Equity II—The Adequacy of Education
Since 1989, when the Kentucky Supreme Court took the dramatic and unprecedented step of declaring the entire state system of elementary and secondary education unconstitutional for failing to provide all children an adequate education, the concept of "adequacy" has moved to center stage in discussions of fairness in school finance systems. The Kentucky decision highlighted an important lesson from the educational reforms of the 1960s, 1970s, and 1980s: improving the fiscal capacity of schools may be necessary, but certainly isn't sufficient, to achieve equality of educational opportunity. To some, the idea of adequacy offers a promise of overcoming problems that previous equity-oriented reforms failed to remedy, and to do so by linking school finance decisions explicitly and centrally to the quality of education provided to America's schoolchildren. To others, movement toward adequacy signals a discouraging retreat in the long battle for basic fairness by threatening to perpetuate an education system that tolerates large disparities in the educational resources provided to them. To almost everyone, it is a concept that is still emerging and evolving: there is as yet no consensus on its meaning and only limited understanding about how and what would be required to achieve it.
Nonetheless, a growing number of state court decisions suggests that adequacy is becoming the new equity standard to which state school aid plans should be held. States without such court mandates are also seeking new answers to the old question of what it costs to provide an "adequate" or "basic" or "core" education (Education Commission of the States, 1997). The blossoming of the standards-based school reform movement has caused some to suggest that basing
school finance decisions on the educational needs of students, once ruled "judicially unmanageable" by several courts, may now be an achievable objective.
These developments highlight the importance of exploring what is known about what is called the adequacy movement: how adequacy is defined, what court imperatives are driving it, how states are responding to it, and what conceptual and technical challenges will have to be overcome in implementing adequacy-based school finance systems. A key challenge (addressed more fully in the next chapter) is whether current knowledge about how to improve student achievement is strong enough to make it possible to design a school finance system that fosters this objective.
POSSIBLE MEANINGS OF ADEQUACY
Despite the absence of consensus on the definition of educational adequacy, a good sense of how it differs from other approaches to equity can be obtained by examining it in terms of the five key definitional distinctions mentioned in Chapter 3.
Adequacy is exclusively focused on schoolchildren and does not embrace taxpayers as objects of concern. Conceptually, the unit of analysis could be the individual child or the school, but in practice (especially as litigators have so far employed it in school finance cases), it has been applied to school districts. If education funding becomes more school-based than district-based in the future, as some reformers urge, the unit of analysis for adequacy would probably shift to the school level as well.
One of the major differences often cited between adequacy and other definitions of equity is the former's emphasis on outputs and outcomes. Attorneys have tended to make a distinction between equity and adequacy, defining equity as input focused and adequacy as output focused. As Berne and Stiefel (1999) point out, however, in principle it is entirely possible for inputs, outputs, and outcomes to be equitable or inequitable and for inputs, outputs, and outcomes to be adequate or inadequate.1
Adequacy does indeed place far more emphasis on outputs and outcomes than wealth neutrality or spending equalization; in fact, the latter approaches pay virtually no attention to the results of schooling. For definitional purposes, however, the key characteristic of adequacy seems to be less the input-output distinction and more its greater emphasis on absolute rather than relative stan-
dards. In the past, debates over equity focused on comparisons among children and districts and how well they fared relative to each other. Adequacy appears to demand the setting of absolute standards rather than defining equity in terms of the relative performance of school finance systems.
Adequacy may also result in different groups taking center stage as the focus of special interest. William Clune, a prominent legal scholar (and original member of the Coons team that developed the wealth-neutrality standard for school finance equity), illustrates in the development of his own thinking one direction that adequacy might take. From early descriptions of adequacy that included all children, his writing has evolved to define adequacy as a concept to be applied especially to urban, poor districts and to high-poverty students (Clune, 1995).
It is in determining how adequacy can be evaluated that there is perhaps the least consensus on defining the concept. Much of the remainder of this chapter assesses the various options being explored by courts and states as they attempt to apply an adequacy standard to school finance systems. In practice the distinction drawn here between older measurements of equity using distributional bases and adequacy as an absolute measure is frequently blurred when it comes to defining how the adequacy of a finance system should be evaluated, probably because concepts using absolute rather than relative standards are only now being developed. Some courts (e.g., Harper v. Hunt in Alabama2) have found the state finance system inadequate using comparisons with state and national input and output regulations and standards. Sometimes the ex post tests of adequacy are also comparative rather than absolute in nature, as when outcome measures such as test scores or graduation rates are compared with national statistics.
Developing absolute standards of adequacy (both ex ante and ex post) requires answers to two questions: Adequacy of what? How much is adequate? Strike (1988) has pointed out that the philosophical debate over distributive justice that has taken place over the 30 years since the publication of Rawls's A Theory of Justice (1971) provides some framing ideas for considering these questions.
In seeking an alternative to utilitarianism, which holds that just policies are those that produce the greatest good for the greatest number and defines "good" as happiness or utility, Rawls (1971) sought plural indices of social welfare rather than continuing to rely on happiness or utility. He developed the notion of "primary goods": things that a rational person wants whatever else that person wants, desirable because they are means to leading a wide range of different kinds of lives. He listed the primary goods as rights and liberties, powers and opportunities, income and wealth, and self-respect.
The Rawlsian theory of primary goods suggests a way of approaching an-
swers to the "adequacy of what" question. Interpreted this way, adequacy might be viewed as requiring an allocation of educational resources sufficient to guarantee to every student a minimum set of these educational outcomes that are importantly connected to long term life prospects in our society.
In developing an alternative to Rawls's concept of distributive justice, Gutmann (1987) provided a possible approach to answering the "how much" adequacy question and thereby a means for overcoming equality of opportunity's second limitation, the absence of a principle for establishing levels of achievement or permissible differences in achievement. Unlike Rawls, whose definition of distributive justice featured distributive rules intended to provide an equal opportunity for everyone to pursue a self-chosen plan of life, Gutmann's definition gives greater attention to the value of a democratic society in which decisions are made deliberatively and collectively. Therefore she emphasizes a so-called democratic threshold principle for education, which claims that "inequalities in the distribution of education goods can be justified if, but only if, they do not deprive any child of the ability to participate effectively in the democratic process (which determines, among other things, the priority of education relative to other social goods)" (p. 136). The idea of a democratic threshold at least conceptually addresses the issue of achievement levels ignored by the equality of opportunity approach, and it provides in general terms an anchor for answering the question of "how much" adequacy.
Suggestive as Rawls's primary goods and Gutmann's democratic threshold principle are, however, they provide far from precise answers to the "adequacy of what" and "how much" questions. They point, though, to some of the issues that will arise in seeking greater precision.
To facilitate the discussion, we develop a shorthand whereby we view "adequacy of what?" as a question of qualitative adequacy and "how much" as a question of quantitative adequacy. To apply adequacy to school finance requires figuring out where to set the absolute level of this "bar" called adequacy: how broad or narrow should the bar be (qualitative adequacy) and how high or low should it be (quantitative adequacy)?
Qualitative adequacy involves decisions about what educational opportunities are most directly and powerfully related to the "primary goods" society wishes to distribute fairly among its citizens. These are usually issues of curriculum. Opportunities to become proficient in reading or math might be seen as central. Opportunities to participate in art or physical education might be viewed as more peripheral. The availability of swimming pools or carpeted classrooms might be regarded as unrelated to ultimate life chances. It is not hard to imagine, however, that people will disagree about where to draw the lines, both because they have differing values and because the relationship between opportunities and life chances may be different for different individuals in different circumstances.
Similarly, quantitative adequacy is likely to lend itself to differing interpreta-
tions of "how much" is enough. There might be an achievement level such that those who fall below it would be unlikely to be able to participate in the basic economic and political life of society. This suggests looking for a threshold level of some sort that defines how much input is needed to reach this level. Even if such a threshold level can be agreed upon, there will still be important questions to resolve. Threshold principles often apply to "all that are able," which involves not only judgments about the capacity of individuals but also about the resources society is willing to expend on teaching those with low capacity. Thus, there are choices to be made between setting the adequacy bar at a high level (and accepting that a larger number of students will not meet it) and a low level (which allows for fewer exceptions). On another front, deciding where to draw the line on quantitative adequacy must also take into account the possibility that there might be a law of diminishing returns, such that higher levels of educational resources result in ever-diminishing amounts of improvement in educational achievement or ultimate life chances.
Qualitative and quantitative adequacy thus raise numerous issues when considered independently; they also interact in ways that will pose additional challenges for the development of adequate school finance systems. For example, the interaction has interesting implications for the outcome of a state decision about whether or not to allow individual districts to provide more resources than called for by the state-determined "adequate" level. On one hand, let's say that a state school finance system embodies a broad definition of qualitative adequacy, incorporating most or all of the primary educational goods that will affect life chances significantly and/or that its quantitative adequacy level is sufficiently high that spending above the adequate level will not yield much in additional returns on the critical dimension of life chances. In this case, districts' freedom to spend whatever they want beyond the threshold may result in nice extras for students but will not compromise equality of opportunity. Such a "broad or high adequacy" policy is likely to be quite expensive and redistributive, however, and therefore politically unpopular with wealthier parents who would like to keep their resources in their local schools to benefit their own children. "Narrower or lower adequacy," on the other hand, erodes commitment to equality of opportunity by setting a floor under achievement that guarantees only participation in the basic institutions of society, not equal opportunity to enjoy all of society's primary goods.
Despite these difficulties, adequacy holds promise for overcoming two serious theoretical weaknesses with the more common concept of equality of opportunity. First, equality of opportunity—the idea that all children should have an equal chance to succeed and that education is one of the most efficient tools for ensuring this—is insufficiently attentive to which educational outcomes matter most in the sense of making an important contribution to the life prospects of individuals. Second, equal educational opportunity requires no particular level of achievement, nor does it forbid significant inequalities in achievement between
high-achieving and low-achieving individuals so long as variations in achievement are not associated with "morally irrelevant" characteristics. (A morally relevant characteristic, in this view, might be individual student ability or effort, but not an accident of birth.) Adequacy might overcome these limitations in the equal opportunity definition of equity.
A final general observation about adequacy: while the concept is enjoying a newfound prominence, it actually is an idea with old roots in school finance theory and practice. Specifically, adequacy and traditional foundation aid programs have much in common. In recognizing this, we open the possibility that there are historical lessons from which to benefit in the current round of adequacy discussions. In particular, we should be alert to the possibility that adequacy (especially a low standard of quantitative adequacy) could reproduce the disequalizing consequences of traditional foundation plans.
Foundation plans began in the early part of the 20th century and efforts by school finance reformers to overcome problems with flat grants, the early form of state aid to school districts. Cubberley (1919a; 1919b) and Strayer and Haig (1923) elucidated and developed the idea of the foundation program. Foundation grants set a minimum level of spending per pupil below which a state does not permit a district's spending to fall. Each district is required to levy a property tax at a fixed rate; the state supplements the revenues from this levy up to the foundation level.
Foundation plans3 implicitly define the foundation level as "adequate," but seldom were these levels determined by a systematic assessment of what was required to fund an adequate education. "Adequate" was determined through a political bargaining process, as legislators and governors negotiated on how much state revenue was available or could be generated through additional taxation. Sometimes states made gestures to the notion of adequate inputs by tying their foundation plans to their requirements for teacher certification or to maximum allowable class size levels or pupil-teacher ratios.
State foundation plans were frequently criticized by school finance experts (e.g., Odden and Picus, 1992:176–177) because of their disequalizing effect (they permitted districts that could afford to do so to spend above the foundation level) and because they incorporated no formal adjustment mechanisms to account for changes in educational costs (so over time the foundation level tended to fall below what districts needed to address basic needs). When these plans first came under legal challenge, it was their distributional equity, not their adequacy, that was at issue. Recently, however, the adequacy of foundation and other formulas for distributing education aid has also become a central concern.
THE SHIFT TOWARD ADEQUACY
While there were court decisions in the 1970s in New Jersey, Washington, and West Virginia that gave attention to the adequacy as well as the wealth neutrality of school finance systems, it was the Kentucky Supreme Court's decision in 1989 overthrowing the entire state education program that galvanized the shift toward adequacy in courthouses and statehouses. Between 1989 and mid-1998, courts in Alabama, Massachusetts, New Hampshire, New Jersey, New York, North Carolina, Ohio, Tennessee, and Wyoming also have ruled that their state constitutions' education clauses guarantee students an adequate level of educational opportunities that should allow them to achieve certain desired educational outcomes. In addition, claimants in Arizona won an adequacy case concerning capital costs of school construction; and adequacy-based lawsuits were pending in mid-1998 in Louisiana, Minnesota, Pennsylvania, and South Carolina.
Part of the reason that the 1970s decisions in New Jersey, Washington, and West Virginia did not spur more adequacy cases was that it was not evident at the time that the courts were doing anything significantly different than they did in traditional wealth-neutrality cases. Kentucky, however, clearly marked a sea change. In Rose v. Council for Better Education, 790 S.W.2d 186 (Ky. 1989), the state supreme court used ''efficiency" language in the constitution's education clause4 to declare the entire educational system inadequate and unconstitutional. The court established the objectives of an adequate education, proclaiming that it would provide students with the opportunity to develop at least the following seven capabilities:
sufficient oral and written communication skills to enable students to function in a complex and rapidly changing civilization;
sufficient knowledge of economic, social, and political systems to enable the student to make informed choices;
sufficient understanding of governmental processes to enable the student to understand the issues that affect his or her community, state, and nation;
sufficient self-knowledge and knowledge of his or her mental and physical wellness;
sufficient grounding in the arts to enable each student to appreciate his or her cultural and historical heritage;
sufficient training or preparation for advanced training in either academic or vocational fields so as to enable each child to choose and pursue life work intelligently; and
sufficient levels of academic or vocational skills to enable public school students to compete favorably with their counterparts in surrounding states, in academics, or in the job market.
The influence of the Kentucky case on other states was direct and apparent. Courts in Alabama, Massachusetts, and New Hampshire relied specifically on the Kentucky court's definition of an adequate education when providing guidance to their own state legislatures about crafting remedies for finance systems that had been declared inadequate.5
Other state courts developed their own specifications about what constitutes educational adequacy. In 1993, the Tennessee Supreme Court found that the state constitution required the education system to provide districts with sufficient funds to permit attainment of certain broadly defined educational outcomes: "The General Assembly shall maintain and support a system of free public schools that provides at least the opportunity to acquire general knowledge, develop the powers of reasoning and judgment, and generally prepare students intellectually for a mature life" (Tennessee Small School Systems v. McWherter, 851 S.W.2d 139, Tenn. 1993). Similarly, in 1994, the Arizona Supreme Court ruled that the state's system for funding school facilities was unconstitutional because certain districts lacked the resources necessary to maintain adequate school buildings (Roosevelt Elementary School District v. Bishop, 877 P.2d 806, Ariz. 1994). That decision, while limited to capital funding, also suggested that similar claims of adequacy might apply to school districts' operating costs. In 1995, the highest court in New York used a civic rather than an economic rationale to undergird its finding that the state is constitutionally obligated to create and maintain an education system that provides children with "the basic literacy, calculation, and verbal skills necessary to enable [them] to eventually function productively as civic participants capable of voting and serving on a jury . . . [and] minimally adequate physical facilities and classrooms . . . to permit children to learn" (Campaign for Fiscal Equity v. State of New York, 86 N.Y.2d 307, N.Y. 1995).6
Courts that mandate adequacy do not always themselves define the objectives that an adequate education system should serve. In 1997, the Ohio Supreme Court ruled that by permitting dramatic deficiencies in facilities, materials and supplies, and class sizes in some of the poorer school districts, the state had violated its constitutional duty to provide students with a "thorough and efficient" education system (DeRolph v. Ohio, 79 Oh.St.3d 297, Oh. 1997). The court,
however, did not discuss what it would consider acceptable, leaving it to the state legislature to determine how to meet the constitutional requirement. In Wyoming, in a 1995 development in a court case that went back to 1980 (Washakie v. Herschler, 606 P.2d 310, Wyo. 1980), the supreme court ruled (Campbell v. State, 907 P.2d 1238, Wyo. 1995) that the legislature's response to its earlier decision had been deficient and ordered it to devise an acceptable remedy: "The legislature must first design the best educational system by identifying the 'proper' educational package each Wyoming student is entitled to have. . . . The cost of that educational package must then be determined and the legislature must then take the necessary action to fund that package. Because education is one of the state's most important functions, lack of financial resources will not be an acceptable reason for failure to provide the best education system. All other financial considerations must yield until education is funded."
Adequacy decisions in the courts do not always address the question of whether districts can provide education above the adequate level, but when they have, they have spoken with several voices. In Wyoming, local supplements are expressly forbidden. The Wyoming Supreme Court noted that "historical analysis reveals local control is not a constitutionally recognized interest and cannot be the basis for disparity in equal educational opportunity." The New Hampshire Supreme Court (Claremont v. Governor of New Hampshire, 703 A.2d 1353, N.H. 1997) and the North Carolina Supreme Court (Leandro v. State of North Carolina, No. 179PA96, 1997) both indicated that variations in spending resulting from local add-ons would be permissible so long as all districts are able to provide students with the constitutionally guaranteed minimum of opportunities. The finance system enacted in Kentucky in response to the 1989 ruling there allows local spending above the state-determined adequate level and so far has not faced legal challenge.
As happened in the wealth-neutrality litigation, not every state's high court has been receptive to adequacy arguments. In Illinois, where the state constitution's education clause explicitly requires the state to "provide for an efficient system of high quality public educational institutions and services," the supreme court rejected attempts by plaintiffs to evaluate whether the quality of education offered in many of their districts met that constitutional standard. According to the court, "questions relating to the quality of education are solely for the legislative branch to answer" (Committee for Educational Rights v. Edgar, 672 N.E.2d 1178, Ill. 1996). The high courts in Rhode Island and Florida relied on a similar rationale in rejecting adequacy-based claims. In Florida: ''appellants have failed to demonstrate . . . an appropriate standard for determining 'adequacy' that would not present a substantial risk of judicial intrusion into the powers and responsibilities assigned to the legislature" (Coalition for Adequacy v. Chiles, 680 So.2d 400, Fla. 1996). In Rhode Island: "what constitutes an 'equal, adequate, and meaningful' [education] is 'not likely to be divined for all
time even by the scholars who now so earnestly debate the issues'" (City of Pawtucket v. Sundlun, 662 A.2d 40, R.I. 1995).
Another possible dimension of an adequate education has been proposed in two Minnesota cases seeking, among other substantive education remedies, racial integration.7 Advocates there have sought to have an adequate education defined in such a way as to include a racially integrated education. If this approach were to succeed, it would provide a way, through the state constitution's education clause, to remedy unintentionally created racial isolation, a situation that the federal constitution—through the equal protection clause—would not redress.
STATE RESPONSES TO THE ADEQUACY MOVEMENT
The adequacy movement is comparatively new; there is limited evidence to date on what is required to reform school finance systems on this basis. Early indications can be gleaned from a review of state responses so far.
The Kentucky example proved that moving to educational adequacy as the basis for reform may require legislative solutions going far beyond matters of school funding. The fact that Kentucky's legislature promptly enacted a comprehensive statewide education reform package in response to the court's decision is encouraging but, given particularly favorable political incentives in the state (Carr and Fuhrman, 1999), probably not reflective of how other states may respond to similarly far-reaching court mandates. The Kentucky Education Reform Act (KERA) recreated the state's entire elementary and secondary education system, encompassing finance, governance and program changes; increased school district revenue by 34 percent (19 percent adjusted for inflation) between 1990 and 1993; and reduced disparities in spending among districts and in the relationship between district wealth and spending (Adams, 1997). KERA also featured a strong accountability program based on a new assessment system and providing financial rewards for exceptional performance and significant sanctions for poor performance.
Legislative efforts to comply with adequacy rulings in states other than Kentucky have been noticeably more contentious. Chapter 3 introduced the long saga of school finance litigation in New Jersey, which had adequacy overtones from its early days and assumed a more explicit adequacy cast in the 1980s, when school finance reform advocates filed a new suit on behalf of the children in 29 of the state's poorest districts. The Abbott v. Burke, 575 A.2d 359 (N.J. 1990), decision decreed that the children in the 29 districts identified in the court case as special needs districts must have educational opportunities equal to those of
students in the wealthiest districts in the state; in a later ruling the court made clear that equal opportunity included access to all educational and extracurricular activities that were available in the wealthiest districts as well as funding for supplemental programs to address the special educational needs of urban districts. The politics of finding legislative remedies satisfactory to the court were complicated by the fact that the court created an educational entitlement only for children in the 29 poorest districts and did not address the finance systems' constitutionality in the remaining 450 districts. Opposition to increased aid to urban districts and to the tax increases necessary to fund this aid derailed initial legislative attempts to comply with the court's ruling. Nonurban legislators who voted for the initial legislative reform package almost all lost their seats in 1991 and opposition to the reform package played a significant role in the defeat of Governor James Florio in 1993 (Carr and Fuhrman, 1999). The case was only settled in 1998 when the state supreme court, after rejecting another legislative reform in 1997, finally approved a detailed and comprehensive reform package.
Experiences in other states have also highlighted the political difficulties that frequently follow adequacy rulings by courts. Despite the similarity of the judicial decisions in Alabama and Kentucky, the fate of the former has been quite different from the latter. Antitax groups mobilized against comprehensive reform in Alabama; the Alabama Education Association opposed the bill as well because of its accountability features. In 1994 the state elected a governor (Fob James) who explicitly opposed the court order and tried (unsuccessfully) to have it overturned. To date finance reform has not occurred in Alabama and plaintiffs are still seeking judicial remedies. In Wyoming, an initial legislative effort to define and impute costs to a "proper education" was quickly challenged in court by plaintiffs seeking yet more revenue for Wyoming schools. By 1998, though, the legislative and executive branches came together to pass a reform that they believe will hold up against judicial scrutiny, a reform that raised the state education budget by approximately 10 percent and elevated per-pupil revenue into the $7,000 range. Ohio's legislature voted a new funding plan designed to provide an adequate instructional system, but voters refused to approve the necessary tax increase. Arizona's legislative and executive branches made several efforts to revamp the state's system for funding capital facilities and costs after the state supreme court ruling in 1994; in 1998 the court for the fourth time declared the state's effort unsatisfactory and gave the legislature 60 days to develop an acceptable plan.
States that are not under court order to implement educational adequacy have also attempted to link their educational reform and school finance agendas by asking "What is a core education and what does it cost" (Education Commission of the States, 1997). Their experience as well as the experience of states under court mandates to orient their school finance systems toward educational adequacy indicates that there are a number of conceptual and technical challenges to overcome.
CONCEPTUAL AND TECHNICAL CHALLENGES
Deliberations in courthouses and statehouses suggest that a desire to implement adequacy may be outpacing current understanding of how to define and achieve it. In comparison to wealth neutrality and equal spending, equity defined as adequacy requires difficult value choices, as well as policy decisions in areas in which the available technical knowledge is weak. Implementing adequacy requires establishing anchors for identifying what is adequate; determining the costs of an adequate instructional delivery system; making adjustments for student, school, and geographic characteristics; adjusting for inflation from year to year; and developing an assessment system for measuring whether adequacy has been achieved.
Establishing Anchors: What is Adequate?
We have already observed that adequacy requires both identifying desired educational outcomes and then making decisions about what kinds of educational experiences (qualitative adequacy) and how much achievement (quantitative adequacy) will be sufficient to meet the standard. We have also noted that sometimes courts provide the answer to at least the first of these dilemmas (e.g., in Kentucky), but often the question of outcomes and always the question of "adequacy of what" and "how much" are left to legislatures and governors to determine.
These are not easy questions to resolve through political processes. Value conflicts make it difficult for Americans to reach consensus about goals for education, or at least about which goals should receive priority. There are large uncertainties about which educational experiences are central to the achievement of specific goals. There are also large uncertainties about what levels of achievement are necessary for students to reach specific goals.
These are not insurmountable problems, and in fact headway is being made in addressing them. The Wyoming legislature, under a deadline from the court to define and cost out a "proper" educational package, developed a basket of education goods and services consisting of some 30 courses and kinds of knowledge designed to ensure that students acquired a common core of knowledge and a common core of skills. The basket served as the basis for subsequent efforts to identify and cost out the resources sufficient to achieve these objectives.
The standards-based reform movement (discussed further in Chapter 5) has inspired all states to undertake serious consideration of their educational objectives. Standards-based reform has encouraged states to deliberate on their own goals, to undertake the development of shared understandings about what students need to know and be able to do (i.e., to set content standards for education), and to develop systems to measure whether students do in fact master this content (performance standards). It is likely that standards-based reform reinforces the willingness of policy makers and judges to shift toward an adequacy standard for
educational equity, by creating a receptive political climate and perhaps by providing judicially manageable standards that courts had a difficult time identifying in the past.
Nevertheless, the tortuous path to standards-based reform is also a cautionary tale about how long and difficult the process can be. As a recent National Research Council committee pointed out, it is not yet clear whether the guiding assumptions of standards-based reform are correct or that policies built on them will have their desired effect (National Research Council, 1997:33–46). The rhetoric that "all students can learn to high standards" leaves unresolved important philosophical and logistical issues, such as balancing high, uniform standards with students' unique educational needs and abilities. The extent to which consensus can be reached on curriculum and performance standards is unclear. Controversies over the specification of outcomes (e.g., in Pennsylvania), the content of curriculum frameworks and performance assessments (e.g., in California), and the content of voluntary subject-matter standards (e.g., the lopsided U.S. Senate vote condemning voluntary national standards in history) "suggest that consensus dissolves once the public moves beyond a general belief in the need for standards and assessments to questions about what those standards should be and how students should be taught and tested" (National Research Council, 1997:38). Moreover, major uncertainties remain about whether student performance can be measured validly and reliably and whether instruction consistent with the standards can be implemented in individual schools and classrooms.
The technical challenges in measuring student performance become increasingly important as the concept of adequacy shifts the focus of attention to the outcomes of education. A key issue is whether existing tests define and measure achievement in ways consistent with standards-based reform or other statements of desired educational outcomes. Large-scale, standardized tests8 are tools for determining what students know and can do in specified domains. No large-scale assessment measures all aspects of student achievement. Moreover, many tests currently in use do not capture critical differences in students' levels of understanding, do not adequately reflect higher-order thinking skills called for by many new education standards, and do not reflect more comprehensive goals for student achievement that go beyond subject-matter knowledge to other valued skills and abilities. For example, a recent National Research Council (NRC)
committee evaluating the National Assessment for Educational Progress (NAEP) pointed out a number of dimensions of achievement not sufficiently reflected in the current NAEP frameworks and assessments: problem representation, use of strategies, self-regulatory skills, explanation, interpretation, and individual contributions to group problem solving (National Research Council, 1999a:138). Assessment systems employing multiple measures of achievement (i.e., large-scale tests augmented with alternative assessments) may be necessary to capture the range of outcomes comprising an adequate education. The recent NRC evaluation of NAEP discusses one approach to a multiple-measure assessment system (National Research Council, 1999a), but at present it is the case that "policy and public expectations of testing generally exceed the technical capacity of the tests themselves" (National Research Council, 1999b:30).
Despite the difficulties, however, adequacy litigation and legislation are beginning to define ranges of achievement levels and exposure to knowledge and skills that can serve as ends toward which to orient a practical school finance distribution system. The next challenge facing policy makers is to determine the costs of an acceptable instructional system oriented toward these objectives.
Determining the Costs of an Adequate Instructional Delivery System
In a perfect world, the determination of the costs of providing educational adequacy would take into account differences among individual students that affect their educational needs. Policy makers, however, cannot easily design a resource allocation program for individuals. They must instead concentrate on building finance systems, ideally with an eye to providing districts, local schools, and even classroom teachers with resources and inducements to tailor instruction to the specific characteristics of students. These systems may on occasion prove blunt when measured against individual student needs; the methods currently available will probably appear primitive down the road when more research and better data allow existing approaches to be refined.
Varying approaches to designing an instructional delivery system differ in crucial ways, both in how they assign costs and how they make reference to educational outcomes. States with early "adequacy" rulings (e.g., Washington, West Virginia, and Kentucky) have continued to rely on traditional legislative processes, which depend on political bargaining to make allocation decisions about educational inputs and leave the links between outcomes and the costs of an adequate instructional delivery system largely unspecified.
More recently, policy analysts and researchers have begun to explore several approaches for calculating the costs of adequacy that link outcomes and instructional delivery more explicitly: (1) inference from outcomes by statistical analysis; (2) inference from outcomes by empirical observation; (3) professional judgment; and (4) a market-oriented approach based on the development of whole-school designs that school districts can "buy." (More detail about these
methods can be found in Guthrie and Rothstein, 1999; Duncombe and Yinger 1999; and Odden and Busch, 1998.) The first and second of these approaches usually depend on states' having sophisticated student achievement testing systems that provide standardized statewide measures of student performance, with data linking this performance to student background characteristics. In states where such testing systems do not exist, then the third and fourth approaches, based on professional judgment or whole-school designs, seem at present to be the only alternatives, where "getting to adequate" necessitates building instructional resource models to which costs can subsequently be assigned.
Each of these alternatives results in an estimate of the cost of an adequate education for a presumed or hypothetical typical student. Conceptually the next step must then be to adjust this cost (or perhaps redefine the goal of adequate outcomes) for students in different socioeconomic circumstances and locations. The statistical analysis approach accomplishes this second step simultaneously with the step of identifying costs for the typical student. It is important for explanatory purposes, however, to keep the two steps separate, and we address cost adjustments explicitly in the following section. Separating the steps permits us to give appropriate attention to the cost adjustment issue and also to make clear that methods of cost determination may have different strengths and weaknesses, depending on whether they are being used to find average costs or to adjust costs for student and district differences.
Inference from Outcomes by Statistical Analysis
This approach represents the efforts of econometricians to apply the tools of statistical modeling to the determination of educational costs. The approach originated in efforts to determine how intergovernmental aid formulas should be adjusted to take into account public service costs beyond the control of local jurisdictions. Researchers tried to develop cost indices that would measure how much higher (or lower) the costs of providing a given level of services were in each district compared with a statewide average. Thus, initially, the studies focused exclusively on the differences in costs, with little attention paid either to the particular service level or what it would cost a typical district to provide that level. More recent studies have extended the approach, so that a specific service level (defined, for example, in terms of the percentage of students achieving various educational goals) can be specified and the cost of providing that service estimated for the district with average characteristics. Ladd and Yinger (1994) and Downes and Pogue (1994) made important theoretical contributions on which have been constructed cost models applied to education data from New York (Duncombe and Yinger, 1999) and Wisconsin (Reschovsky and Imazeki, 1998).
Rather than attempting to determine the costs of an adequate education as the sum of the costs of individual instructional components, statistical models take a "black-box" approach, choosing an acceptable level of pupil performance or
proficiency and then using multiple regressional analysis to determine the dollar amount associated with it based on analysis of extensive data from individual districts. In many studies of this type, no attempt is made to control for the efficiency or inefficiency with which a district is operating. Implicitly such studies take as given the average amount of inefficiency across districts and assume that the efficiency with which a district operates is not correlated with other district characteristics that are included in the statistical model. In effect, inefficiency is assumed to be distributed randomly among school districts. Provided this assumption is reasonably valid, districts with above-average inefficiency are not rewarded when adequacy is defined or aid is distributed on the basis of results from this approach (as they would be, for example, if aid were distributed on the basis of actual expenditures). Nor, however, does this approach provide any incentive for the typical district to operate more efficiently.
The more sophisticated versions of this approach (Duncombe et al., 1996; Duncombe and Yinger 1997, 1999) enter a measure (albeit an imperfect one) of efficiency directly into the model and hence, to the extent that the model is correct, can adjust estimates of adequacy and/or state aid programs to provide incentives for districts to become more efficient.
These versions are more sophisticated in other ways as well. For example, Duncombe and Yinger used a statistical approach to determine the desired objectives of the education system as well as the costs of achieving it. Their approach determined which performance indicators are valued by voters, as indicated by their correlation with property values and school spending. The resulting "index of educational performance" for school districts in New York state includes the average share of students above the standard reference point on 3rd- and 6th-grade Pupil Evaluation Program tests for math and reading, the share of students who receive a more demanding Regents diploma (which requires passing a series of exams), and the high school graduation rate. While this approach results in a performance yardstick, Duncombe and Yinger note that it cannot determine the point on the yardstick that school districts should be expected to meet or that defines an adequate performance. The performance target must be based on the judgment of public officials, though once the target is set its costs can (at least in theory) be calculated via the statistical approach.
Because these indicators (test scores, graduation rates, and Regents diploma awards) may not accurately measure the totality of school outcomes that voters care about, Duncombe and Yinger also estimated models using "indirect" controls for school district performance. In such models, they define average performance as the level achieved by the district with average income, tax price, and voter characteristics, given its teachers' salaries and environmental factors. These (abstract) communities can then be used to observe how much per-pupil spending is necessary to achieve average educational outcomes, again while controlling for other cost or discretionary factors.
Reschovsky and Imazeki (1998) also utilized a statistical method to estimate
the cost of adequacy in Wisconsin districts. They measure outcomes by 10th-grade test scores, controlled for 8th-grade scores of the same students. By this means, they attempt to isolate the "value-added" by school districts, reasoning that the 8th-grade score may reflect students' social capital and instruction in other locations as well as the effectiveness of instruction in the present district. "Adequate" outcomes were defined as the average 10th-grade value-added throughout the state; Reschovsky and Imazeki conclude that the cost of achieving this adequacy, before adjusting for student need and geographic differences, is $6,331 per pupil.
Applying sophisticated statistical models to the determination of adequate educational costs is in its infancy. In part because of its technical complexity, statistical modeling is unlikely to appeal to policy makers as the primary way of calculating the cost of an adequate education. Modeling holds great academic and theoretical interest and may suggest insights that would stimulate productive further research into the relationship between spending and student achievement, as when different modeling techniques lead to different cost estimates. However, in addition to complexity, statistical modeling has shortcomings that, given the present state of the art, limit its usefulness as a policy tool for determining the costs of providing an adequate education to the typical student.
The first of these relates to the limited number of outcome measures these models can incorporate without becoming impossibly complex mathematically. Duncombe et al. (1996) acknowledge this limitation, but assume that it may not be too serious since "outcomes often are highly correlated with each other." Whether or not this assumption is correct, it is also true that many of the desirable educational outcomes with which courts are concerned (e.g., in Kentucky) or legislatures (e.g., in Wyoming) are presently not measured and therefore cannot be quantified for use in such a statistical model.
Furthermore, even if it were possible to quantify all outcomes, such models could at best tell us what resource levels are generally associated with acceptable achievement (with inefficient practices removed, to the extent known), not what resource levels would be necessary, if used efficiently, for this achievement. To reach this level of analysis, the statistical controls would have to include alternative pedagogies and curricula, something beyond current sophistication. If the policy goal is for a legislature to adopt (or a court to mandate) the minimum level of resources necessary to achieve acceptable outcomes, this becomes a crucial distinction.
A more serious shortcoming of the statistical approach relates to its theoretical dependence on educational production function models. One of the longest-running debates in school finance concerns what is called the "production function" in education: whether it exists in any measurable way (Murnane and Nelson, 1984; Monk, 1990) and whether systematic relationships can be found between inputs and student achievement. Hundreds of studies have failed to yield agreement among scholars about the effects of resources on outcomes.
While the statistical models being described in this chapter are cost functions, not production functions, they can be derived directly from production functions and embody the same information. Continuing uncertainty about the underlying common relationships therefore raises significant doubts about the extent to which statistical models yield reliable information about the costs of an adequate education for the typical student.
Statistical methods for determining adequate educational costs appear to have a greater level of precision than the other methods discussed in this section, so it is important to keep in mind the assumptions and judgments behind them. Given restrictions on current ability to quantify desirable outcomes and the weaknesses in the production theory on which cost models are constructed, the apparent precision of statistical models may be misleading. While these methods may, especially as they are improved, provide important comparisons with methods of determining costs that are less elegant, they are not yet ripe for use as the primary means for policy makers and the public to discern or understand these costs. (They may at present be more useful in determining cost adjustments, once average costs have been identified—discussed in the next section.)
Inference from Outcomes by Empirical Observation
Another black-box approach involves establishing a level of acceptable pupil performance on an agreed-upon set of outcome measures, identifying school districts or schools that achieve the desired goals, and determining what these "successful" districts or schools spend. This level of resources is then deemed to be adequate. "The underlying assumption is that any district should be able to accomplish what some districts do accomplish, provided they have a similar amount of revenue and that amount is modified for individual districts to take into consideration cost pressures they face that are beyond their control" (Augenblick, 1997:4).
The empirical approach is described in detail in a 1995 investigation undertaken by Augenblick, Alexander, and Guthrie for the State of Ohio, and then revised in a report by Augenblick in 1997 (Augenblick et al., 1995; Augenblick, 1997). It initially involved constructing a representative pool of Ohio school districts, comprised of all Ohio districts save those that were characterized by high and low extremes of property wealth and per-pupil spending. Once such outliers had been removed, remaining districts were ranked by a composite of student performance measures in reading, mathematics, writing, and science. Districts whose average student performance was at the 70th percentile or higher on most measures were defined as providing a minimally adequate education.
Augenblick et al. next examined instructional arrangements of the districts that met the performance criteria. These districts' mixes of instructionally related components such as ratio of professionals to pupils, class sizes, school sizes, and course offerings were distilled and taken to be instrumentally exemplary for
districts attempting to reach specified levels of achievement. These exemplary conditions and practices can be taken as a model instructional program, one empirically verified by student performance. It then becomes possible to assign costs to these empirically derived instructional components.
A problem with this approach is its suggestion that the identified instructional components and mixes are highly desirable. It can encourage state policy makers to fund specific instructional components and therefore restrict the discretion and initiative of local districts to organize resources and instructional delivery differently if they believe they can accomplish the same objectives more effectively. (This same criticism can also be leveled against the professional judgment approach described below.)
In response to this concern, Augenblick's 1997 report revised the earlier approach by eliminating any empirical observation of school inputs, only observing the average per-pupil spending level that was correlated with acceptable outcomes. In addition, and in response to other criticism, the more recent report abandons a norm-referenced outcome measure (the 70th percentile of the statewide district achievement distribution) and adopts a set of criterion-referenced measures (percentage of students passing minimum competency levels).
For his 1997 study, Augenblick identified 102 (out of 607) Ohio school districts whose students met 17 of 18 performance thresholds, or output criteria (outlier high and low property wealth and/or high and low spending districts were again eliminated from consideration). In addition to a dropout rate of 3 percent or less and an attendance rate of at least 93 percent, the remaining 16 criteria consist of specified passage rates such as 75 percent on the state's minimum proficiency tests. Once having identified a pool of districts that did not exhibit extremes of wealth or spending and in which students had met these state measured performance criteria, Augenblick constructed a weighted mean per-pupil revenue amount from among eligible district expenditure patterns.
The per-pupil dollar amount derived from this process was $3,930, based on 1996 Ohio spending levels (Augenblick, 1997). This became the Augenblick definition of "adequacy" for Ohio school districts, before adjustments for differences in cost pressures beyond district control. He adjusts for these cost pressures using a regression model that assumes that districts spend about what they need to spend to respond to the pressures they face. The model determines the cost-pressure adjustments by regressing district per-pupil expenditures on various cost pressures (e.g., the proportion of special education students in the district).
A similar empirical observation approach was used by the Illinois Commission on Education Funding in 1996 to estimate the cost of the foundation of an adequate education. The commission calculated a foundation level of adequacy (i.e., $4,225 per pupil in 1995–96) and then recommended that the foundation level be adjusted for regional cost-of-living differences and for student poverty
rates. Despite the governor's endorsement, the commission's report was not adopted by the state legislature.9
A seeming virtue of the Ohio and Illinois approaches is their transparency. Compared with statistical analysis, the assumptions appear clear and comparatively easy for policy makers and the public to understand. This seeming transparency, however, may be an illusion. In the case of the Augenblick approach, for example, the amount of funding determined to be adequate depends on the order in which the model's various steps are carried out and on how many of the steps are included and how (e.g., the adequate amount changes depending on whether or not the data are weighted by district size and depending on how extreme districts in terms of wealth and income are handled). Moreover, the approach controls poorly for differences in the socioeconomic background of the students. Because the ranking of how well districts do on the performance measures is undoubtedly correlated with student backgrounds, the approach of taking average costs in districts meeting the standards is a biased estimate of what costs would be in a typical district. This problem would be reduced if the calculated cost were viewed as a minimum and then carefully adjusted for the costs of more difficult to educate students. But the assumption behind the adjustments for cost pressures—that all districts currently adjust fully for the pressures beyond their control—seems questionable and possibly unfair to the districts facing the highest pressures.
Another limitation of the empirical observation approach, as of the statistical models discussed above, is that the minimum proficiency tests measure only certain cognitive outcomes, not the full range of cognitive, value, and behavioral outcomes that courts and legislatures have used to identify an adequate education. As noted above, it may be that further study will determine that all outcomes are positively related, i.e., a district most of whose students pass proficiency tests in reading, mathematics, writing, and citizenship is also likely to be a district that adequately develops student interests in the creative arts or a district most of whose students have adequate social ethics to interact constructively with others in society. However, with no available research on these relationships, it would be premature to jump to such conclusions.
A third strategy for determining what an adequate education looks like is to rely on professional judgment to construct an ideal-type delivery system, without
either statistical or empirical inference from actual measured outcomes. The components of such a system can then be identified and costs assigned to them. While, at first glance, such an approach may seem unscientific, the approximations inherent in professional judgment may be no less precise than those embedded, though more hidden, in statistical or empirical methods. It is possible that professional judgment, if carefully exercised, may be better able to adjust for the vast multitude of factors involved than is a statistical or empirical approach.
A school finance system in which the state funded, or guaranteed funding, for a defined set of resources in each district (including class sizes, teacher salary levels, a specific number of administrators and clerical staff, etc.) was once common, particularly in southern states (Augenblick and Myers, 1994). It is no longer widely used, however, and this system predated a concern to link these resource models to a notion of "adequacy." The notion of input adequacy, however, was implicit in these systems. Once adequacy became an explicit concern, a professional judgment approach was developed by Jay Chambers and Thomas Parrish in proposals they made for funding adequate education systems in Illinois in 1992 and in Alaska two years later (Chambers and Parrish, 1994). Because they recognized that no precise technology exists for linking resources to outcomes in education, they declined to term their goal "adequate," using the term "appropriate" instead. Calling their method the resource cost model (RCM), Chambers and Parrish convened committees of teachers, administrators, and public officials to deliberate and determine what resources were necessary to deliver an appropriate education. They toured facilities across the states and met with local educators and policy makers. In Illinois, for example, they concluded that teacher staffing resources should be provided so that a regular grade 1–3 class should have 22 pupils; that a speech therapist should have a caseload of 62 pupils; and that school buildings insulated to a proper standard should have resources to purchase energy to maintain a year-round building temperature of 70 degrees.
The charge of these committees was not entirely to specify the resources of an appropriate education, because they were also told they must "keep a balance between the resources they would like to see specified for each educational program and what they believed to be affordable" (Chambers and Parrish, 1994:53) given the states' fiscal and political realities. Operating under these guidelines, the process resulted in a recommendation for an appropriate funding level that was 2 percent greater than present total funding in Illinois, and 16 percent greater than present total funding in Alaska. However, the Chambers-Parrish specifications of appropriateness, developed through this process, would have required substantial redistribution of resources from district to district within these states.
Having specified an appropriate level of resources by this consultative process, employing professional judgment, Chambers and Parrish utilized a statistical analysis to estimate the costs and the within-state cost differences of providing these resources. The result, however, was that "policy makers [in both states] tended to find the overall system somewhat incomprehensible and complex"
(Chambers and Parrish, 1994:72). The RCM was not therefore ultimately adopted as a basis for policy in these states.
More recently, a consulting group led by James W. Guthrie utilized a professional judgment approach relying on both consultation with local experts (as did Chambers and Parrish) and reliance on research and whole-school design models (see next section) to calculate an adequate level of resources to be distributed to Wyoming school districts (Guthrie et al., 1997). The Wyoming approach differed from the earlier work of Chambers and Parrish in four important respects.
First, Guthrie et al. had been retained by the state legislature to design a system to fulfill a mandate of the Wyoming Supreme Court in Campbell v. Wyoming, 907 P.2d 1238 (Wyo. 1995). The court prohibited the legislature from considering the total cost of a new education funding system, requiring that the "best" (i.e., adequate) system be funded regardless of cost: "lack of financial resources will not be an acceptable reason for failure to provide the best education system." Therefore, unlike the Alaska and Illinois experts for the RCM, the Wyoming professional expert groups were not asked to balance adequacy against total costs in making their recommendations.
Second, in defining adequacy, Guthrie et al. consulted with professional expert groups in Wyoming and nationally but did not rely exclusively on the opinions of practitioners. Rather, these opinions were used to inform the consultants' views, based on national research and prior experience, regarding the resource elements necessary to produce adequate outcomes.
Third, learning from the Illinois and Alaska experiences, Guthrie et al. did not use a complex statistical method (regression models) to calculate resource costs or cost adjustments, believing that they would be unlikely to be able to explain how these calculations were made in a manner that would be understood and accepted by policy makers, educators, and citizens. Rather, less sophisticated but more easily understandable methods, still based on economic theory, were employed.
Fourth, because the Wyoming legislature was ordered by the court to come forth with recommendations for adequacy on very short notice, Guthrie et al. calculated the costs only of the main elements of an adequate education, using less precise methods to estimate other costs. (For example, the cost of utilities was calculated by taking the average cost of Wyoming districts in the prior year, with no attempt to specify resources necessary to reach a target temperature for classrooms when controlled for building insulation standards.)
In this case, the professional judgment approach was used not only because of concerns about poorly specified outcome measures in education generally, but because the state of Wyoming did not utilize a standardized achievement test like that in Ohio, Illinois, Mississippi, or Texas, even for narrowly defined academic outcomes, and so even poorly specified outcome data were not available. In many states without adequate assessments, the professional judgment and whole-school design methods may be the only alternatives available, without resorting
to the sorts of indirect voter preference models suggested by Duncombe and Yinger.
Like inference from empirical observation, the professional judgment approach opens the possibility of centralizing a great deal of instructional decision making at the state level, if state policy makers use the approach not only to determine the amount of resources to make available to each district but also to fund individual components. This possibility has been of concern in Wyoming, as it was in Ohio. In Wyoming, state officials are seeking to avoid the danger of implementing adequacy in an overly prescriptive was by viewing state resources as a block grant that districts are free to use as they see fit to construct an adequate educational program.
The professional judgment approach can, as noted, he used when concerns about poorly specified outcome measures in education and/or the unavailability of outcome measures preclude inferences about adequacy from either statistical models or empirical observation. It is imprecise, but it has the virtue that its imprecision is transparent. It also involves experts explicitly weighing the best available knowledge about instructional components and their connection to desired educational outcomes, instead of taking the black-box approach to these issues, as do the approaches discussed above.
The method may be vulnerable to criticism because of a different black-box issue. Each of the individuals on the expert panels constitutes a black box of one. It is impossible to be sure of the basis of their professional judgment or to guarantee that each person is free of the kind of conflict of interest that would lead them to make judgments based on their own circumstances or the consequences for their arriving at particular ''adequate" revenue levels. Moreover, the professional judgment approach may not turn out to be reliable; panels may arrive at different conclusions while using similar information. This did not happen initially in Wyoming, where two panels operating six months apart did arrive at similar judgments, but larger samples will be needed before the issue of reliability can be satisfactorily assessed.
Inference from Whole-School Designs
It may now or soon be possible to specify adequate resource levels based on a distillation of national empirical research about effective schools and judgments of professional researchers regarding effective practices. Such specifications might be based on "whole-school designs," off-the-shelf school blueprints intended for adoption in their entirety by schools (Odden, 1997; Odden and Busch, 1998). The New American Schools organization has adopted seven of these designs for promotion to schools, including Atlas Communities, based primarily on the School Development Program (SDP) developed by James Comer; the Audrey Cohen College System developed at the college of that name in New York City; Co-NECT, a school design developed by a Cambridge (Mas-
sachusetts) consulting firm; the Expeditionary Learning program affiliated with Outward Bound; the Modern Red Schoolhouse, designed by the Hudson Institute; the National Alliance for Restructuring Education, which cooperates with schools (e.g., in Kentucky) to restructure their resources to meet higher academic standards; and Roots and Wings/Success for All, developed by Robert Slavin's team at Johns Hopkins University. Other well known designs include the Edison Project, the E.D. Hirsch Core Knowledge Curriculum, the Accelerated Schools model developed by Henry Levin at Stanford, Theodore Sizer's Coalition of Essential Schools, and the CMCD program (Consistency Management and Cooperative Discipline) now being disseminated in Texas, Chicago, and Norfolk, Virginia (Fashola and Slavin, 1998; Glennan, 1998; Northwest Regional Educational Laboratory, 1998; Stringfield et al., 1996).
Whole-school designs are constructed to elavate student achievement. They are also intended by their designers to be salable to local school districts. They can be "purchased" to be implemented. Usually the design teams insist that they be engaged to oversee the training necessary for a school district staff team to implement the design. Because, however, these design teams realize fully that the marketability of their models depends importantly on having a satisfactory price, they are sensitive to costs.
None of the above-listed designs can yet be said to be firmly established by research, in the sense that the achievement of students in schools following these models has been proven superior in replicated controlled empirical or experimental studies. However, many education policy makers are impressed with anecdotal evidence concerning the success of some or all of these programs, with some limited empirical data that tends to confirm it. These designs will become more formidable if research continues to accumulate regarding their effectiveness. The resources specified by each of these designs (with the exception of the National Alliance for Restructuring Education, which does not promote a single design as such, but tailors its recommendations to individual affiliated schools) could be priced, and the sum might be considered the cost, at the school level, of an adequate education. One complicating factor, though, may be that this approach does not yield a very precise definition of the cost of adequacy. One recent estimate of 1996–97 first-year costs of implementing New American Schools designs ranged from $82,600 to $354,000 per school of 500 students above the core costs of a principal and regular classroom teachers, or $165 to $708 per pupil (Odden and Busch, 1998). Moreover, cost estimates may turn out to be quite sensitive to local circumstances.
Making Adjustments for Student, School, and Geographic Characteristics
It has long been recognized that school districts face differences in input prices and differences in the educational needs of their students, both of which
affect the cost of providing educational services. Traditional approaches to school finance equity (such as wealth neutrality and equal spending) could in theory accommodate adjustments to take account of these differences, but in reality these cost differences have been tackled in unsystematic and sporadic ways. The shift toward adequacy, however, makes the issue of adjusting costs unavoidable. It is unfair to hold a high-cost district or school to the same standard of adequacy in its educational program as a lower-cost district or school unless the district or school is given enough resources to compensate it for the higher costs than are outside its control. Indeed, if policy makers fail to take these cost differences into account as they revise their formulas for distributing education aid, they may leave high-cost districts serving disadvantaged students worse off than they were before.
While at least one approach to cost adjustment accounts simultaneously for geographic and student differences (see the discussion of Duncombe and Yinger's work below), it is useful, at least initially, to discuss these different adjustments separately to emphasize the need to take both kinds of cost issues into account. Doing so helps illuminate an important fact. There has been quite a bit of analytical work done on geographic adjustments, and the conceptual issues have been relatively thoroughly explored. Yet school finance distribution formulas seldom explicitly incorporate geographic adjustments. (It is possible that adjustments are being made implicitly, in the sense that weights and other features are added that drive dollars to areas that are perceived to have higher input prices.) In contrast, the knowledge base about the differential costs of educating children with differing needs (such as those from non-English-speaking or economically disadvantaged backgrounds) is comparatively weak. Despite this weak conceptual basis, many school finance programs attempt to account for these differential costs, either by adjusting the per-pupil weights in allocation formulas to account for the presence of special-needs students or via separate categorical programs aimed at these students.
School finance experts have historically approached the problem of adjusting input costs from place to place by focusing on personnel costs, since personnel represent 80 percent of local school budgets (Peternick et al., 1998). Three examples of personnel-based indices developed to assess resources and costs are Barro's (1992) average-teacher-salary (ATS) index, McMahon and Chang's (1991) cost of living index, and Chambers' (1995) teacher cost index (TCI). (See footnote 12, Chapter 3 for an explanation of these indices.) The TCI is the most sophisticated of these approaches for examining national differences in teacher salaries and distinguishing the "cost" of education from actual education expenditures (which are influenced by district decisions about the level or mix of resources to use as well as by conditions outside district control). It is of more theoretical than practical use at this stage, however, as it is still considered developmental and its complex regression methodology and national adjustment factors are difficult for state policy makers to explain and defend, especially in the
absence of adoption and certification by an authoritative official body (Rothstein and Smith, 1997). Instead, the handful of states which currently take account of differences in the cost of education from place to place do so by using state wage indices (Ohio), consumer price indices (Colorado, Florida, Wyoming), or regression analysis using state rather than national data (Texas). The advantages and disadvantages of the various methods (which are described in Mishel and Rothstein, 1997, and Rothstein and Smith, 1997) illustrate that there is no precise or correct method of making the proper adjustment for geographic differences, but that each method brings the state using it closer to a fairer distribution of education resources than would exist in the absence of the adjustments. Moreover, Mishel and Rothstein (1997) suggest that further advances in developing usable education cost indices might come from more research into the numerous geographic cost indices that have been developed for noneducation programs and exploration into the regional patterns that might become evident across methodologies.
Compared to input-price differences, states and the federal government have been more likely to recognize differences in the costs of educating students who are at risk educationally for a variety of reasons. This is accomplished either through special funding programs or by adjusting the allocation formulas in general aid programs to take account of the presence of at-risk students. For the purposes of this discussion of calculating the costs of adequacy, it is the cost adjustments in the general aid formulas that are of most interest.
In 1993–94, 36 states used some form of weighting procedure to adjust basic education support for the additional educational services needed by specific student populations (Gold et al., 1995). (Not all these adjustments necessarily reflected at-risk populations; some states used weights based on differences in grade levels and educational programs.) There appears to be little consistency among states in how these adjustment factors are established. Gold et al. (1995:36) report, for example, that the weights for students in special education programs "appear to vary from being close to arbitrary to being very directly related to actual costs."
Researchers conducting cross-state spending or revenue comparisons (e.g., Parrish et al., 1995, 1998) frequently adjust per-pupil numbers for student needs by using weights of 2.3 for special education students and 1.2 for poor children and children with limited proficiency in English. As Parrish et al. (1995) point out, however, these cost factors are very imprecise due to a lack of relevant data.
The 2.3 factor for special education comes from a study (Moore et al., 1988) using data from a nationally representative sample of 60 school districts in school year 1985–86 and represents the average cost of serving a special education student as a multiplier of the cost of serving a regular education student. This single number, besides perhaps being unreflective of current practice, masks considerable variation in the costs of providing services to students whose disabilities differ substantially in severity and in the intensity of services required.
The 1.2 weight often assigned to children in poverty is based on the average federal Title I allocation for a school year. Levin (1989) derived the 1.2 weight based on average revenues and Title I (then called Chapter 1) allocation per student for 1987. While this indicator may be the best currently available for determining a weighting for students in poverty and is easily understood, it results from federal budget decisions about what to spend on Title I, not on a calculation of the costs of educating poor children and of compensating for prior deprivation that may affect their educational performance. It makes no allowance for the possibility that the additional costs of educating poor children may be higher on a per-pupil basis when there are large concentrations of poor children in a school than when the proportion of poor children is lower. Orland (1990), for example, suggests that attending school with other students from poverty backgrounds is more detrimental for a poor child than being poor but attending school with middle-income students. There is something about the impact of a concentration of low-income students that is itself detrimental. It may be simply that the social capital of one's classmates, or the lack of it, affects learning. Moreover, the 1.2 ratio also understates even the amount of additional money made available for compensatory services to poor students in 1987, since it did not take account of state compensatory programs. The 1.2 weight typically assigned to limited-English-proficient children is merely an extension of the poverty adjustment. Parrish et al. explain the selection of this weight by citing the absence of reliable and representative data on the costs of services to these students and on the logical conclusion that such students are unlikely to cost less to educate than poor students.
Assigning weights to at-risk students, of course, assumes that such students can be identified; but even this is problematic. Poor children, for example, are frequently identified by their eligibility for the free or reduced price lunch program, but this is an imprecise indicator at best. Participation rates vary widely, even among schools that may be socioeconomically similar. Older children are often reluctant to be included in the program and their schools do not push the issue; participation rates in elementary school also relate to how aggressively school administrators and teachers promote it. Counting the number of children with disabilities is complicated because each state can use its own operational definitions for classifying such children, and districts can and do use different criteria for determining whether a child with a disability is counted in the special education or regular education population.
Econometricians, as noted above, have recently made important theoretical and empirical strides in developing statistical models to adjust for the cost of public services beyond a local government's control and have applied these models to education (Duncombe and Yinger, 1999; Reschovsky and Imazeki, 1998). These models contain adjustments for both input-price differences and for the presence of special-needs students.
The most sophisticated application of this approach is found in the work of
Duncombe and Yinger (1999). Using data from 631 school districts in New York, they estimated the average effects on district spending of the following "cost factors," that is, characteristics of each district that are outside the immediate control of school officials that affect the costs of educating students: input prices, district size, percentage of children in poverty, percentage of female-headed households, percentage of students with severe disabilities, and the percent of students with limited English proficiency. This work is more sophisticated than other efforts to develop cost estimates in that the authors use a creative (but somewhat controversial) method to control for the relative efficiency or inefficiency of each district and in that they explicitly account for the fact that educational outcomes, spending on education, and production are all simultaneously determined.
Emerging from their analyses are estimates of cost indices for each district. A district's cost index is 100 if it has average values of all the cost factors, exceeds 100 to the extent the district faces a harsher than average environment for educating students (as measured by its cost factors), and is less than 100 to the extent that the district faces a less harsh environment for educating students. For example, based on their most complete model, the authors estimate that the cost indices for the state's upstate large cities (which include Buffalo, Rochester, and Syracuse) average 189. This figure implies that the cost of providing an adequate education in those districts would exceed the cost of adequacy for the average district by 89 percent. In contrast, costs in 47 small upstate cities are estimated to exceed the average by only 9 percent, while those in upstate suburbs fall short of the average by about 9 percent. The most striking cost indices are for Yonkers (with a cost index of 192) and New York City, with a cost index of 397. (All figures are in Duncombe and Yinger, 1999: Table 8-2, column 2). Taken literally, the New York City estimate would imply that the city would need almost four times as much revenue per pupil as the typical district to educate its students to an adequate level. However the estimates for New York City are quite dependent on the particular specification of the equation and range from 112 to 397. These very large differences underscore the challenges of specifying the model correctly.
This econometric approach to estimating cost differentials highlights the important observation that education costs may well vary quite dramatically across districts depending on the characteristics of the district. As Duncombeand Yinger document, their estimated cost indices exhibit much greater variation across districts than those based on the weighted pupils method currently used by New York State, which weights secondary, disabled, and other-at-risk students more heavily than the typical student, or those based on average teacher salaries. The implication is that relative to the district with average cost factors, the state is currently providing inadequate assistance to some districts (those with high cost indices) and is giving relatively too much state assistance to other districts (those with low cost indices).
Although not yet sufficiently developed to be translated into actual state determinations of funding adequacy, this type of research is extremely important in that it highlights the potentially large variation in cost differences across districts. It is not yet ready to be used by policy makers because a number of technical issues deserve further attention (such as the best way to account for inefficiency), the results vary with the specification of the equation (as evidenced by the variation in results for New York City), insufficient attention has been paid to some important issues (such as ensuring that the teacher salary measures used in the analysis are outside the control of schools so they are not confounded with teacher quality, and accounting for how the cost indices would be likely to vary depending on the desired level of educational outcomes), and the lack of transparency of the approach to policy makers. This last limitation should not be viewed as diminishing the potential usefulness of this approach. As this approach is perfected further, the cost indices it generates can be used as a standard against which more straightforward and easy-to-understand measures can be compared. The ultimate goal would be to develop simple measures that appropriately capture the cost variations that emerge from a conceptually and statistically sound model.
Adjusting for Scale Economies
School finance distribution formulas in approximately 15 states, ranging in enrollment size, income and ethnic diversity, and economic complexity from California, New York, and Florida to Utah and Wyoming, contain a size adjustment factor (Gold et al., 1995). Some states adjust for large size. They accord large districts added per-pupil revenue, presumably justified because there are diseconomies in operating unusually large organizations. Other states adjust for unusual small size, districts, schools, or both.
In some instances, small schools or districts are accorded added revenues on a sliding or weighted scale commensurate with enrollment levels. In other instances, states provide added funds to ensure that selected core curriculum offerings are affordable, even when enrollment levels are too small to generate the necessary funding under other formula provisions. Other states have adopted a per-pupil multiplier by which revenues are increased for students enrolled in small schools (Odden and Picus, 1992).
No two states have adopted a distribution scheme that contains a similar formula component or approach specifying additional revenues a school or district should receive because of enrollment size. The general understanding that per-unit operating costs are known to increase with small or large enrollments (diseconomies of scale) does not translate to policy agreement on specific levels of added costs. This is true both for school and district diseconomy adjustments.
Research on scale economies has been impeded by a general lack of school level spending data. School districts assert that they spend more per pupil in
quite small schools. They assuredly do. However, seldom can they report specifically how much more such schools spend and in which accounting categories. Until more accurate spending data are available over a broader sphere of districts, it is unlikely that a technical base for scale adjustments can be constructed.
School administrators, having been informed that scale economies exist for schools, have often opted to construct large schools. Some city school districts have constructed elementary and secondary schools accommodating literally thousands of students. Whatever, if any, economic advantages such large size may confer, it is far less clear that there are commensurate educational advantages. Indeed, given the prospect of estrangement and anomic behavior among students in unusually large schools, administrators might do well to reexamine their preferences for large scale.
Existing research regarding scale economies suffers from another deficiency. Researchers have adopted overly simple models that, even when school-level spending data are employed, do not take student achievement sufficiently into account. There is evidence (Barker and Gump, 1972; Haller et al., 1990; King, 1994; Lee and Smith, 1997) to suggest that small school settings are educationally advantageous to the students enrolled in them. There are reported benefits such as greater student academic engagement and commitment to meeting instructors' expectations. Also, there appears often to be a productive linking of community to school in smaller settings. If research were conducted properly, policy makers would have a better sense of the balance of advantages and disadvantages between school and district size and student performance.
The solution to scale diseconomies that was most widespread in the middle of the 20th century was to offer financial inducements for districts to close small schools and for states to consolidate small school districts. However, in light of the shortage of evidence that elimination of small operating units in fact has saved money and may have diluted favorable learning settings and communitarian engagement, states have been reducing consolidation pressures.
Recent slowing of state efforts to eliminate small units has not, though, resulted in research about the funding levels needed for them to operate productively. The absence of appropriate data on which to construct a firm research base has led to each state constructing idiosyncratic revenue distribution arrangements. These arrangements are far more of a political than a technical construct.
One of the vexing situations facing state policy makers is to distinguish between a community's preference for small schools and the necessity of having small schools. Presumably, population sparsity and rural remoteness can trigger the need for a "necessary" small school. Here, it is generally acknowledged, the state has an obligation to allocate added revenues to ensure that students are fairly treated. However, if a community merely chooses to have a small school in order to take advantage of the sense of engagement that such an institution may offer, is the state still obligated to pay the added costs? State policy makers generally answer this question in the negative. However, specifying statutorily what is to separate a necessary from a preferred small school may prove difficult. Among
the adjustments are distance. A necessary small school can be defined as one in which a student must travel a specified distance, say more than two miles, in order to attend. States have also attempted to impose a criterion for distance between schools. If schools are located within a specified radius of one another, then it is thought that this is a sign of a preferred, not a necessary, circumstance.
Adjusting for Inflation
After the challenges of determining what an adequate education costs in any given year are met, another task remains: adjusting for inflation each year so appropriations tied to adequacy keep up with changes in the costs of education. Failure to adjust for inflation was one key reason why state foundation grants, a common finance mechanism earlier in the century, over time fell badly below the amounts districts needed to meet basic educational needs.
Disagreements among analysts about how to take account of inflation make it hard to draw conclusions about how the production efficiency of schools has changed over time. The same conceptual dilemmas will face analysts seeking to adjust adequacy levels for inflation. On one hand, school inflation will generally be more rapid than consumer price or gross national product inflation. The latter are heavily influenced by improvements in manufacturing productivity that cannot be matched by schools that rely heavily on labor and where opportunities for technological improvements are comparatively limited. Therefore using consumer price or gross national product indices to adjust for inflation in education will underestimate the real increase in education costs. On the other hand, efforts to develop price indices specifically for education have been few (the major example being an index developed by Halstead [1983; Research Associates of Washington, 1993] that extends back to 1975) and are controversial. In particular, treatment of teacher salaries in education-specific indices has been questioned. The Halstead index is based on actual salary changes and therefore does not reflect the fact that teacher pay reflects district choices about whether to pay teachers more or less than comparable workers, choices that are presumed to affect teacher quality (Rothstein and Mishel, 1997). Efforts to develop inflation adjustments that can be used in specific states to update adequacy levels are in their infancy. An early example for Wyoming is described in Guthrie and Rothstein (1999); the authors also note that how best to account for inflation in any given state may depend on specific characteristics of that state, such as the competitiveness of its teacher labor market.
PROMISES AND PITFALLS
The shift toward adequacy as the equity standard to be sought is appealing because it offers the promise of reorienting debates over school finance toward fundamental issues of education reform.
Adequacy focuses school finance discussions squarely on issues of educa-
tion and student achievement. It thus brings education finance into the central policy dialogue about education, which is concerned with improving school and student performance. More than other equity concepts, adequacy is therefore directly rather than just indirectly related to goals 1 and 2, which have to do with student achievement. Adequacy suggests, but does not guarantee, that educational policy will increasingly be driven by a concern for spending resources in instructionally effective ways, which speaks to goal 1.
Adequacy implies that discussion of the level of educational resources to be made available should include explicit consideration of what the public and policy makers want the educational system to accomplish and what kinds of educational opportunities students must be given to meet these objectives. This orientation stands in contrast to much past and current decision making about education finance, which often is dominated by political bargaining over how to distribute available funds.
Adequacy promises to be especially important for addressing the second of the three goals: reducing the nexus between student background characteristics and student achievement. It requires, rather than just permits, attention to cost adjustments, a topic addressed further in Chapter 7. It also may lead to more focus on urban areas, where education is generally thought to be least adequate. Court cases and policy debates over the adequacy of school funding in cities should help resolve the question of whether money is the key problem in urban schools and focus attention on all the factors affecting the productivity of these schools.
While attracted by the appeal of these promises, the committee notes that there are major unresolved questions that must be addressed if school finance is to be held to an adequacy standard. By raising these questions, we do not intend to prejudge the ultimate usefulness of the adequacy concept; earlier concepts of equity were also daunting in their infancy, but over time much progress was made in defining and measuring them. What we do intend to point out is that adequacy as an equity concept still requires much development and must be used with an awareness of this fact. As courts and policy makers seek to apply an adequacy standard and as researchers attempt to improve understanding of adequacy, they must keep in mind the importance of addressing the following kinds of issues.
The meaning of adequacy is still unclear. Major questions remain open: is it a wide, high standard or a narrow, low standard? Does it focus attention and resources primarily on the disadvantaged or does it contribute to improving achievement for all students? These are not technical questions for which scientific answers can be provided; they will require difficult political judgments that may be subject to the same kinds of public resistance that have faced finance reformers in the past.
What will it mean to extend the concept of adequacy as an equity standard to federal, school, and student-level policies, not just to the district level, where
most court decisions to date have focused? How can the concept be applied to school finance decisions that might arise under different institutional arrangements for public funding of education, e.g., to a system that embraces charter schools or vouchers or both as well as traditional public schools?
If states permit district add-ons to the state-determined ''adequate" spending level (as some court decisions already explicitly allow, while others do not), is this not likely to result in spending disparities that continue to put children from poor families or children living in poor communities at a relative disadvantage in terms of the educational opportunities they are offered? If add-ons are allowed, might not this reduce the support of people in wealthy communities for having the state set a high level of adequate spending?
What happens to the definition of an adequate education when it collides in the political arena with demands to "adequately" fund a host of other worthy objectives, such as better health care, a cleaner environment, well-maintained roads and bridges, safer streets, and so on? Even when a state court (as in Wyoming) finds grounds in the state constitution for giving education pride of place before all other public services, is it really possible to resolve basic resource allocation decisions about education in some rational and technical fashion rather than in the political arena? In this sense, is adequacy really something new under the sun or mainly a new and perhaps misleading label for the customary and unavoidable process of political bargaining?
How will courts or legislators determine if funding is adequate? It would clearly be a mistake to require that all students attain a certain level of achievement, since by that criterion any school that did not reach that goal would be entitled to additional funding no matter how inefficient it might be. The intent instead should be to ensure that schools have the resources that they would need to achieve the desired outcome, assuming that they produced education reasonably efficiently. The challenge arises in figuring out what that amount is.
This last point emphasizes the fundamental problem with adequacy as a new standard for school finance—namely, that it requires better knowledge than currently exists about how to determine what an adequate education costs. In fact, the state of the knowledge base is such what we are far from being able to provide empirically sound answers to the key questions of "adequacy for what" and "how much is adequate." This fundamental problem becomes clearer as we take up the issue of educational productivity in the next chapter.