A Costing-Out Primer
A Costing Out Primer
An education adequacy costing-out study determines the amount of money actually needed to make available all of the educational services required to provide every child an opportunity to meet the applicable state education standards. A variety of approaches for undertaking such studies have been used in recent years in many states, including Kansas, Kentucky, New York, Ohio, Maryland, Montana, and Texas–in some cases as part of the development of a new funding system ordered by a state court. This policy brief describes briefly the concept and history of costing out and offers an overview of the four methodologies used for this purpose.
Ever since states began to appropriate money to local communities to assist with the cost of education more than a century ago,1 state education finance systems have purported to provide a base adequacy amount. In its first incarnation, such state funding took the form of a flat state grant for each school child, theoretically in an amount sufficient to provide a minimum education. Because of insufficiencies in state funds and the inequity of providing the same amount of funding for students in both poor and wealthy districts, during the 1920s many states began adopting “foundation” programs. These required local school districts to levy taxes at a rate that was aimed at generating enough revenue to fund a minimum education, with the state supplementing the amount actually raised by poor districts when they did not yield the minimum “foundation level.”
From the beginning, however, good intentions to support a meaningful foundation level were never realized. No real system was established to determine what the minimum foundation amount should be, and rarely were these foundation amounts set in accordance with any realistic analysis of the actual cost of a minimum education. Instead they tended to be established by the legislature based on the amount of funding currently available for educational funding without any relationship to actual needs.2 Even the base amounts initially established tended to erode over time because of budget pressures and competing political priorities.
The first sustained attempt to overcome the limitations of the historical foundation funding approach was undertaken in the early 1980s by Jay Chambers and Tom Parrish in two groundbreaking studies performed in Alaska and Illinois. In an attempt to “develop a basis for providing cost-based adjustments to the education funding allocations school districts received from the state,”3 they created a “Resource Cost Model” (“RCM”), which sought to identify and enumerate the specific resources that would allow for adequate educational opportunity.
The Resource Cost Model was further developed and enhanced in the 1990s in response to the standards-based reforms adopted in almost all of the states. These reforms established academic standards in each major subject area that reflected the state’s expectations of the specific knowledge and skills that students should possess at the culmination of their high school education. The standards provided education finance researchers useful output measures for determining the “quantity and precise mix of resources needed to reach desired goals in education.”4 The significance of these state outcome standards was further enhanced in 2001 by passage of the federal No Child Left Behind Act,5 which requires states to ensure that all students are proficient in the standards by 2014.
Judicial orders in education adequacy cases provided further impetus for methodological advances. Of particular significance in this regard was the 1995 order of the Wyoming Supreme Court,6 which required the state to calculate the cost of the “basket of goods and services” needed to provide all students with a “proper” education. Refining and expanding the Chambers-Parrish RCM, James Guthrie and Richard Rothstein developed a “Professional Judgment” approach, in which teams of professionals were asked to design an educational program that would meet stated proficiency goals, and to identify all of the specific resources that would be necessary for its success. After the basic prototype was established, the members of the professional judgment panels were also asked to consider whether extra resources would be required to provide certain types of students, such as those from low-income families, students with disabilities, or English Language Learners, with an adequate education. Once team members had identified the set of inputs required to achieve the stated goals, researchers determined the precise cost of obtaining those goods and services for Wyoming school districts through an extensive series of economic analyses and market pricing assumptions.7
A second major methodology was created by John Augenblick and John Myers in response to the education adequacy order of the Ohio Supreme Court in DeRolph v. State.8 This technique was originally called the “empirical” approach, but has come to be known as the “successful school district” method. It sought to identify those school districts that are currently meeting state standards, and then to use their average expenditure amount as a fair estimate of the actual cost of an adequate education. After removing “outliers,” the Ohio researchers chose a sample of successful school districts by reference to six specific measures of student achievement and eight input measures, such as pupil teacher ratio and average teacher salary.9
The significance of the new costing-out methodologies is that they attempt to determine a true foundation level by identifying the specific conditions necessary to provide all children a reasonable educational opportunity and systematically determining the amounts necessary to fund each of these needs.10 Over the past 10 years, professional judgment and successful schools studies have been undertaken in over two dozen states, and two additional methodologies, the “expert judgment” and the “cost function” approaches, have been utilized in a more limited number of instances.11
A number of states have begun using accepted costing-out methodologies to determine the actual costs of meeting the mandates of the federal No Child Left Behind Act (NCLB). The Act includes requirements that states develop assessments based on state standards in grades 3-8, meet yearly proficiency targets, and implement a variety of remedies and sanctions when districts and individual schools fail to meet these targets. Studies in Hawai’i and Minnesota have focused on the administrative costs of implementing the Act, while studies such as those in Texas and Ohio have begun to address the issue of identifying the substantially larger costs of meeting yearly proficiency targets. A sound, consistent methodology for identifying and comparing the costs of achieving proficiency in all 50 states has not yet been developed.
Professional judgment has been the predominant costing out approach in recent years. In addition to the original and a follow-up Wyoming study, professional judgment has been utilized in at least 13 other states: Oregon, South Carolina, Maryland, Kansas, Nebraska, Indiana, Colorado, Missouri, Kentucky, North Dakota, Washington, Montana, and New York. Most of these studies have utilized techniques modeled after the original Wyoming model, but a number of variations on the basic theme have emerged. For example, the Oregon Council on the Oregon Quality Education Model, a 23-person body of legislators, educators, business leaders, advocates, and other community representatives, appointed an expert staff and four separate subject-area work groups that devised prototype elementary, middle, and high schools. For each prototype the Council set forth a detailed list of “program elements,” such as core staff, program staff, additional instructional time for students to achieve standards, and district administrative overhead. “Tangible assumptions” having a direct relation to cost, such as class size, age of building, and numbers of computers per pupil were then determined and specific cost assumptions for each prototype school calculated.12
The most extensive of the recent professional judgment studies is the New York Adequacy Study undertaken jointly by the American Institutes for Research and Management Analysis and Planning in 2003 and 2004.13 This effort involved 10 professional judgment panels (two consisting of educators from New York City, two from other urban districts, two rural, two suburban, and two focusing on special education). The results of their deliberations were synthesized through an elaborate set of computer analyses and then reviewed by a panel of outside experts, a stakeholders’ panel, and a summary professional judgment panel before the specific elements of the synthesized prototype educational models were subjected to economic analyses and market pricing review. A geographic cost of education index was also compiled for this study and a small successful schools study was undertaken both to identify candidates for the professional judgment panels and to provide comparative data on salary patterns. Extensive public engagement processes were employed to identify the precise output standards that should be utilized by the panels and to provide further perspectives on the panels’ recommendations.
In two states, Arkansas and Kentucky, studies were recently conducted that derived resource needs from the literature on “proven effective” school reform models and from the judgments of “experts” who have developed or analyzed those models. In other words, in place of the diverse team of teachers, administrators, and school business officials that comprises the typical professional judgment panel, judgments in these studies are made by a small group of educational policy experts.
For example, in a recent Kentucky study, Allan Odden, Lawrence Picus, and Mark Fermanich set forth and then costed out each element of a high quality instructional program that was based on the latest “state of the art” research. This model included publicly funded pre-school programs for children aged 3 and 4 from poverty backgrounds, full day kindergarten, school sizes of 300-600 at the elementary level and 600-900 at the secondary level, school-based instructional facilitators, class sizes of 15 in grades K-3 and of 25 in other grades, collaborative professional development and extra help strategies for struggling students, family outreach, and technology.14 Overall this study concluded that Kentucky would need to increase spending by $740 million, or 19% above actual expenditures in the base year, to finance adequacy.
Since the original Ohio study, additional successful school district analyses have been undertaken in at least eight other states: Mississippi, Illinois, Maryland, Kansas, Louisiana, Colorado, Missouri, and New York. “Successful schools” is essentially a statistical modeling approach that calculates the cost of an adequate education based on specific data regarding resource inputs, student test scores, and other precisely defined outcome measures. To undertake such a study one needs a clear definition of an agreed-upon set of input and output standards and a way to measure them consistently for all school districts in the state. One reason that fewer states have used this methodology than the professional judgment approach may be that sufficient data is simply not available in all locales.
Recent successful schools studies have added additional factors to the straightforward inputs and outputs analyzed in the original Ohio study. For example, a report prepared for the New Hampshire Adequate Education Costs and Municipal Grant Distribution Commission offered four alternative ways of identifying high-performing districts based on various combinations of input and output factors. One of these alternatives uses “efficiency factors” that eliminate from the pool of model school districts those that provide services beyond a specified maximum level.15 An inverse variation on this theme is the model proposed by the Council of Great City School Districts, which bases the adequacy amount on the total per pupil expenditures of the 10% highest achieving districts in the state.16 In Maryland, the empirical analysis was done at the individual school level because the state has a total of only 24 school districts.
Identification of the standard for success has also taken on added complexity. In the recent New York study undertaken by Standard and Poor’s, “success” was defined four different ways, resulting in differing numbers of “successful” school districts.17 Each definition resulted in substantially different expenditure levels needed to eliminate the identified adequacy gap. For example, under the most rigorous definition of “success” (which included only school districts that met the States 2007-08 performance index targets under NCLB, had a Regents diploma rate above the state average, and had a dropout rate below the state average), the statewide spending gap, using the New York Geographic Cost of Education Index and certain specified special needs weightings, was $8.75 billion; if, however, the criteria for “success” was those districts that have a simple, unweighted average of 80% or more of their test takers scoring at our above the current proficiency level on grade 4 English Language Arts and Math tests and receiving passing grades on five Regents graduation exams (the definition used by the State Regents for their analyses), the spending gap utilizing the same variables would be reduced to $6.03 billion.
Cost function studies attempt to determine, through analyses of performance measures and cost indices, how much a given school district would need to spend, relative to the average district, to obtain a specific performance target, given the characteristics of the school district and its student body.18 The cost function approach differs from the successful school district approach in that it attempts to determine not only a level of spending that is correlated empirically with academic success, but also how that level may change for districts with different characteristics serving different student populations.
Cost function analyses cannot be done in every state because they require extensive state-wide data on per-pupil school expenditures, student performance, and various characteristics of students and school districts. The statistical analyses undertaken to equate levels of expenditure with specified outcome targets are quite complex and often are difficult for policy makers to understand. For these reasons, cost function analyses to date have mostly been theoretical modeling exercises rather studies that are ordered or utilized by legislatures or courts.
Recently, however, both the plaintiffs and the defendants in West-Cove School District v. Neeley, a Texas education adequacy case, submitted cost function analyses to the Court. Cost function was considered a practical and appropriate methodological approach in Texas both because of the exceptional range of data that is available in that state and because of the large variations in district characteristics and the large number of districts whose needs must be assessed. The first study, undertaken for the Legislature’s Joint Select Committee on Public School Finance by Lori Taylor and other researchers at Texas A& M University, concluded that, for 2004, current aggregate spending in the state was at a level sufficient to provide all districts in the state the resources needed to allow 55% of their students to meet the state performance targets for that year; if no funds were to be re-distributed away from districts spending at levels higher than the adequacy level designated by the study, an additional $226-$408 million would be needed. The cost function analysis prepared by Andrew Reschovsky and Jennifer Imazeki for the plaintiffs, which used 3 differing outcome standard definitions and a number of different policy judgments, concluded that between $1.65 billion and $6.17 billion would be needed to meet the 55% successful performance measure. After considering at length the methodological and judgmental differences between the studies, the trial court accepted the Reschovsky/Imazeki study.19
1. The original pattern of local funding of education emerged in an agricultural economy when wealth was relatively evenly distributed and property taxes provided the visible, tangible and predominant basis for taxation. Ellwood P. Cubberley, Public Education in the United States (2d ed. 1934) at 734.
2. For example, in 1979 the Tennessee School Finance Equity Study, a two-year in-depth independent analysis of tax and equality issues, found that the educational needs of the average fourth grade classroom was $832 per pupil, while the foundation formula provided only $318. Theodore Meyers, Thomas C. Valesky and Marilyn A. Hirth, K-12 Education Funding in Tennessee: Equity Now—Adequacy Coming (20 J. EDUC. FIN. 394,396, 1995). In New York State in 1997-1998, the foundation amount was $3900, even though the average district expenditure was $9,800 and even the lowest spending district in the state spent $7700. The State Education Department of New York, The State of Learning, A Report to the Governor and the Legislature on the Educational Status of the State’s Schools (2000).
5. Jimmy Kim and Gail L. Sunderman, Large Mandates and Limited Resources: State Responses to the No Child Left Behind Act and Implications for Accountability (The Civil Rights Project, Harvard University, February 2004) at 6.
7. The complex methodology actually used to carry out these tasks is summarized in James W. Guthrie and Richard Rothstein, “Enabling ‘Adequacy’ to Achieve Reality: Translating Adequacy Into State School Finance Distribution Arrangements” in Equity and Adequacy in Education Finance: Issues and Perspectives (Helen F. Ladd, Rosemary Clark and Janet S. Hansen, eds., Committee on Education Finance, National Research Council, 1999), 228-246.
9. These methodologies are described in more detail in Kern Alexander et al., Proposals for the Elimination of Wealth-Based Disparities in Public Education, Report to the Ohio Legislature (July, 1995), and John Augenblick and John Myers, Recommendations for a Base Figure and Pupil-Weighted Adjustments to the Base Figure for Use in a New School Finance System in Ohio (July 17, 1997).
10. Consistent with their understanding of contemporary needs as articulated by the standards-based reform movement, the courts deciding recent adequacy cases have defined “adequacy” as requiring education at “more than minimum level.” See, e.g., Brigham v. State, 692 A. 2d at 680; See also William H. Clune, The Shift From Equity to Adequacy in School Finance (EDUC. PLICY 376, 1994) at 8 (describing the thrust of the cases as calling for a high minimum level). This “high minimum” approach focuses on what would be necessary to assure that all children have access to those educational opportunities that are necessary to gain a level of learning and skills that are now required, say, to obtain a good job in our increasingly technologically complex society and to participate effectively in our ever more complicated political process.” Minorini and Sugarman, “Educational Adequacy and the Courts,” in Equity and Adequacy in Education Finance, supra n. 7 at 188.
11. As will be described below, “expert judgment” is essentially a variation of professional judgment and “cost function” can also be viewed as a variant of the statistical modeling that is at the heart of the successful schools approach.
13. American Institutes for Research (A.I.R.) and Management and Planning, Inc. (MAP), New York Adequacy Study: Providing All Children With Full Opportunity to Meet the Regents Learning Standards (March 2004).
14. Lawrence O. Picus and Associates, A State-of-the-Art Approach to School Finance Adequacy in Kentucky (February 2003). The model omits instructional aides “because research generally shows they do not add value.” Id. at 21.
18. See, e.g., William D. Duncombe and John M. Yinger, “Performance Standards and Educational Cost Indexes: You Can’t Have One Without the Other,” in Equity and Adequacy in Education Finance, supra n. 7 at 260; Andrew Reschovsky and Jennifer Imazeki, Achieving Educational Adequacy Through School Finance Reform (CPRE Research Report, Oct. 2000).
19. West Orange Cove Consolidated Independent School District et al. v. Shirley Neeley, Commissioner of Education, et al., No. GV-100528 in the 250th District Court of Travis County, Texas. The Court also chose not to rely on a professional judgment study undertaken by MAP, which also had been submitted by the plaintiffs.
Updated June 1, 2006