' demographic characteristics on their rates of Title I program participation. The results are shown in table 1. Males participate at a slightly higher rate than females. Participation rates are highest in the West and lowest in the Northeast. Both African Americans and Hispanics participate at more than three times the rate of whites. The children of poorly educated parents have the highest participation rates. And rates are highest both in the central city and in rural areas. Despite the growth of schoolwide programs and the expansion of Title I to grades four through twelve, these participation rate patterns across demographic groups are similar to those reported for 1998 by the U.S. Department of Education. 42

Table 2 shows the distribution of Title I participation across schools in 1996, by type of school. Participation rates are highest among elementary schools; schools in a city, town, or rural area; and schools with a high percentage of poor children. Among elementary schools with 75 percent or more on free or reduced-price lunch, the program participation rate is 98 percent. Once again, these patterns are similar to those reported by the department for 1998.

Table 3 displays the NAEP reading and mathematics skill levels of participating and nonparticipating students. Available data permit us to calculate these for reading scores in 1994 and for mathematics scores in 1996.

The second column of table 3 reports program participation rates. Combining the mathematics data, which are for 1996, with the reading data for 1994 shows a significantly increased rate of participation by fourth graders during this two-year period, from 13.6 percent to 22.0 percent. For eighth graders the participation rate almost doubled during this period, from 6.5 percent to 11.7 percent.

The percentage distributions of these students across skill levels demonstrates what has long been known and was emphasized in the Prospects study, the largest and most definitive evaluation of Title I ever undertaken. Title I students are very low performing, and the program has [End Page 78] generally been unable to narrow the gap between their performance and that of nonparticipants. 43 Thus, among fourth graders, 80.4 percent of participants scored below basic in reading, while this was the case for only 33.1 percent of nonparticipants. Similarly in math, 69.2 percent of participants scored below basic, whereas only 26.4 percent of nonparticipants scored this low.

Consistent with the move toward schoolwide programs, the department's recent report compares NAEP average scores over time, separately for low- and high-poverty schools, instead of for individual participating and nonparticipating students. In reading, the primary focus of Title I, high-poverty schools averaged 190 in 1988, 184 in 1994, and 188 in 1996. Low-poverty schools averaged 217 in 1988, 220 in 1994, and 225 in 1996. If anything can be concluded from this, it is that the large performance [End Page 79] gap between low- and high-poverty schools widened significantly between 1988 and 1996, from 27 points to 37 points.

How Have the Funds Been Spent?

Table 4 shows the distribution of Title I funds across expenditure categories during 1990-91. 44 Although the specific distribution of funds has changed from year to year, it remains the case that the great majority of Title I funds are spent on the salaries of classroom teachers, Title I specialists, and instructional aides.

Specialists and teachers are at the top of the salary scale, with annual compensation (including fringes) often reaching $40,000 or more. Aides are at the bottom, typically paid on the order of $8-11 per hour. Thus, each Title I dollar buys significantly more aide than teacher instructional hours. Title I currently employs approximately fifty thousand aides and more than eighty thousand teachers and specialists, working in approximately forty-five thousand schools. 45 The remainder of Title I funds are [End Page 80] spent on program administration, counselors and other personnel, materials, and funding for specialized training and intervention programs.

Table 5 shows the instructional activities paid for by Title I funds during 1985-86 and 1990-91. These are the most recent detailed breakdowns available, and the distribution of activities likely has changed in recent years. In particular, since the 1994 reauthorization, pullout programs have been significantly reduced, and schoolwide programs have been enormously increased. However, our recent observations across a variety of schools and districts indicate that the instructional activities listed here continue to be the principal mechanisms by which Title I instruction is delivered. When it comes to working with students, not much is new under the sun.

What do these activities represent in practice? To the extent that they depart from ordinary classroom instruction, they involve small group and one-to-one instruction, as well as assistance in maintaining order [End Page 81] within a larger classroom. The greatest amount of this work occurs in elementary schools and focuses on reading instruction, although other grade levels and other subjects (particularly mathematics) are also covered. As for instructional techniques, they tend to be whatever is current practice in the particular district or school. Unfortunately, in recent years this has typically been whole language instruction, and our recent observations suggest that this continues to be the case. Thus, we are not surprised to learn that Title I instruction has been unable to narrow the performance gap for the children it serves. But does this mean that program services have been of no value whatsoever?

The Effectiveness of Title I Service

Certainly evaluation studies have reliably determined that Title I has failed to accomplish its mission of closing the achievement gap between [End Page 82] low- and middle-income students. As reported by authors of the most recent and best available evaluation: "After controlling for student, family, and school differences between Chapter 1 participants and nonparticipants, we still find that participants score lower than nonparticipants and that this gap in achievement is not closed over time." 46

But does this imply that Title I services have been completely ineffective, a total waste of more than $100 billion? It does not. As also noted by these authors: "This inability to find a compensatory effect of Chapter 1 does not mean the program was a failure. Limitations of this study do not allow us to determine whether Chapter 1 students would have been academically worse off without the assistance they received." 47

The problem is methodological. No one has undertaken a random assignment study, in which eligible students are randomly assigned to Title I or no Title I services. Instead, evaluation designs have been based on comparing outcomes for Title I students with outcomes for students who did not receive these services. At least the time period under study, 1991-93, was before the 1994 reauthorization, so most Title I students received personalized, add-on services. But at this time, Title I regulations required that the lowest performing students receive services. So the two groups--Title I participants and nonparticipants--are not comparable, and the Title I participants could have performed even lower in the absence of services. That is, the evaluation design was biased against finding an effect of the program.

Does this mean that Title I may be having a large positive effect on the students served? This is unlikely, for then the effect would have overwhelmed the bias and been visible in this and previous studies. But it is possible, perhaps even likely, that, under some circumstances, Title I has had a modest positive effect on the students served. Given that more than 100,000 teachers and aides, many of them people of good will and at least some skill, have worked in Title I for many years, it is difficult to imagine that their efforts have not brought some positive results for low-income children.

Displacement of Local Funds

Federal revenue sharing is typically accompanied by the displacement of local funds. Federal funds are used to pay for goods and services that [End Page 83] would have been paid for by local funds, so that the local governmental unit can save its own funds (for example, by avoiding a local tax increase). In the evaluation of manpower and training programs, the magnitude of such displacement has been one of the central empirical issues. 48 The substitution of federal funds for local funds that would otherwise have been spent occurs at a rate of about 50 percent. That is, only about half of federal monies granted to local government agencies are used to provide goods and services that would not have been provided anyway.

What about the displacement of local funds by Title I funds? In the legislation's earliest years, the issue went completely unaddressed. The 1965 ESEA included no provision to prohibit the use of federal funds to supplant state and local funds. This was soon understood to be a problem, however, as a 1969 Senate report on that year's amendments to ESEA stated that evidence indicated that Title I funds were not being properly used. An inquiry at the time by the federal Office of Education found cases of supplanting, excessive use of Title I funds for capital outlays, use of Title I funds to maintain segregated school systems, and use of the funds for general school needs. To address this problem, the 1969 amendments prohibited supplanting state and local funds with federal ones and stated that the intention was for Title I funds to be supplementary, allowing local school districts to provide goods and services that they would not otherwise be able to afford.

This prohibition of supplanting was reemphasized in the 1974 amendments, as was the statement that Title I funds are to be used solely for "excess costs," defined in the legislation as costs "directly attributable" to programs assisted under Title I that exceed average per pupil expenditures of a local school district (expenditures for any state and local services for special programs to serve the educationally disadvantaged, for bilingual education, or for special education for the disabled are excluded from the average).

More recently, the language of "no supplanting" has given way to "maintenance of effort." Thus, the 1978 amendments to the ESEA included a provision in which an LEA can receive Title I funds only if the state authority finds that the combined fiscal effort per student or aggregate expenditures are not less than the expenditures of the previous year. If they are, Title I funds are to be proportionally reduced. A waiver provision for "unforeseen circumstances" is included. In 1981 this maintenance [End Page 84] of effort provision was made more specific by requiring that the year's expenditures must be no less than 90 percent of those expenditures the previous year. This requirement has remained in place through the 1994 amendments to the ESEA.

The intent of the legislation has been to provide add-on services that would not be provided otherwise. Yet our observations in local school districts suggest that, as with other federal revenue-sharing programs, Title I displacement is a significant problem, appreciably reducing the add-on benefits received by the lowest performing children.

Schoolwide programs are particularly worrisome, because one of their main effects is to make it easier for administrators to spend Title I funds in this manner. Yet the use of these programs has exploded since they were targeted for increase in the 1994 ESEA reauthorization.

It is easy to see how Title I reformers believed that giving such flexibility to the schools most affected by neighborhood poverty would be a positive step, empowering principals to use the money to improve schools as a whole. Unfortunately, the reformers ignored the possibility that central administrators would direct, or otherwise induce, individual school administrators to spend the funds on goods and services that the district would otherwise have paid for out of its own funds, reducing or eliminating the add-on instructional services (particularly aides and tutors) that had previously been provided to the lowest performing children. But this is occurring. Substitution of federal for local funds should be expected to occur in school districts under constant financial pressure, where fungibility is the first instinct of many central administrators.

Such displacement has significantly reduced the impact of the program on low-income children. Even before the 1994 reauthorization dramatically increased schoolwide programs, the authors of the Prospects evaluation concluded as follows: "We cannot assume that funds or services labeled 'Chapter 1' represented additional resources, or even different services, from those schools or districts would have undertaken without Chapter 1 funds.... Chapter 1 funding was so long established, and so endemic to the U.S. public school system, that the allocation of Chapter 1 funds to a school district or school may not have necessarily meant that the resources for that district or school were increased on net." 49

And this was for the time period before the 1994 reauthorization enormously expanded displacement-prone schoolwide programs. Clearly the [End Page 85] displacement issue has been given far too little consideration in recent discussions of Title I.

The 1994 Reauthorization and Its Consequences

Since its inception in 1966, Title I has had a number of major flaws. It has not been a coherent program, but a funding stream that local districts have spent on whatever activities they were already engaged in. The additional services provided to each student have typically been so modest that the chances of significantly assisting that student to reach grade-level performance have been low. 50 Teachers and aides have been inadequately trained in research-based instructional methods, in particular, phonics-based reading instruction. The tendency toward displacement of funds has always been present.

The Department of Education's 1999 report Promising Results, Continuing Challenges provided an opportunity to see the results of the 1994 reauthorization. The good news is that a continuing emphasis has been placed on standards and accountability. When this is combined with the NRC report on reading instruction and the department's efforts to reference that report in its implementation of the Reading Excellence Act, a continued push is evident toward the use of research-based instructional techniques.

Unfortunately, much of the remaining news is bad. What was never a coherent program is even less coherent now. What were once minimal services delivered to each eligible student have been watered down even further. What was once a focus on grades one through three--the time when low-income children can still be saved from falling hopelessly behind grade level--has been decreased. What was once a mild temptation toward displacement of Title I funds has become an open invitation to use them in this manner.

Each of these outcomes resulted from policies deliberately pursued in the 1994 reauthorization. Policy analysts seemed to have had an inadequate understanding of how the program operates on the ground and of the most important issues to be addressed for the program to succeed.

National politics also played a role. Politics may have been, and continues to be, decisive in driving the program toward its current unfortunate state. Both Democrats and Republicans have seen political gains in [End Page 86] expanding the program to serve more grades in more schools with more flexibility for districts and principals to spend the funds as they see fit. The inevitable result has been that the neediest children with the most to gain from an intensive treatment--very low income first through third graders who are already falling behind--are having less and less of the funds directed toward them.

In particular, currently about one-third of the program participants are in grades one through three; the program used to focus almost exclusively on those grades. Furthermore, the program's budget now must serve almost twice as many participants. Why so many, each receiving so little? With the fiscal 1999 Title I appropriation at $8.3 billion, 11 million participants average $755 each. 51 But 6 million would average $1,383 each. And more to the point, if the bottom one-fifth of students in grades one through three were targeted, 2.1 million children would be served (assuming 3.5 million total at each grade level), which would allow $3,952 per child. If per child resources of this magnitude were spent effectively to provide add-on services to these children, the original goal of bringing these children up to grade level might be achieved.

But these services are reduced, dollar for dollar, when the funds are used to displace local district funds. Yet such actions are invited in schoolwide programs. And, between 1988 and 1998, schoolwide programs increased from two hundred to sixteen thousand. And in schoolwide programs, goods and services do not have to be targeted at add-on services for the lowest performers. Instead, furniture for the entire school, computers for the entire school, copiers for the entire school, books for the entire school, and so on can be purchased. What school district could resist such an opportunity to conserve its own funds? 52 Yet the department does almost nothing to police this issue, and no evaluation studies have attempted to estimate the magnitude of the increases in displacement, which have likely accompanied this recent increase in schoolwide Title I programs. 53

Why did the 1994 reauthorization take this direction, which has led to such a sharp discontinuity with the previous twenty-eight years of program implementation? The answer, we believe, is that Title I's failure to close the cognitive skill gap led to a desire for strong action and a mistaken analysis of the problem led to the wrong policies being pursued.

What was this mistaken analysis? It was based on two principles. First, pullout should be reduced or eliminated, because it is stigmatizing and the [End Page 87] pulled-out student misses in-class instruction. Instead, inclusion is to be fostered. Second, the flexibility of district administrators to spend Title I funds, by increasing schoolwide programs and extending Title I to grades four through twelve, should be increased. Flexibility seems to have been promoted because these analysts did not know how to improve the program after its admitted failure for so many years; they chose to "let all flowers bloom" by delegating responsibility to the district level. Such flexibility is so politically popular that it has become the mantra of both political parties. It is also beloved by the many entrepreneurs in the enormous grants economy of the education field, because each of these hopes to thereby increase their share of the pie. As noted by Stanley Pogrow, "Creating local standards in medicine is a form of malpractice." 54

Yet both pillars of the 1994 reauthorization are built on misunderstandings. Pullout is necessary for one-to-one instruction, and such instruction is the most effective way to raise the performance of low-performing students. 55 This is why the first thing middle-class parents do when their child is having difficulty in school is to hire a one-to-one tutor. Only with one-to-one tutoring can three vital instructional conditions be achieved: individual assessment followed by instruction on the student's current performance level (instead of teaching to a small group or class average); complete privacy within which to risk failure, combined with no other child being present to give the answer; and the presence of a caring and encouraging adult, able to detect and immediately correct the student's difficulties, whatever their source.

The reason Title I was ineffective in the past is that it involved pulling out groups of children, each with a different performance level and central instructional problem. These students were taught by teachers and aides using relatively ineffective, whole language techniques. Because the program was kept separate from the rest of the school, the children pulled out were the lowest performers, and Title I staff were unable to succeed with them, morale was low, and the children and program were stigmatized. The call for inclusion thus understandably arose. Yet a better plan would have been to employ a research-based (phonics) curriculum, provide Title I staff with adequate training and supervision, and have them apply this curriculum one-to-one. Only under these conditions might low-income, low-performing students be brought up to grade level.

As for stigmatization and missed class time, neither is a problem when the program is properly implemented. For example, the Reading One-to-One [End Page 88] program, the Howard Street program, the Book Buddies program, and others have demonstrated that when tutoring is one-to-one, using an effective curriculum and adequate training and management, students look forward to tutoring, and nontutored students beg to be included. 56 At program start-up, teachers are often concerned that pulled-out students will miss class time. But this concern is met in several ways. First, the teacher chooses when during the day he or she would like the student pulled out. Thus, scheduling can minimize the loss of instructional time. Second, the students tutored are typically the lowest performing in what is already a low-performing class. These students are far behind the demands of the curriculum, are often immature, and have attention deficit problems. They are unable to concentrate in a class of twenty to thirty students, and their presence greatly increases the teacher's difficulty in working with the other students. But if five tutors come to the classroom door and take out the five lowest performers for one-to-one instruction, the teacher is better able to succeed with the remaining students, and the pulled-out students receive intensive, one-to-one instruction designed to focus on what they need to help them regain full participation in the class. 57

The movement toward flexibility and the resulting increases in funds allocated to schoolwide programs and to grades four through twelve have done great damage to the chances of Title I achieving its goal. There are four reasons for this. First, these older children are typically so far behind grade level that it is almost impossible to catch up. Second, even less is known about successful intervention in grades four through twelve than about grades one through three.

Third, the institution of a schoolwide program almost guarantees a reduction in add-on services to the lowest performing children. The purpose of schoolwide programs is to free administrators from the demand to target add-on services to the lowest performing students. This is based on the idea that, in very low income schools, all students need help. Yet this is false. As noted by the authors of the Prospects study, "Differences in academic achievement within schools are far greater than average differences between schools.... Poor schools are not filled solely with low-achieving students.... There are good students in all schools, even in the most troubled places." 58 The lowest performing students in these schools need help, and what would help them the most is intensive personal attention, not a nebulous schoolwide program that involves spending [End Page 89] Title I money on computers, furniture, or even additional classroom teachers.

Fourth, middle and high school principals have little or no knowledge of effective compensatory programs on which to spend their newly acquired Title I funds. The obstacles to the creation of such programs are overwhelming. The difficulties of bringing low-performing elementary school children up to grade level are vastly greater for middle and high school students. They are now very far behind and have had six or more years of daily school failure. Their social-psychological defenses against trying hard at schoolwork are well developed. And their primary focus is elsewhere--on adult activities, which hold much more attractiveness than schoolwork. Little wonder that these principals either spend their Title I funds on completely ineffective programs or simply use the funds to purchase goods and services they would have purchased anyway with local funds. 59

The Department of Education's Direction for Title I

Promising Results, Continuing Challenges indicates the direction the Department of Education has charted for Title I. The continued emphasis on high standards and accountability is welcome. But the goal that "all children reach challenging standards" and the specific policies to achieve this are sadly out of touch with the reality of the children's current performance level and the extent to which the current pattern of Title I expenditure is significantly assisting them.

The reality is a Title I program in which, in 1994, 80 percent of all participating fourth graders had reading skills below the basic level. This massive level of failure continues to be widespread among children in high-poverty schools. Even this department report, with its claim of "promising results," acknowledges that, in 1998, 68 percent of all fourth graders in high-poverty schools have reading skills below the basic level. (Note that the department's 1998 figures are based on all students in high-poverty schools, even though the Prospects study reported that these schools have children who are doing fine. The department is including high performers in high-poverty schools, who have received no personally targeted services, among program participants.)

Beyond a general focus on "standards and accountability," the department's [End Page 90] recommendations come down to two. First, a general emphasis is placed on "staying the course" with the changes brought by the 1994 reauthorization. Second, a plan is put forth to reduce the number of Title I aides, so as to increase the number of Title I teachers. The reason given is that these aides "lacked educational credentials required to deliver high-quality instruction." 60 However, the issue is not credentials, it is skills and training. Having failed, over a thirty-year period, to put a serious training program in place for the aides, the department now wishes to fire them because they "lack credentials."

Most of the reading curriculum experts in this program were trained in, and still practice, whole language instructional techniques. Program funds also pay for equipment and materials, few of which are of much assistance to the lowest performing children. (Despite exaggerated claims, most low-performing elementary school children placed in front of computers play with them for a while and then lose interest. And like so much else in low-income schools, computer-assisted learning tends to be inadequately implemented.) Finally, some of the funds pay for special programs. But many of these are not effectively targeted on the lowest performing children and do not have the personnel, curriculum, training, and administrative support to be successfully implemented for these children. And, with the move to schoolwide programs, add-on direct services to the lowest performing children have decreased.

The department wants to stay the course with this program. Instead of training both teachers and aides in effective practices, the department plans to simply scapegoat the aides.

We have observed many Title I aides and teachers at work in the schools. In addition, one of us has provided in-service reading instruction training for all the Title I teachers and aides in one of the nation's larger cities. We have observed these teachers and aides to be heterogeneous in their skills and abilities. Certainly the teachers operate at higher skill levels than the aides, but they frequently cost three to four times as much. And when one of us trained both teachers and aides, he found that their knowledge of reading instruction was not all that dissimilar. In particular, the teachers knew almost nothing about such research-based phonics techniques as the use of decodable text and explicit instruction in phonological processing, blending and segmenting, and other word-attack skills. In our opinion, an aide who had been trained in these matters would be a better reading instructor than a teacher who had not been so trained. [End Page 91]

Further, the department's latest recommendation--to fire the aides--reflects the general lack of attention to costs in educational cost-benefit calculations. As with the proliferation of high-cost programs such as Reading Recovery, where a $50,000 per year teacher is tutoring at most eight children one-to-one, the department is ignoring the possibility that this $50,000 could pay for three aides and even more tutors, who could serve three or more times as many students on a personal basis. Because of the instructional advantages inherent in the one-to-one setting, less credentialled but properly trained and supervised aides or tutors can be effective in this setting. 61

Proposals for Structural Reform

Frustration with the cognitive performance being achieved by low-income children is widespread. As charter schools, voucher experiments, and city and state takeover of urban districts become commonplace, similar proposals have been suggested for the reform of Title I. The most dramatic of these goes to the heart of the issue of how effectively the funds are being spent by changing the recipients of the funds.

Change the Recipients of the Funds

Many of the central determinants of poor Title I performance--a focus on politics rather than performance; too little attention to skills, training, and management; the use of whole language instead of research-based methods; and a tendency to displacement of funds and other sorts of mismanagement--are directly traceable to the control of these funds by a separate bureaucracy within the central administration of local school districts. A tempting solution is to bypass this bureaucracy, sending the funds directly to the principals and teachers of eligible children, allowing the children's parents to decide how the funds should be spent, or sending the funds as a block grant to state education agencies.

A PORTABLE ENTITLEMENT. Perhaps the most radical suggestion is to fully fund Title I (increase the funding so that every low-income child receives a grant) and tie the money directly to each child, with parents having authority over how it is spent. As described by Chester Finn, Marci Kanstoroom, and Michael J. Petrilli: [End Page 92]

How to empower these parents? By insisting that Title I funds follow children to the school or education provider of their choice--be it the neighborhood school, a public school across town, a private school, a tutoring company, an after-school program, or a summer program. Instead of funding school districts, the federal government would fund children, much as Pell grants do for higher education. All poor kids would have funds "strapped to their backs." And we say ALL poor kids, because we favor the idea of expanding Title I funding to serve all eligible youngsters so long as the aid is truly portable, child by child and school by school. 62

Similarly, Diane Ravitch wrote, "Turn the key federal program for poor kids--Title I--into a portable entitlement, so that the money follows the child, like a college scholarship. Presently, federal money goes to the school district, where bureaucrats watch it, dispense it and find manifold ways to multiply their tasks and add to their staffs. As a portable entitlement, Title I's $8 billion would allow poor children to attend the school of their choice instead of being stuck in low-performing schools." 63

What would this mean in practice? Finn explains it as follows. 64 The idea is to change Title I from a district-based program to a child-based program, with each eligible child entitled to a certain amount of Title I aid no matter where he or she goes to school. The states would manage this program, and it would be their job to get the appropriate amount of Title I money into each school's bank account based on how many Title I students enrolled in the school.

The states would also set the limits for just how portable the funds would be. In particular, each state would decide whether the funds could go only to public schools or whether they could also go to private schools. Each state would decide whether the parents would be able to direct any portion of the funds to a nonschool provider. (An example would be a tutor or tutoring program, operating in an after-school program or in the evening. The Reading Excellence Act already provides funding for Tutorial Assistance Grants in which, in sufficiently low-income schools, parents have a choice of the after-school tutoring program--one provided by the school, and another by an outside vendor--they wish to instruct their child.)

This plan has a number of positive features. Even if a state does not permit parents to direct funds to a nonschool provider, local Title I bureaucracies would be reduced in size, perhaps even eliminated. Displacement [End Page 93] might also be reduced, as district administrators would have less opportunity to make Title I dollars fungible against goods and services they would otherwise have provided. 65 Overall, it would be an evolutionary step, in which principals and teachers in low-income schools would be empowered to make their own best choices as is now the case in charter schools.

FULL PARENTAL EMPOWERMENT. A more revolutionary change would follow if a state allows parents to direct funds to a nonschool provider. Imagine that each low-income parent has a "Title I expenditure account" for each of his or her children, from which approved expenditures can be allocated at will. These might include school-provided services during the school day or in an after-school program, as well as tutorial services provided in the home or neighborhood after school and during the summer. If this were combined with a program in which local high school students with good grades were trained as tutors to be certified and hired by the parents with Title I funds, perhaps under the supervision of the eligible student's school in an after-school program, benefits might be provided both to the high school and elementary school student. Furthermore, the funds would be kept in the neighborhood, without the overhead and profit charged by private sector tutoring firms. (Sylvan Learning Systems routinely pays its tutors $10 per hour while charging parents $30 per hour.)

Such a program would have multiple benefits. The high school students would be able to earn money to pay for college. Tutor training and the experience of tutoring would strengthen the tutor's reading skills, and assisting an elementary school student would enhance the tutor's self-esteem and perhaps even inspire the tutor to eventually enter the teaching profession.

The student being tutored would have a high school student to bond with and look up to. In neighborhoods where English is a second language, the tutors would be bilingual, so they would be ideally suited to assist elementary school students with the transition to English. These tutoring jobs could serve as an incentive and reward for high school students to work hard at their studies. The efforts of these tutors would go where they are needed the most--to increase the reading and homework assistance for low-income students in elementary school. And the funding would go where it is needed the most and where it can do the most good for the future--to high school students striving to escape the traps [End Page 94] present in low-income neighborhoods and better their future by attending college.

One danger in such a plan is that parents would use their control over Title I funds to defund the Title I programs in their local schools, or to escape these schools altogether. This possibility could be eliminated by sending the majority of each child's grant directly to the child's school, leaving only a modest sum of money under the parent's discretion. School-based efforts to involve low-income parents in their child's education, which are rarely successful under current conditions, might be undertaken much more energetically and successfully if these schools had a financial incentive to be in contact with the parents and to respond to their children's needs.

Perhaps a greater danger would be fraud and misuse by parents of their Title I account. Checks and balances would have to be installed. For these reasons, a plan based on (even partial) parental decisionmaking should be first tried as a limited experiment, accompanied by a rigorous evaluation.

Would such a plan immediately allow low-income schools to bring their students up to the performance level of middle-income children? Not if the families of low-income children still provide them with less prereading and other developmental skills during the preschool years and less home instruction and homework assistance during the school years. And not if the teachers of low-income students are still using less than fully effective instructional and classroom management techniques. But it would do something very important. By putting Title I funds in the hands of parents, whose motivation is solely to provide add-on services for their own child, this plan would completely eliminate the rampant displacement and reduction in personalized, add-on services to low-income, low-performing children that currently plagues Title I. For this reason alone, this is an attractive policy alternative.

BLOCK GRANTS TO STATES. An alternative method for disbursing Title I funds is as block grants to states. As with vouchers for parents, this has been suggested by those analysts--Chester E. Finn Jr. and Paul T. Hill, in particular--who wish to keep the funds away from district bureaucrats and who see the many different federal funding programs as distractions that have kept the schools from focusing on their main job.

Where, as in California and Texas, state-level education departments have, at least recently, promoted research-based instructional methods, [End Page 95] this might be a positive step, which would give state-level bureaucrats further leverage in their efforts to improve the instructional practices of local school districts. Many local school districts will, as they have done in California and Texas, promise to change their instructional methods to receive the money and then do nothing. Further, a single consolidated state grant of revenue sharing from the federal government would likely encourage the displacement of local funds at rates even higher than is currently the case. Nor will such a plan do anything to improve the home situations of low-income children. A possible result of consolidating federal education aid as a block grant to states would be to reduce targeted aid to low-income schools, at least in some states that are financially hard-pressed.

Schoolwide Reforms

Much attention is currently being given to schoolwide reform models. Many of these programs are organized around positive practices. Some of them, although far too few, have evidence of effectiveness. 66 But these programs are often expensive to implement and too often are not effective. Learning to implement successfully what is already known appears to offer a higher cost-benefit ratio than further invention and testing of new programs. An additional disturbing aspect exists to these schoolwide programs. Typical of the field of education, bringing in outside staff and spending a great deal of money seem to be necessary simply to induce schools to undertake management, training, and curriculum implementation that use common sense and best practices. Surely this mechanism is too inefficient to rely on for the reform of all the schools in the nation.

Reform the Profession

Finally, an overarching problem must be confronted: an education profession with too little focus on effective school management, a nonscientific culture, and entrenched elites wielding power under a system of perverse incentives, where true expertise is minimal, management and training are poor, and the most desirable promotion is the one out of the classroom. Successfully raising the achievement of low-income children cannot be done without significantly altering at least some of these conditions. [End Page 96]

The 1960s, 1970s, and 1980s saw relatively little progress in these matters. But when the Bush administration began, and the Clinton administration continued, a focus on national standards and accountability in education, some progress began to appear. As the Department of Education has sought to faithfully implement the Reading Excellence Act, with its focus on research-based methods of instruction, more progress has been made. This momentum of reform must continue and be extended to the schools of education as well as the state and local education agencies. This effort to change the culture of the education profession may ultimately yield the greatest gains in improving student achievement.

Some will say that such a cultural change is unlikely in the near future. Yet the volatility of the education profession--always swinging from one new reform to another, always pursuing the latest funding opportunity--indicates that change is possible. Now what is needed is the firm hand of scientific practice to stabilize the process with a focus on effective outcomes.

Paul Hill said, "It is now clear that every aspect of public K-12 education, from financing, school staffing, use of time, the authority of principals, instructional methods, and school accountability are all profoundly influenced by federal priorities." 67 If this is so, why not use this power to change the culture of the profession?

Current Department of Education staff are enforcing the provisions of the Reading Excellence Act to disseminate scientifically validated instructional techniques. The method is identical to that used by the National Science Foundation--empanel appropriate groups of scientific experts to be the judges for grant competitions. The department should audit Title I schools for effective training and management practices. New teachers should have an expert teacher in the classroom with them to provide hands-on training in effective practices. A minimum number of in-service training hours from appropriate experts should be required. A behavioral management program must be functioning in the school. It would not be difficult to train auditors who could assess Title I schools for the presence of these activities.

What Should Be Done?

Several policies would improve the effectiveness of Title I. These include reducing the displacement of local funds by federal funds, redirecting [End Page 97] services toward the lowest performing students, and building a more coherent program.

Reduce Displacement and Redirect Services to the Lowest Performers

First, reduce displacement, and redirect services to the lowest performing students. Children who are significantly behind the rest of the class in first grade, and who are completely lost to the system by third grade, require the most powerful possible intervention, including personalized instructional attention over an extended period, if they are to catch up to the rest of the class. The growth of schoolwide programs that displaced Title I funds away from add-on services to these children has been a disaster. Title I must be rededicated to provide add-on services to the lowest performing children in low-income schools. This could be accomplished by a simple mandate in the regulations or by changing the regulations back to once again reduce the number of schoolwide programs. This, however, does nothing to address the after-school and summer assistance needs of these children. Accordingly, some version of parental empowerment would allow the parent to contract with a tutor or other provider for such additional assistance for their child.

The reduction of services to the neediest by schoolwide programs has been widely observed by teachers in the schools. For example, consider this statement from the International Reading Association's recommendations for Title I reauthorization:

At its core, the idea of schoolwide programs was to offer increased instructional time for students who need it most. Unfortunately, in most cases this has not happened. Resources are often diverted from the core mission of enhancing the basic instructional program to other projects.... Schools receiving schoolwide funding should be required to demonstrate a commitment to seeing that the neediest students are provided instructional time and materials on appropriate levels, aligned with but above and beyond their regular classroom reading and/or math instruction. 68

Redirect the Program to Grades One through Three

The expansion of Title I to grades four through twelve has gone against everything that is known about the development of reading skill and cost-effective interventions to improve it. It has also, no doubt, greatly [End Page 98] increased the displacement of Title I funds. That only 37 percent of Title I participants are in grades one through three is unconscionable. This effect of the 1994 reauthorization must be rolled back, and the program retargeted on grades one through three.

Build a Coherent Program

Both Democrats and Republicans have pushed Title I toward decentralization and revenue sharing. The result has been enormous variability in instructional techniques, with little focus on the most effective practices.

The Department of Education should work directly to implement what is already known about best practices in curriculum, training, and management. Because in many cases what should be done is already known, the focus should be on making sure it is implemented. Such actions would surely have the highest cost-benefit ratio of any that could be taken. Unfortunately, the decentralized education system, and the small share of local educational costs contributed by the federal government, makes this difficult to achieve.

CURRICULUM. Two recent events have significantly altered the state of the reading wars. First, the NRC report sought to end the conflict by noting the advantages of both phonics instruction (particularly for beginning readers) and of whole language instruction (particularly for more advanced readers), arguing for a "balanced approach" using the best of each method. Second, Congress passed the Reading Excellence Act, which included language mandating the use of "scientifically validated" instructional techniques. The Department of Education has followed this up by stressing that scientifically validated methods will be a key criterion in judging the interstate competition for the approximately $240 million available under the act during fiscal 1999. This is one example of how the department can utilize revenue sharing to nudge the states and localities toward best curriculum practices.

How successful is such nudging likely to be, and would the curriculum be better off without it? The answer is that success rates are likely to be mixed. Trickle-down has a long way to go from the department to individual classrooms, and along the way are many stops, some manned by curriculum experts with a long-standing antipathy to phonics. However, the very existence of large numbers of such individuals, heavily entrenched in school districts and education schools, suggests the department [End Page 99] must continue waving the NRC report and arguing for the use of scientifically based instructional methods. Absent this, a push back toward whole language instruction may be inevitable.

TRAINING. The problem with teacher training is that it is not very good and there is too little of it. Teachers report that their education school training provided relatively little of value in the classroom. Following this training, they are typically placed into classrooms (often the least desirable ones--those with low-income children) and left on their own. Useful and extensive in-service training and in-class modeling of instruction by an experienced teacher are rare. Instead, beginning teaching is too often a sink-or-swim experience, and when the class is composed of low-income children, far too many new teachers find themselves sinking.

A portion of the Title I money should be targeted on training and should be spent on hands-on instruction in scientifically validated methods. Even better would be to require that the principal or other instructional leader spend at least a portion of each new teacher's first few months with him or her in the classroom, modeling effective classroom management and instructional technique. (This is common practice in the more effective schools serving low-income children.)

MANAGEMENT. Authors such as Chubb and Moe have argued that what is needed is to "get the incentives right," perhaps by moving toward privatizing the schools, for good management to follow as it does in the private sector--via survival of the fittest. However, the public schools are not likely to be privatized any time soon. Yet many schools serving low-income children are inadequately managed and operate within a district culture and structure that do relatively little to improve this situation.

What can be done to improve these conditions? Individual public schools can be provided with more control over their own resources and greater decisionmaking power in general. What the NRC report and the Department of Education's emphasis on this report did for reading can be done for school management. A few basic, desirable management actions should be identified and advertised widely. Then, Title I schools should be monitored to see that they are implemented. Because learning occurs in the classroom, these actions should be classroom-based and apply to all classes in the school. In each case an administrator or trainer and the teacher should be jointly responsible for [End Page 100] the outcomes. These might include (1) training on, and rating of, classroom management techniques that are effective in creating and maintaining the order and discipline necessary for learning, (2) provision of sufficient time-on-task and steady coverage of an appropriate curriculum in core subjects such as reading and mathematics, and (3) use of effective instructional techniques to get this curriculum across to the students. Each administrator and teacher pair should rate itself and be rated by the principal on each of these items, and Title I audits should focus on these instructional delivery issues instead of solely on financial accounting matters.

Further, effective schooling for low-income children requires starting them early and providing them with skilled teachers. Title I principals should be rated on their ability to provide effective reading and mathematics instruction in kindergarten and their ability to recruit, train, and retain good teachers, while dismissing poor ones.

PAID AND VOLUNTEER TUTORS. The Clinton administration's America Reads initiative has demonstrated that relatively large numbers of college students and community residents can be fielded at low unit cost to work on reading with low-income students, one-to-one. At least some research has shown that when these tutors are properly trained and managed, utilizing a serious instructional curriculum and decodable text books with beginning readers, significant gains can be attained. 69 Local high school students could also be used extensively as tutors for elementary school students in low-income neighborhoods. Ideally, these high school students could provide the home assistance that low-income parents are unable to provide.

RESEARCH AND DEVELOPMENT. Much of the culture and practice in the education field stands on a shaky scientific basis. Furthermore, scientific research played a key role in recent successful attempts to re-insert phonics into the practice of beginning reading instruction. Doesn't this argue for increased emphasis and expenditures on research and development if the field of education is to attain the capabilities and success rate of, for example, medicine?

The answer is yes and no. On the one hand, scientific knowledge is the basis of all technical progress, and too little such knowledge exists in the education field. On the other hand, instead of looking for even more new programs, perhaps those proven elements of programs should be implemented. For example, direct instruction's effectiveness was demonstrated [End Page 101] thirty years ago and continues to this day. But teachers were discouraged by what they perceived to be its rigid implementation style and took to whole language instead. 70 More generally, few disagree that effective schools would embody the curriculum, training, and management elements listed above.

Continued scientific research, particularly testing the effectiveness of alternative management strategies, would be valuable for effective program implementation. Such research should certainly be undertaken in a much more far-reaching and rigorous fashion than has been the case heretofore. But cognitive gains for low-income children are more likely to come from successfully implementing what is already known, than from the invention and testing of entirely new programs. This is much more a matter of engineering application than of basic research.

THE VIEW FROM THE CLASSROOM. A great many creative proposals have been suggested for reforming Title I. Their sometimes radical nature is an appropriate response to more than thirty years of failed efforts. Yet all this creativity must be held to the highest and most difficult standard: What will be the effect on the classroom and the students? The school system and the education profession function relatively poorly for low-income children. Changes could be instituted that make low-income children, their teachers, and their families even worse off than they are now. As Chester Finn warned, "First, do no harm." 71

To help low-income children succeed, their teachers must be engaged. It is therefore useful to ask, what is the view from the classroom? Given the opportunity, how would the teachers currently on the front line, in the low-income classrooms, like to see the Title I money spent?

The teachers we have spoken with might respond as follows. Yes, show me the most effective instructional techniques. Yes, assist me to implement a behavioral management program that works. Yes, start the children off earlier and provide them better preschool instruction, so that they come to my class better prepared to do the work. Yes, provide the students with a tutor at home if their parents cannot help with the homework. Yes, provide me with the assistance (aides or tutors, both in the class and pullout) to help the lowest performing children catch up while allowing me to work more intensively with a smaller group of children who are ready to learn the assigned curriculum.

THE VIEW FROM THE LOW-INCOME HOUSEHOLD. As the preponderance of research has shown, the learning deficits of low-income children [End Page 102] begin and end in the household. Shoring up this household by providing parents with vouchers that they can spend on after-school, evening, and summer tutoring for their children is the strongest action that could be taken to improve this situation. Is there any doubt about how these parents would respond if they were asked whether they wanted this opportunity?


The many difficulties standing in the way of increased school achievement by low-income children suggest the necessity of undertaking mixed strategies. In particular, Title I reauthorization should include the following: (1) a plan to reduce displacement and to retarget Title I funds on providing add-on services to the lowest performers in grades one through three; (2) at least some funds sent directly to principals and teachers in low-income schools, to be used at their discretion to assist the education of at-risk students; (3) at least some funds sent as a Title I account to the parents of low-income students, so that they can hire tutors to assist students after school and during the summer (this should perhaps first be tried as a carefully evaluated experiment); (4) serious attempts to tie funding to demonstrable and audited improvements in the use of scientifically validated curriculum, training, and management; and (5) a continued and serious effort to improve the scientific basis and outcomes-orientation of schools of education, in particular, and the profession, in general.

Realistically, the chances that this particular turn in the five-year ESEA reauthorization cycle will single-handedly and forever fix the Title I program are slim to none. Instead, a commitment should be made to the long haul and to detailed implementation efforts to shore up the classrooms and homes in low-income neighborhoods to the greatest extent possible. Embarking on a serious and dramatic reform of Title I resembles the attempt to reform the Russian economy. It must be undertaken as a long-term process, with the expectation that there will be many twists and turns along the way. This undertaking must be accompanied by a serious effort to provide effective, personalized add-on services to the lowest performing low-income children, whose cognitive skill improvement is the reason for the program.


1. Michael J. Puma and others, Prospects: Final Report on Student Outcomes (Department of Education, Planning and Evaluation Service, Office of the Undersecretary, April 1997); and L. F. Carter, A Study of Compensatory and Elementary Education: The Sustaining Effects Study (Department of Education, Office of Program Evaluation, 1983).

2. Connie Juel, "Learning to Read and Write: A Longitudinal Study of Fifty-Four Children from First through Fourth Grade," Journal of Educational Psychology, vol. 80 (1988), pp. 437-47; Robert E. Slavin, Nancy L. Karweit, and Nancy A. Madden, Effective Programs for Students at Risk (Boston: Allyn and Bacon, 1989); Connie Juel, Learning to Read in One Elementary School (New York: Springer-Verlag, 1994); and George Farkas, Human Capital or Cultural Capital?: Ethnicity and Poverty Groups in an Urban School District (New York: Aldine de Gruyter, 1996).

3. Barbara Heyns, Summer Learning and the Effects of Schooling (New York: Academic Press, 1978); Doris R. Entwisle and Karl L. Alexander, "Winter Setback: The Racial Composition of Schools and Learning to Read," American Sociological Review, vol. 59 (June 1994), pp. 446-60; Doris R. Entwisle and Karl L. Alexander, "Summer Setback: Race, Poverty, School Composition, and Mathematics Achievement in the First Two Years of School," American Sociological Review, vol. 57 (February 1992), pp. 72-84; and Daniel M. O'Brien, "Family and School Effects on the Cognitive Growth of Minority and Disadvantaged Elementary School Students," paper presented at the Twentieth Annual Research Conference of the Association for Public Policy Analysis and Management, New York, November 8-10, 1998.

4. See Nancy Madden and others, "Success for All: Longitudinal Effects of a Restructuring Program for Inner-City Elementary Schools," American Educational Research Journal, vol. 30 (1993), pp. 123-48, especially p. 125.

5. Farkas, Human Capital or Cultural Capital?; and Christopher Jencks and Meredith Phillips, eds., The Black-White Test Score Gap (Brookings, 1998).

6. Stanley Pogrow, "Title I: Wrong Help at the Wrong Time," in M. Kanstoroom and C. Finn, eds., New Directions: Federal Education Policy in the Twenty-First Century (Washington: Thomas B. Fordham Foundation, March 1999), pp. 41-60.

7. Farkas, Human Capital or Cultural Capital?; Jencks and Phillips, The Black-White Test Score Gap; and Lawrence C. Stedman, "An Assessment of the Contemporary Debate over U.S. Achievement," in Diane Ravitch, ed., Brookings Papers on Education Policy 1998 (Brookings, 1998), pp. 53-85.

8. Jencks and Phillips, The Black-White Test Score Gap.

9. See Grover Whitehurst and C. J. Lonigan, "Child Development and Emergent Literacy," Child Development, vol. 68 (1998), pp. 848-72; Grover Whitehurst, "Getting Ready to Read: Family Influences in Context," paper presented at the Green Center Conference on Attaining Universal Literacy, University of Texas at Dallas, May 1998; and Barbara Foorman and others, "The Case for Early Reading Intervention," in Benita A. Blachman, ed., Foundations of Reading Acquisition and Dyslexia: Implications for Early Intervention (Mahwah, N.J.: Lawrence Erlbaum Associates, 1997), pp. 243-64.

10. Phonemic awareness is the ability to hear and manipulate the distinct sounds in spoken language. Skills include identifying beginning and ending sounds, blending and segmenting sounds in spoken words, dropping sounds and pronouncing the remaining sounds, swapping sounds, and rhyming.

11. Whitehurst, "Getting Ready to Read."

12. Anthony S. Bryk, Valerie E. Lee, and Peter B. Holland, Catholic Schools and the Common Good (Harvard University Press, 1993); and Mary Zahn, "Head of the Class: Inside the Back-to-Basics World of the Marva Collins School," Milwaukee Journal Sentinel, February 21, 1999.

13. Marilyn Adams, Beginning to Read: Thinking and Learning about Print (MIT Press, 1990); and Bill Honig, Teaching Our Children to Read: The Role of Skills in a Comprehensive Reading Program (Thousand Oaks, Calif.: Corwin Press, 1996). See also Benita A. Blachman, ed., Foundations of Reading Acquisition and Dyslexia: Implications for Early Intervention, part IV: Implications for Intervention (Mahwah, N.J.: Lawrence Erlbaum Associates, 1997).

14. Russell A. Berkley and Joseph Biederman, "Toward a Broader Definition of the Age-of-Onset Criterion for Attention-Deficit Hyperactivity Disorder," Journal of the American Academy of Child and Adolescent Psychiatry, vol. 36 (1997), pp. 1204-10; and Joseph Biederman, "Are Stimulants Overprescribed for Children with Behavioral Problems?," Pediatrics News (August 1996).

15. For a presentation of the Reading One-to-One curriculum we have developed, see George Farkas, "Reading One-to-One: A Program Serving Large Numbers of Students While Still Achieving Strong Effects," in J. Crane, ed., Social Programs That Work (New York: Russell Sage Foundation Press, 1998), chapter 2; and George Farkas, Reading One-to-One Tutor's Manual (University of Texas at Dallas, Center for Education and Social Policy, 1998).

16. John W. Meyer, W. Richard Scott, and Terrence E. Deal, "Institutional and Technical Sources of Organizational Structure: Explaining the Structure of Educational Organizations," in John W. Meyer and W. Richard Scott, eds., Organizational Environments: Ritual and Rationality (Beverly Hills, Calif.: Sage Publications, 1983), pp. 45-70.

17. John W. Meyer and Brian Rowan, "The Structure of Educational Organizations," in John W. Meyer and W. Richard Scott, eds., Organizational Environments: Ritual and Rationality (Beverly Hills, Calif.: Sage Publications, 1983), pp. 71-98, quoted matter on p. 75; and Meyer, Scott, and Deal, "Institutional and Technical Sources of Organizational Structure," pp. 58-60.

18. Terry M. Moe, "Politics and the Theory of Organization," Journal of Law, Economics, and Organization, special issue, vol. 7 (1991), pp. 106-29.

19. John E. Chubb and Terry M. Moe, Politics, Markets, and America's Schools (Brookings, 1990).

20. Walter W. Powell, "Expanding the Scope of Institutional Analysis," in W. Powell and P. DiMaggio, eds., The New Institutionalism in Organizational Analysis (University of Chicago Press, 1991), pp. 190-91.

21. Powell, "Expanding the Scope of Institutional Analysis," p. 191.

22. Consider the following interchange when the Dallas school board was appointing an interim superintendent to take the place of a superintendent who had been sent to prison (Dallas Morning News, September 21 and 25, 1997): "When board President Kathleen Leos announced Dr. Hughey's name, members of the mostly black audience seemed confused. 'What's the race? If he's not African-American, we're definitely opposed,' shouted Thomas Muhammad, a black activist." "Some say the current battle between blacks and Hispanics is, at least in part, a fight over who will control jobs and contracts in a school district with more than 17,000 employees and a $967 million annual budget. 'It's about power and it's about jobs,' said Jesse Diaz, a vice president of a league of United Latin American Citizens chapter."

23. These observations are based on ten years of interaction with thousands of teachers and administrators across hundreds of schools in a great many school districts where we have presented or implemented the Reading One-to-One tutoring program. See Farkas, Human Capital or Cultural Capital?; George Farkas, "Ten Propositions about Schooling, the Inheritance of Poverty, and Interventions to Reduce This Inheritance," Research in Social Problems and Public Policy, vol. 6 (1997), pp. 125-69; Farkas, "Reading One-to-One"; and Farkas and others, "Can All Children Learn to Read at Grade Level by the End of Third Grade?," in D. Vannoy and P. Dubeck, eds., Challenges for Work and Family in the Twenty-First Century (New York: Aldine de Gruyter, 1998).

24. See Adams, Beginning to Read; and Honig, Teaching Our Children to Read.

25. Catherine Snow and others, eds., Preventing Reading Difficulties in Young Children (Washington: National Academy Press, 1998).

26. See the literature cited in Adams, Beginning to Read; Sonya Simons, Vera Woloshyn, and Michael Pressley, eds., "Special Issue: The Scientific Evaluation of the Whole-Language Approach to Literacy Development," Educational Psychologist, vol. 29, no. 4 (Fall 1994); Honig, Teaching Our Children to Read; and Foorman and others, "The Case for Early Reading Intervention."

27. For further explanation, see "Focus: Reading Recovery, Preventing Reading Failure," Effective School Practices, special issue, vol. 15, no. 3 (Summer 1996).

28. These gross characterizations are violated in particular cases. Nevertheless, they represent real tendencies in the alignment of forces in the reading wars.

29. Further examination of the perennial appeal to educators of such constructivist ideologies of instruction would be useful. Examples include the cross-national adoption of whole language (for example, it dominates reading instruction in Mexico despite the strong phonetic basis of Spanish) and the popularity of constructivist mathematics instruction, which, despite its damaging neglect of drill to achieve automaticity in addition and multiplication, is strongly advocated by many mathematics instructional specialists in education schools and school districts.

30. Bonnie Grossen, Gail Coulter, and Barbara Ruggles, "Reading Recovery: An Evaluation of Benefits and Costs," in "Focus: Reading Recovery, Preventing Reading Failure," Effective School Practices, special issue, vol. 15, no. 3 (Summer 1996), pp. 6-24.

31. When we have discussed Reading Recovery practices with school district people over the past ten years, many have volunteered, "They're like religion." In particular, Reading Recovery practices include the following: instructional techniques are passed on by a shared experiential training, with the belief that only this training makes a true practitioner; beliefs are based on faith instead of scientific research--these include the belief that intervention can be restricted to first grade and that, once "discontinued," the student will not require further remedial assistance; an unwillingness to release data for secondary analysis by others; and the use of nonscientific techniques in much of their research, including selecting "those who can profit" for tutoring, and dropping from the study sample those who did not succeed in the program.

32. "Focus: Reading Recovery, Preventing Reading Failure"; and Honig, Teaching Our Children to Read.

33. With the exception of those trained by Reading Recovery, many district curriculum directors have little formal training, or even experience, in the curriculum field. In the districts we have visited, it is common for the curriculum director to have been promoted from many years as a principal or other district administrator, based largely on personal ties to the superintendent. In many Texas school districts, an experienced Reading Recovery teacher typically earns in excess of $40,000 per year. With fringes this totals $50,000 or more. If he or she can tutor eight children one-to-one per year, the cost is $6,000 per child. At these prices only a small percentage of at-risk students can be served, but this fact is typically ignored by administrators, even when they are offered much lower cost alternatives (Farkas, Human Capital or Cultural Capital?, chapter 12). For studies questioning Reading Recovery's research practices, claims to effectiveness, and unit costs, see Yola Center, K. Wheldall, and L. Freeman, "Evaluating the Effectiveness of Reading Recovery: A Critique," Educational Psychology, vol. 12 (1992), pp. 263-73; Yola Center, K. Wheldhall, and L. Freeman, "An Experimental Evaluation of Reading Recovery," Reading Research Quarterly, vol. 30 (1995), pp. 240-63; Foorman and others, "The Case for Early Reading Intervention"; Patrick Groff, "Questions and Conclusions from a Discussion of Reading Recovery," in "Focus: Reading Recovery, Preventing Reading Failure," Effective School Practices, special issue, vol. 15, no. 3 (1996), pp. 25-29; Grossen, Coulter, and Ruggles, "Reading Recovery"; T. Rasinski, "On the Effects of Reading Recovery," Reading Research Quarterly, vol. 30 (1995), pp. 264-70; and T. Shanahan and R. Barr, "Reading Recovery: An Independent Evaluation of the Effects of an Early Instructional Program for At-Risk Learners," Reading Research Quarterly, vol. 30 (1995), pp. 958-96.

34. Kathleen Kennedy Manzo, "District Misuse of California Reading Funds Alleged," Education Week, May 7, 1997. We have observed similar activities on a widespread scale in Texas.

35. For further discussion see Farkas, Human Capital or Cultural Capital?, chapter 12.

36. Statement of Dr. G. Reid Lyon at Overview of Reading and Literacy Initiatives, hearings before the Senate Committee on Labor and Human Resources, April 28, 1998.

37. However, the most recent reauthorization in 1994 was more prescriptive of program operations than had previously been the case. James Berke and Michael Kirst, Federal Aid to Education: Who Benefits? Who Governs? (Lexington, Mass.: Lexington Press, 1972).

38. The sources of the numbers for the earlier 1990s are Martin Orland and Stephanie Stullich, "Financing Title I: Meeting the Twin Goals of Effective Resource Targeting and Beneficial Program Interventions," in Margaret Wong and Kenneth Wong, eds., Implementing School Reform: Practice and Policy Imperative (Philadelphia: Temple University Center for Research in Human Development and Education, 1997), chapter 1; and J. Moskowitz, S. Stullich, and B. Deng, Targeting, Formula, and Resource Allocation Issues: Focusing Federal Funds Where the Needs Are Greatest (Department of Education, 1993). However, Title I allocations have been in flux during the later 1990s, with an expansion to higher grades. See Promising Results, Continuing Challenges (Department of Education, Planning and Evaluation Service, Office of the Undersecretary, 1999).

39. Orland and Stullich, "Financing Title I"; and Promising Results, Continuing Challenges.

40. Orland and Stullich, "Financing Title I"; and Promising Results, Continuing Challenges.

41. Orland and Stullich, "Financing Title I"; and Promising Results, Continuing Challenges.

42. Orland and Stullich, "Financing Title I"; and Promising Results, Continuing Challenges.

43. Puma and others, Prospects.

44. This is from Mary Ann Millsap and others, The Chapter 1 Implementation Study, interim report, prepared by Abt Associates Inc. (Department of Education, 1992), which seems to be the most recent available data. These are the numbers cited in Department of Education, Office of the Undersecretary, Use of Federal Education Funds for Administrative Costs (Washington, 1998.)

45. Orland and Stullich, "Financing Title I"; Promising Results, Continuing Challenges; and Ralph Frammolino, "Title I's $118 Billion Fails to Close Gap," Los Angeles Times, January 17, 1999.

46. Puma and others, Prospects, p. 40.

47. Puma and others, Prospects, p. 55.

48. See George Johnson and James Tomola, "The Fiscal Substitution Effect of Alternative Approaches to Public Service Employment Policy," Journal of Human Resources, vol. 12 (Winter 1977), pp. 3-26; George Johnson, "The Labor Market Displacement Effect in the Analysis of Net Impact of Manpower Training Programs," in Farrell Bloch, ed., Evaluating Manpower Training Programs (JAI Press, 1979); George Farkas, D. Alton Smith, and Ernst W. Stromsdorfer, "The Youth Entitlement Demonstration: Subsidized Employment with a Schooling Requirement," Journal of Human Resources, vol. 18, no. 4 (Fall 1983), pp. 557-73; and Edward M. Gramlich and Harvey Galper, "State and Local Fiscal Behavior and Federal Grant Policy," Brookings Papers on Economic Activity, no. 1 (1973), pp. 15-58.

49. Puma and others, Prospects, p. 53.

50. The services provided to each student average fewer than thirty minutes per day.

51. Orland and Stullich, "Financing Title I"; and Promising Results, Continuing Challenges.

52. Anyone doubting that one of the major activities of district administrators is getting, spending, and shifting funds only needs to spend time with them, as we have done. For a description of how Dallas administrators used the 1994 reauthorization to pull large amounts of Title I moneys out of principals' budgets, see chapter 12 of Farkas, Human Capital or Cultural Capital? Concerning the general attitude of administrators to the use of funds, consider the case of the previous Dallas schools superintendent. Upon coming into office she announced she was bringing the FBI into the district to weed out what she had discovered to be long-standing corruption. The investigators convicted her of using district funds to purchase furniture for her home, and she was sent to prison. As for administrators' attitude toward accountability, the Austin assistant superintendent has just been indicted for altering student identification numbers so that these students' scores on the state accountability examination Texas Assessment of Academic Skills would be void, and some schools would receive higher rankings from the state.

53. Republicans used to be relatively vocal on this subject. It was one of their principal concerns during the many debates over the Comprehensive Employment and Training Act (CETA) and other job training programs during the 1970s and 1980s. Apparently their interest in decentralizing Title I and converting it into block grants has reduced their concern for this issue. For estimates of the magnitudes of displacement under these programs, see Farkas, Smith, and Stromsdorfer, "The Youth Entitlement Demonstration"; and Johnson and Tomola, "The Fiscal Substitution Effect of Alternative Approaches to Public Service Employment Policy."

54. Pogrow, "Title I."

55. Barbara Wasik and Robert Slavin, "Preventing Early Reading Failure with One-to-One Tutoring: A Review of Five Programs," Reading Research Quarterly, vol. 28 (1993), pp. 179-200; Marcia Invernizzi and others, "At-Risk Readers and Community Volunteers: A Three-Year Perspective," Scientific Studies of Reading, vol. 1, no. 3 (1997), pp. 277-300; and Farkas, "Reading One-to-One."

56. Farkas, Human Capital or Cultural Capital?; Farkas, "Reading One-to-One"; D. Morris and others, "Helping Low Readers in Grades 2 and 3: An After-School Volunteer Tutoring Program," Elementary School Journal, vol. 91 (1990), pp. 133-50; D. Morris, "First Steps: An Early Reading Intervention Program," ERIC Document Reproduction Service No. ED 388 956 (1995); Connie Juel, "Beginning Reading," in R. Barr and others, eds., Handbook of Reading Research, vol. 2 (New York: Longman, 1991), pp. 759-88; Connie Juel, "What Makes Literacy Tutoring Effective?," Reading Research Quarterly, vol. 31 (1996), pp. 268-89; Marcia Invernizzi and others, "A Community Volunteer Tutorial That Works," Reading Teacher, vol. 50 (1996); and Invernizzi and others, "At-Risk Readers and Community Volunteers."

57. During the eight years in which Reading One-to-One has tutored more than ten thousand students in more than one hundred schools across more than twenty districts, teachers have never had a problem with pullout once they saw the program in operation. See Farkas, Human Capital or Cultural Capital?; and Farkas, "Reading One-to-One."

58. Puma and others, Prospects, p. 66.

59. For one example, see Program Z discussed in chapter 12 of Farkas, Human Capital or Cultural Capital?

60. Orland and Stullich, "Financing Title I"; and Promising Results, Continuing Challenges, executive summary, p. 3.

61. Wasik and Slavin, "Preventing Early Reading Failure with One-to-One Tutoring"; and Farkas, "Reading One-to-One."

62. Chester Finn, Marci Kanstroom, and Michael J. Petrilli, "Overview: Thirty-Four Years of Dashed Hopes," in M. Kanstoroom and C. Finn., eds., New Directions: Federal Education Policy in the Twenty-First Century (Washington: Thomas B. Fordham Foundation, March 1999).

63. Diane Ravitch, "Clinton's School Plan Is a Good Start: Let's Go Further," Wall Street Journal, January 20, 1999.

64. Personal communication to George Farkas from Chester E. Finn Jr., April 1999.

65. However, the ability of these bureaucrats to influence principals out in the schools, whose career advancement they control, should never be underestimated. See chapter 12 of Farkas, Human Capital or Cultural Capital?

66. Rebecca Herman, An Educators' Guide to Schoolwide Reform (American Institutes for Research), at

67. Paul T. Hill, "Getting It Right the Eighth Time: Reinventing the Federal Role," in M. Kanstoroom and C. Finn, eds., New Directions: Federal Education Policy in the Twenty-First Century (Washington: Thomas B. Fordham Foundation, March 1999), p. 147.

68. "Title I Revision: IRA's Position," Reading Today (February/March 1999), p. 8 (emphasis in original).

69. Farkas "Reading One-to-One"; Invernizzi and others, "A Community Volunteer Tutorial That Works"; Invernizzi and others, "At-Risk Readers and Community Volunteers"; Juel, "Beginning Reading"; Juel, "What Makes Literacy Tutoring Effective?"; Morris and others, "Helping Low Readers in Grades 2 and 3"; and Morris, "First Steps."

70. Debra Viadero, "A Direct Challenge," Education Week (March 17, 1999), pp. 41-43.

71. Chester E. Finn Jr., "The Federal Role in Education Reform: First, Do No Harm," testimony prepared for delivery to the Senate Committee on the Budget, February 11, 1998.

Previous Article


Next Article


Additional Information

Print ISSN
Launched on MUSE
Open Access
Archive Status
Archived 2007
Back To Top

This website uses cookies to ensure you get the best experience on our website. Without cookies your experience may not be seamless.