3112513 mm III/IIIWI////I/I4/4//I//4II//I/III/IIt {MAIL WWI-r -* 21 Unfifiafl' 254 This is to certify that the thesis entitled A STUDY OF COST AND INSTRUCTIONAL INPUT: DOES DOLLAR EXPENDITURE EQUITY EQUATE TO PROGRAMMATIC EDUCATIONAL OPPORTUNITY? presented by MICHAEL A. BOULUS has been accepted towards fulfillment of the requirements for Doctoral degree in Admin. 8 Higher Educ. Maw M}ajor professor Date__WA ° 0-7639 [I Iu:‘\\\‘ ‘ :33" “I,” \II” OVERDUE FINES: 25¢ per day per item RETURNING LIBRARY MATERIALS: Place in book return to runove charge from circulation records A STUDY OF COST AND INSTRUCTIONAL INPUT: DOES DOLLAR EXPENDITURE EQUITY EQUATE TO PROGRAMMATIC EDUCATIONAL OPPORTUNITY? By Michael A. Boulus A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Administration and Higher Education 1981' 1;qu «16.3 Alla ..\\ Q ABSTRACT A STUDY OF COST AND INSTRUCTIONAL INPUT: DOES DOLLAR EXPENDITURE EQUITY EQUATE TO PROGRAMMATIC EDUCATIONAL OPPORTUNITY? By Michael A. Boulus The purpose of this research was to attempt to answer the question of whether dollar expenditure equity equates to pro- grammatic educational opportunity for children. A means of defining educational equity was developed which reflects the instructional time and programs directed by the personnel most immediately associated with providing educational opportunity to students--the instructional staff at the school building level. The problem addressed in the research is the need to demon— strate the relationship of cost input differences with programma- tic differences and correlate these differences with other recog- nized variables of educational equity. Equity was measured in terms of programmatic personnel in- puts available to students. This staffing measure consists of noncategorically funded instructional and support staff. Per- sonnel factors at the elementary building level were quantita- tively weighted in terms of teacher experience and education in order to obtain a comparable measure of input equity and input units. The equity measure was analyzed against comparable cost data to answer the question: does educational cost difference measure programmatic differences? Michael A. Boulus The findings show the economically advantaged suburban districts have the most favorable measure of programmatic inputs per stu- dent among the twelve districts in Ingham County. The city school district of Lansing ranks eleventh among the twelve dis- tricts surveyed. However, the cost per input unit is highest in the Lansing School District leading one to the conclusion that greater spending does not necessarily yield a more favorable staffing ratio among the districts surveyed. While the urban school district of Lansing has the higher program input costs, there is less general program diversity and opportunity. The most obvious explanation is differences among the dis- tricts in staff compensation. The larger the membership and the more urban a district, the higher the average salary and the higher the cost per student. The findings further indicate that a district's wealth as measured by State Equalized Valuation per pupil is not a predic— tor of a district's instructional costs. Are then higher cost districts taxing themselves at a higher rate in order to provide comparable programs and services than their less costly counter— parts? The answer is yes. The correlation between cost per input unit and millage effort reveals that high cost districts are highly dependent upon a high tax effort. High millage districts are typically your high cost districts, and they are higher cost because of higher salaries. Three variables which describe the community in terms of potential educational needs were then analyzed to identify Michael A. Boulus communities likely to need extra services from their schools--ser- vices which add to the cost of education. The data revealed that Lansing had the highest indicators of special needs which the district must meet and the lowest capacity to provide additional services as measured by tax load in relation to income. Virtually all studies which seek to answer whether a system of school finance is an equitable one select local district wealth, and generally property wealth, as the discriminatory criterion of interest. The studies ask "are educational re- sources (i.e. revenues, staff, programs) distributed among the local districts according to local property wealth?" While this principle provides the basis upon which many state general membership aid formulas are based, it fails to acknowledge equity in terms of ability to pay, cost differen— tials and educational needs. Cost is the key determinant relative to how much a district must pay in order to provide comparable programs, staff and services. If equity is to be achieved, districts facing exces- sive costs over which they have little or no control must receive help from the state in meeting those costs so that an equitable proportion of funds will be available in each district to meet the proper needs of its students. DEDICATION To my loving wife, Marian, and to my wonderful parents whose support and sacrifice have made this possible, this volume is dedicated. ii ACKNOWLEDGMENTS This study would not have been possible with- out the guidance and assistance I have received during the period of my graduate study. I am especially indebted to Dr. C. Robert Muth who has served as my advisor and good friend. It was through his encouragement and counsel that the completion of this study became a reality. I express my gratitude to Dr. Philip Cusick, Dr. Samuel Moore and Dr. Kenneth White who served so ably as members of my committee. Finally, special thanks and appreciation to Richard Mancino for his research assistance and Connie Parker for the typing and preparation of the manuscript and for tolerating the extra work often at the expense of their own time. 111 TABLE OF CONTENTS CHAPTER PAGE I. THE PROBLEM...... .......................... 2 Introduction............................ 2 Need for the Study.... ...... .. .......... A Statement of Problem and Purposes...... 7 Research Procedures ..... ...... ..... ..... 9 Research Questions to be Investigated... ll Assumptions and Limitations............. 12 Isolating the Equity Function........... 12 Measuring Equity ............ ............ 1A Accounting for."Quality Differences in Personnel Factors"............ ..... .. 15 Source and Nature of Cost Variations Supply and Demand Functions.......... 16 Sample Size............................. 16 Definition of Terms..................... 17 Overview................................ 18 II. REVIEW OF THE LITERATURE............ ....... 19 Introduction..... ..... .................. l9 Judicial Standards of Equity............ 19 Distribution of Resources and Fiscal Need...... ....... .................... 3A Cost Differences and Educational Opportunity.......................... Al Measuring Equity........................ 53 Development of Comparable School Cost Data: Rationale for Isolating In- structional Expenditures............. 59 Summary.... ......... .................... 61 III. DESIGN OF THE STUDY ........................ 66 Introduction......... ..... .............. 66 Pepulation.............................. 66 Sample Selection ...... .................. 67 Instrument Development and Use.......... 68 Isolating Basic Educational Expendi- tures......................... ....... 7A Modifying State Per Capita Cost Data.... 76 Identification of Need Variables........ 80 Collection of Data...................... 81 iv CHAPTER IV. PRESENTATION AND ANALYSIS OF DATA.......... Introduction ........... ...... .......... Relationship Between Programmatic and Cost Inputs ........ .............. Cost Per Input Unit and Program Equity.. Cost and a District's Wealth............ District's Ability to Pay.... ..... ...... District's Need Variables............... Comparison of Published Per Capita Cost Data ................ . ........... Summary of Findings ........... .......... V. SUMMARY4OBSERVATIONS AND RECOMMENDATIONS... Summary Observations.................... Recommendations......................... Implications for Further Research....... BIBLIOGRAPHY.... Appendix A. Appendix B. Appendix C. Appendix D. Appendix E. Appendix F. Appendix G. Appendix H. 0.000000000000000 0000000000000000 Survey Instrument................. Select School District Data...... Introductory Letter............... Correlation Matrix................ Average Salaries by Enrollment Sizes.00000.000000000000000.0.0 Percent of Successful School Mil- lage Elections..... ....... ..... Staff Compensation and District 812800000000.000000000000000000 Comparative Increases in Tax Effort. ..... . ..... ............. 128 128 132 1A2 1414 IA? 150 153 15“ 155 156 157 158 LIST OF TABLES TABLE I. Ls) 1N. _ O \0 (EN O\U'l 12. 13. 1M. 15. 16. Basic School District Data Ingham County School Districts, 1979- 80 ...................... ..... ...... Select Data for Michigan's 530 K- 2 School Districts Grouped by Membership, 1979- 80 .................... Schedule of Input Units and Salary Costs by School Buildings and School Districts, 1978- 79....... ..... Students Per Input Unit by District.......... ......... Cost Per Input Unit by District....................... Correlation Matrix .................................... Average Teacher Salary by District and Size, 1978-79. SEV Per Pupil and Cost Per Input Unit by School Dis- trict..... ................. Operational Millage by School District, Type, Cost, Salary and Input Units ............ ................. SEV/pp, School/Community Millage Rate, Average Per Capita Income and School/Community Millage Rate Per $1,000 of Income by District................... Percentage of Title I Students by District............ Percentage of Students Obtaining Less than 75% of Reading Objectives on the Michigan Educational Assessment Program (MEAP) Test, 1978............... Student Socioeconomic Status by District.............. A Comparison of Operating Expenditures by District.... Correlation Coefficients for Select Per Student COSt Data. 0000000 0000 000000 00000000000000.000000000 Current School Aid Inequity .................. .. ..... .. vi 87 88 9O 93 9A 99 101 105 107 111 11A 115 116 121 123 136 PLEASE NOTE: page l seems to be missing in numbering only. Text follows. Filmedaas received. UNIVERSITY MICROFILMS. CHAPTER I THE PROBLEM Introduction Since the decision in Serrano v. Priest1 in 1971 hold- ing that the quality of education may not be a "function of the wealth of a pupil's parents and neighbors,"2 a number of states have passed major school finance reform legisla- tion. Common to most of them is a formula which recognizes that all communities do not have equal preperty tax bases. To compensate for these differences, poorer districts tend to receive greater financial assistance from the state while wealthier districts receive less assistance. In this manner some semblance of balance, at least in terms of funds avail- able to provide potentially similar programs for each child in the state, is achieved. Support which achieves equal dollars per pupil when state and local monies are combined,however, does not necessarily guarantee that equal amounts of goods and ser- vice can be purchased with each available dollar. 13:15 probable that the purchasing power of the dollar will differ by district or region, depending on factors such as size, pepulation density, distance from transportation centers, cost of living and similar variables. Due to these varia- tions, fluctuations in the cost of providing public 1. 5 Cal. 3d 584, 487 P.2d 1241, 96 CalRptr 601(1971). 2. Ibid., at 1244. 3 education will obviously occur. For example, where teacher negotiations have produced significantly higher salary sched- ules and have reduced student-teacher ratios, the cost of pro— viding teachers on a per pupil basis might be expected to in— crease. Similarly, sehool district differences and geographic differences may affect disparities in nonsalary costs. It is apparent that expenditures per pupil vary widely across the state and that school staffing ratios are similarly unequal. In 1977—78, for example, general instructional ex- penditures in Michigan ranged from less than $707 per pupil in some districts to more than $1,707 in others, while in- structional staffing ratios ranged from fewer thaan2 to more than 76 staff per thousand pupils across the districts.3 While there have been attempts to deal with the equal dollars aspect of school funding, the questions of whether equal dollars equates to a comparable measure of programma- tic equity has not been thoroughly studied. A major objective of the researcher in this study is to determine the extent to which cost differences across dis— tricts can be compared and correlated with variations in the purchasing power of the dollar as reflected by quantitative programmatic measures. In pursuing this objective, the researcher attempts to answer the question for the sample study: does educational cost differences measure programma- tic differences (equity differences)? 3. James L. Phelps, Michael F. Addonizio and Thomas S. Nicol, "State Aid for Michigan Public Schools: An Analysis of Two Proposals," Michigan Department of Education (mimeo), Lansing, Michigan 1979. Need for the Study Equal dollars, because of input cost variations, may not buy equal services. This basic assumption needs to be researched more closely as it has been ignored in most school finance litigation and educational policy decisions. To cite one instance, the Michigan state aid legislation purports to serve the basic purpose of equalizing educa- tional Opportunities. A number of considerations draw this stated purpose of the legislation into question. Even with a fiscally neutral state aid plan, wide variations in expenditures exist among school districts. This raises the subsequent question: Are expenditure differences related to local preference or to exogenous local conditions that force higher costs for the same or less services? This question is not fully answered by research or by the courts. The legal questions of equal educational opportunity has dealt, until recently, almost exclusively with equality of dollar expenditures. The cost of educational inputs has not been factored into the equity questions and consequently the assumption that equal dollars mean equal opportunity has prevailed. There is a difference, however, between the standard of equality of educational opportunity and the standard of fiscal neutrality. The concept of equality of educational Opportunity roughly stated means treating everyone in a manner befitting their needs. Fiscal neutrality, as advanced by John Coons of the University of California at Cl Berkeley, requires that each state create a finance system that provides every district the same number of dollars per pupil for the same tax rate.” As the subsequent review of the literature indicates, fiscal neutrality works well as a rough legal remedy. It is judicially manageable and does not attack the concept of local control. It does not, how— ever, address the larger questions of what differences other than wealth are legitimate and should be considered. If school districts pay different salaries to teachers and other personnel, as well as different ratios for non— personnel inputs, then clearly the ratio of expenditures to quantity or quality of inputs will vary. Therefore, expendi- tures per pupil are probably not a good index of programma- tic input per pupil. Recently the school finance reform movement has begun to move in a new direction. The Serrano approach to school finance reform, relying solely on variations in local prop- erty tax capacity, has slowed. Courts in Ohio and New York have ruled that a system with equal tax yield for equal property tax effort discriminates against the big cities. In effect, the state courts are intervening in issues that go beyond Serrano and its concept of equity. Courts and legislatures are now putting forward the elusive idea A. John E. Coons, William Clune and Stephen Sugarman, Private Wealth and Public Education, Harvard University Press, Cambridge, Massachusetts 1970. 6 of differential costs of education, but are unsure about how to identify and adjust for uncontrollable costs. As Senator Jason Boe, President of the Oregon State Senate and President of the National Conference of State Legislators, states: "Most legislators are not willing to retreat from their commitment to quality education, but quality can no longer be judged purely in terms of expen- n5 ditures. A major problem confronting researchers and policy- makers attempting to measure cost differentials is the lack of meaningful data provided by most local school districts. Such data are usually maintained on a districtwide basis rather than on a school-by-school or program-by—program basis. An additional problem is that researchers have encountered difficulty reconciling state and local records, thereby raising questions as to the accuracy and validity of records from other sources. While it has been contended that it costs more ix) pro- vide the same educational services in some districts, know- ledge of the extent and impact of these differences in terms of programmatic equity and the complexities of determining comparability are crucial issues for current discussions on school finance plans. If there is to be significant progress toward improving the equity and equality of 5. Jason Boe, "State Legislators and Quality Education" Com act, Summer 1979, at 27. E. K. Forbis Jordan, "Cost Differentials in State Aid Programs in Selected States," paper presented at the Annual Meeting of the American Educational Research Association, (Washington, D.C.; March 31 - April 3, 1975), at 2. educational opportunity, a method needs to be developed for relating operating expenditures to an acceptable standard of programmatic equity. Statement of Problem and Purposes Succinctly, the problem addressed in the research is a need to demonstrate the relationship of cost input differences with programmatic differences and correlate these differences with other recognized variables of educational equity. The purposes of this research are stated as follows: 1. To attempt to answer the question of whether dollar expenditure equity equates to programmatic educational oppor- tunity for children. In order to accomplish this purpose, a means of defining educational equity was developed. After a detailed review of the judicial standards and various methodological studies related to educational equity, the author concluded that a rational and quantifiable system is a measure expressed by instructional personnel inputs at the building unit (see Definition of Terms). Such a measure of equity reflects the instructional time and programs directed by the person- nel most immediately associated with providing educational opportunity to children--the instructional staff of the building unit. The population under study was a representative sample of 2A elementary schools in each of the twelve school dis— tI'icts in Ingham County. Personnel factors of the elementary building level were quantitatively weighted in terms of teacher experience and education in order to obtain a comparable meas- sure of input equity and input unit. The cost per input unit was computed by dividing the total salary of identified per- sonnel (via a direct review of payroll records) by the number of personnel input units. 2. To determine how current state cost data reflect dif- ferences in the cost of personnel input units across school districts. The author examined ways in which generally available cost data can be modified to represent more accurately input equity in terms of dollars. In order to obtain cost data that are reflective of the basic operational costs of the general program of a school district, a method of modifying state published per capita cost statistics was used. This allowed for greater uniformity and comparability of costs across districts. Because state generated data serves as the raw material for educational policy analysis, the question of how state computed current cost data represents realistic programmatic differences across districts becomes extremely important as a research topic. 3. To determine the correlation of recognized school district and community variables with the computed cost of programmatic variables for the sample districts. Educational literature increasingly references such variables relating to educational equity such as student need and taxpayer ability to pay. This research identified such variables and corre- lated them with per student cost and computed programmatic measures. The purpose of the study was thus three-fold. First, to develop a measure of program equity across school districts separate from cost data and to test such measures against cost measures to answer the question: does educational cost differences measure programmatic differences? Secondly to analyze and modify state published cost data and determine how the data reflect programmatic differences. And thirdly, to select and correlate district and community need measures with program and cost measures to determine the extent that need is reflected in programmatic measures. Knowledge of these issues is crucial to current discus- sions on school finance reform aimed at providing equal dol— lars for equal tax effort. This research is exploratory and developmental in nature. It is intended that it will contribute to a better understand— ing of cost variations for comparable services. If the find- ings can contribute to a better understanding of cost varia- tions for comparable services, it will enhance and provide new direction for further research in program equity as a goal of educational finance. Research Procedures The problems and purposes can be broadened in terms of the following procedural steps and methods: (1) A uniform method of weighting and pricing personnel factors (input units) at the elementary building level was constructed for use as a measure of comparing programmatic 10 equity across school districts. (2) Elementary school budgets were purged of categori- Cal expenditures and the programmatic input units they repre- sent in order to obtain a consistent measure of basic programma- tic opportunity. (3) The number of personnel units, the cost per per- sonnel unit, the personnel units per student and the cost of personnel input units per student were calculated for each of the sample buildings and school districts. (A) The data was analyzed to determine the relation- ship across schools between students per personnel.input unit and COSt per personnel unit. (5) State published cost statistics were modified to isolate basic educational expenditures per student. (6) State published district cost statistics as well as basic education cost data (as computed in number five above) were examined to determine the relationship of these data with the personnel unit cost per student as computed for the sample schools. (7) A series of independent variables not fully con- trollable by board policy were identified in an attempt to assess community needs which could have an impact on educa- tional cost. (8) The relationship between the dependent variable (cost per input unit) and each of the independent variables were analyzed to determine what significance existed between cost and district needs. ll Research Questions To Be Investigated (1) Do equal dollars equate to programmatic educational opportunity? (2) What is the relationship between personnel input units per student (as a measure of programmatic equity) and the cost for these units across select elementary schools? (3) Are the districts with the higher input unit cost per student the very same districts with the more favorable programs (as measured by personnel input units per student)? (A) What is the nature and magnitude of the cost varia- tions for comparable services among the sample schOol dis- tricts in Ingham County? (5) Can state published per student cost statistics for school districts be modified in order to reflect more accurately basic operational cost expenditures across dis- tricts? (6) What factors, other than personnel costs, are deter- minants in cost variations among school districts? (7) How does the level of personnel input units per stu- dent correlate with other sehool associated independent need variables? (8) To what extent has Michigan's state school aid formula distributed educational resources among local school districts so as to achieve equity and equality? l? Assumptions and Limitations Defining Equity The author seeks to provide a conceptual framework to correlate educational costs with a measure of programmatic equity. This requires a definition of programmatic equity. Ideally it would be desirable to have some universally accepted concept of educational equity which would permit the comparison of levels of programmatic services across schools. But neither the courts, policymakers nor researchers have established such a concept. For the analysis that follows, equity is measured in terms of programmatic personnel inputs available to students. This staffing measure consists of non-categorically funded instructional and support staff. Other measures of equity and need are examined through the course of the study, but the primary thrust of this research is the issue of input cost treatment of persons in like circumstances and the Second is the different or disparate treatment of persons in unlike circumstances. Isolating the Equity Function The analysis and discussion which follows assumes that programmatic equity can and should be studied in isolation from other factors. Others, such as H. M. Levin, have used a "capital embodiment" approach to suggest that the formal schooling process must be considered in conjunction with other activities that affect the student population such as medical services, dental services, nutritional services, etc.7 The framework that Levin supports is viewed to be valid but a partial view is taken here primarily as a start- ing point for developing a measure of equity in terms of whether state aid should be adjusted to recognize cost differences for comparable services. While it may be argued that equity should be viewed in relationship to the distribution of individual and socie- tal benefits that are produced by the educational system, there are two reasons why this is not appropriate for this study. One, the social and individual benefits from educa— tion are difficult to measure with an acceptable degree of accuracy. Two, it is almost impossible to trace the effect of a school finance plan in terms of societal bene- fits and educational inputs. Recognizing some shortcomings and limitations, the research supports the concept of measur- ing programmatic equity in terms of input units for the fol- lowing reasons: (1) Personnel input units are perceived as being related to the education process; (2) there has been growing support for a shift to a classroom input unit distribution system (e.g. Michigan State Senator Jack Faxon's Weighted Pupil approach for distributing state aid); and (3) unlike other measures of educational equity, per- sonnel input units are manageable, measurable and something that can be understood. 7. H. M. Levin, "Equal Educational Opportunity and the Distribution of Educational Expenditures," Education and Urban Society, Volume 5, 1973, at 1A9-l72. 8. Russel E. Wilson et al., The Equal Quality Plan, Pre- pared by the Task Force on Public School Finance, Lansing, Michigan, June 1969. 1H Measuring4Equity Programmatic equity is assessed in terms of a direct measurement of building level, instructional personnel input units per student. These personnel factors include teachers, support staff, instructional aides, building administrators and other instructional-related building staff. This research, then, measures only the building level cost of instruction input units. The author recognizes that in order to capture all cost variations across schools one must also account for differ- ences in the prices of such components of the budget as \ instructional materials, construction costs, land prices, interest charges, overhead, etc. Total school district costs are used to determine the relationship of programma- tic equity as determined through the author's measurement to the traditional per student cost measures. The major problems associated with state statistics reflecting total costs such as the inclusion of categorical expenditures and the difficulty in isolating non-instructional expenditures to individual schools are reviewed in Chapter Two. The review of the literature which follows in the sub- sequent chapter indicates that variations in the prices of non-salary school inputs are a less significant source of educational cost variations. This supports the design of this study which focuses on instructional salaries, which amount to approximately 80 percent of the building level budget. This researcher isolates cost related to direct programmatic benefits to students. Accounting for "Quality Differences in Personnel Factors" One of the problems which needs to be addressed in uti- lizing personnel inputs as a measure of the equity function is how to assess the differences in "quality" of personnel inputs. For example, as all teachers are evidently not alike, a simple teacher count could possibly be misleading. In order to quantify "quality" differences, the assump— tion is made that the education and experience level of teachers should be recognized in measuring the equity function. The research is mixed as to whether teacher inputs, such as educa- tional training and experience, make a difference in deter— mining outputs, and if so, how important are they? Although difficult to measure, there is a general acceptance that training and experience contribute to teacher competency. This is reflected in almost every master agreement in this state which recognizes differences in experience levels and training and compensates teachers accordingly. Inasmuch as overall instructional expenditures can be attributed to cost differences resulting from variations in teacher experience, education and teacher-pupil ratios, such differences will be controlled in order that comparable resources can be examined. To establish a teacher equity measure, a procedure has been developed for weighting and unitizing these personnel factors. The procedure is de- scribed in Chapter Three of this study. l6 Source and Nature of Cost Variations Supply and Demand Functions It is hoped that an additional outcome of this research would be to provide a better understanding of the source and nature of the variations in the cost of providing comparable instructional services. Specifically analyzed are variables Selected to identify communities likely to need extra services from their schools--services which add to the basic cost of education. The assumption made is that state aid formulas ShOUld not recognize cost variations solely attributable to local demands but only those exogenous to the schobl district. Sample Size This study is not intended to be a statewide examina- tion of the cost of education in Michigan schools. Rather, this researcher seeks to study a model for examining cost differentials in terms of a comparable measure of programma- tic equity. It is hoped that the model and approach used will contribute to a better understanding and awareness of the variations in programmatic equity and input costs per student. I Because of the developmental nature of the model and because the type and nature of data to be collected are not easily attainable, on-site elementary school visitations were required. The researcher has limited the school survey to the twelve school districts in Ingham County. The Ingham County school districts provide a meaningful grouping in terms of size, geographic structure, community and expendi- l7 ture levels (see Appendix B). A representative sample of 2A elementary schools from the selected districts constitute the research sample. Ingham County is very diverse economi- cally and socially. The twelve school districts of the sample include the core city of Lansing with a 28 percent minority student population, five suburban districts and six rural districts. The districts range in student size from 29,260 in Lansing, to 900 in Webberville. Definition of Terms "Instructional Related Staff" -- The following are the instructional staff recognized in the personnel input count: (a) Classroom teachers (b) Auxiliary Instructional Staff (such as music, physical education, certified librarians, coun- selors, aides, etc.) (c) Building principals and assistant principals. "Personnel Input Units" -- A weighted measurement re- flecting the number of instruction-related staff at the elementary building level whose salaries are derived from the general fund budget purged of categorical revenues (i.e. local revenue, state aid membership). The basic unit is the measure weighted in terms of experience and level of education. "Personnel Input Units Per Student" -- The number of personnel input units divided by the number of students in the regular membership. "Cost Per Input Units" -- Total salaries of identified 18 personnel in a building divided by the number of personnel input units. "Operating Expenditures Per Student" -- General fund expenditures of the district pertaining to education (as found in MDE Bulletin 1012) divided by the number of students in the regular membership. "Basic Operating Expenditures Per Student" -- General fund expenditures of the district purged of categorical funds and special allotments, divided by the number of students in the regular membership. "Equity" —- Equal treatment of equals. This principle states the requirement that equals should receive appropriately equal treatment. This definition recognizes that district costs and needs may not be alike and as such should receive disparate treatment. Overview The survey instrument found in Appendix A provides an overview of how building level personnel inputs were factored to provide a comparable measure of programmatic equity across schools. Chapter Two provides a review of the literature pertinent to the study. The design of the study and the design of the instrument is provided in Chapter Three. In Chapter Four, the results are analyzed. In Chapter Five, survey observations, conclusions and recommendations are made. CHAPTER II REVIEW OF THE LITERATURE Introduction The review of the literature is decribed in this chap- ter under the following headings: Judicial Standards of Equity, Distribution of Resources and Fiscal Need, Cost Differences and Educational Opportunity, Measuring Equity, Development of Comparable School Cost Data, Rationale for Isolating Instructional Expenditures and Summary. Judicial Standards of Equity The 1970's have witnessed an increase of activity in the area of school finance reform. Much of this reform has been influenced, indeed compelled, by litigation begun at the state and federal levels which questioned the equity of current distributions of state monies to finance education. The resultant court cases have, in many instances, provided the impetus for reform of the state school finance plans; however, the judicial progress in this area has not been smooth. Its decisions punctuated with a number of conceptual stops and starts, the judiciary has made its way haltingly to the goal of educational equity. Not until recently has the judiciary demonstrated a greater willingness to entertain, in its deliberations of 19 20 the constitutionality Of state finance plans, the more subtle and problematic aspects Of educational Opportunity and educational equity. Prior to the New York decision in Levittown v. Nyquist the judicial response to questions Of school finance equity has been marked by two significant trends. The first trend is the transition in judicial reason- ing from the use Of a simplistic negative standard approach cast in egalitarian terms and which considers only absolute dollar equality toward a more encompassing approach which takes into account variations in costs and educational need. A second trend is the abandonment from school finance litigation Of federal constitutional principles as a viable reforming force. Instead, the challenges to state finance plans increasingly rely upon state constitutional provisions related to education and the provision thereof. The two trends are related. The gradual abandonment of the fiscal neutrality standard in favor of one which accounts for cost and need variations is at least partly due to the movement Of litigation from federal courts back to state courts which, unlike their federal counterparts, are more aware of and amenable to the unique characteristics of their respective states. State courts, perhaps because Of their proximity, seem more willing to apply more sophis- ticated standards Of adjudication than are federal courts. Initial efforts to Obtain judicial relief in the area Of school finance were unsuccessful probably because the Pl plaintiffs presented to the courts issues not well suited for judicial resolution at least at that time. In McInnis v. Shapiro9 the plaintiffs challenged the Illinois school finance system as being in violation of the equal protection and due process clause of the fourteenth amendment. They argued that the system provided some students with a quality education while depriving others with equal or greater educational need of the same. The disparities in educational expenditures were the result of a greater ability and willingness to pay. The plaintiffs demanded, therefore, that "expenditures" be made only on the basis of pupils' educational needs with- out regard to the financial strength of local school dis— tricts."10 Complaining that educational need was a "nebulous concept," the Federal District Court declared that there was no "discoverable or manageable standards by which a court can determine when the Constitution is satisfied and when it is violated."11 Similarly, the federal district court hearing Burruss v. Wilkersonl2 upheld the Virginia system of school finance. The court, relying on McInnis, noted that it had "neither the knowledge nor the means nor the prowess to allocate the public monies to fit the varying needs of these students 9. 293F.Supp. 327 (N.D. 111.1968), McInnis v. Shapiro, 39A U.S. 322 (1969). 10. Ibid. at 336. 11. Ibid. at 335. 12. 310 F.Supp. 572 (W.D. Va. 1969), Burruss v. Wilkerson, 397 U.S. an (1970). throughout the state."13 i The courts clearly were reluctant to adjudicate an issue to which they felt themselves ill-equipped to apply a norm such as educational need. The debate and disagree- ment which surrounded the concept Of educational need at the time certainly worked against the court's acceptance of it as a judicial standard. The courts needed a workable norm the provision of which had not occurred and, conse- quently, the courts presumptive deference to legislative enactments carried the greatest weight in their delibera- tions. In areas like school finance, more a policy than a legal area, the courts are dependent on the litigants and eXperts in the field for information. In the absence of such information the prudent alternative is to refuse to act. As John Coons has pointed out: "Without experience -- without phenomena delivered to the judicial mind -- there is nO substance for the work Of adjudication."14 Hence i inaction. The breakthrough that allowed the courts to assume an active role in school finance reform came in 1969. A work- able judicial norm was delivered to the courts in the form of a study published by John Coons, William Clune and Ste- phen Sugarman entitled Private Wealth and Public Education.15 13. Ibid,, at 574. 14. John E. Coons, "Recent Trends in Science Fiction: Serrano Among the People Of Number," Journal Of Law and Edu- cation, Vol. 6 at 24, 1977. 15. ., William Clune and Stephen Sugarman, Pri- vate Wealth and Public Education, Harvard University Press, Cambridge, Mass., 1970. In it Coons and his colleagues suggested that the equal protection be interpreted not to require any particular funding level tied to educational need, but to prohibit allowing the level of funding to vary as a function Of such irrational factors as the accidental location Of taxable wealth. Denominated as Proposition 1 in Private Wealth and Public Education, the principle stated that "the quality Of public education may not be a function Of wealth, other than of the state as a whole."16 Now known as theory Of fiscal neutrality, the principle requires that each school district receive the same revenue for the same tax rate and that only those expenditure dis- parities resulting from differences in local district taxable wealth are unlawful. Fiscal neutrality provided the courts with a legal theory that seemed easier to grasp and one with which they could adjudicate school finance disputes without becoming bogged down in the issue's complexities. In 1971 the California Supreme Court formally adopted and articulated the principle in its landmark decision, Serrano v Priest.17 The court found for the challengers of the California school finance system, holding that it dis- criminated against the poor in violation Of the equal protection clause Of the California and federal constitu- tions "because it makes the quality of a child's education a function Of the wealth Of his parents and neighbors."18 16. Ibid., at 339 17. 5 Cal. 3d 584, Serrano v. Priest, 96 CalRptrGOl (1971). 18. Ibid., at 1244. The Serrano decision touched Off a footrace Of school finance litigation. As of the 1973 decision by the Supreme Court in Rodriguez, some 59 cases had been filed in more than 30 states. The fiscal neutrality principle appeared to provide the courts with flexible (i.e. judicially manage- able) standard by which tO judge the constitutionality of 19 state finance schema. Van Dusartz v. Hatfield followed a month-and-a-half after Serrano and, in adOpting the Serrano rule for Minnesota, the federal district court stressed the flexibility the case, and its principle, made possible. Because of the early success Of fiscal neutrality, a flurry Of judicial activity followed between 1971 and 1973.20 Nearly all Of these decisions followed the reasoning of Serrano and relied on the fourteenth amendment. The progress Of litigation seemed unyielding when the United States Supreme Court's decision in San Antonio v. Rodriguez applied the brakes and brought to a premature halt this phase of the school finance litigation. The Serrano decision and its immediate descendants had broken the connection between community wealth and educa- tional expenditures, yet fiscal neutrality proved not to be 19. See, Parker v. Mandel 344 F.Supp.1068 UL.Md.l972); Rodriguez v. San Antonio Independent S. District, 337 F.Supp. 280 (W.D. Tex. 1971); Van Dusartz v. Hatfield, 4O U.S.L.W. 2228 (D.Minn. 1971); Hollins v. Shofstall, Civ. NO. C-253652 (Ariz. Sup.Ct., June 1, 1972); Caldwell v.Kansas, Civ. No.50616 (John- son CO.Dist.Ct. , Kansas, Aug.30, 1972); Milliken v. Green, 389 Mich. 1, 203 N.W.2d 457 (1972), vacated, 390 Mich. 389, 212 NW2d 711 (1973). 20. 4O U.S.L.W. 2228 (D. Minn. 1971). 95 the egualitarian panacea its proponents had envisioned. Rely- ing solely on variations in local property tax capacity, the Serrano approach ignored variations in pupil needs or expendi- tures. A negative standard, fiscal neutrality is "content to condemn particular and manifest injustices, leaving definition of the good society to the political process. It merely says no to a particular order of things--wea1th discrimination in education; to every other order it offers a presumptive yes."21 To paraphrase Coons, then, to understand fiscal neutrality is to discover what it ain't, "and principally it is not."22 Fiscal neutrality as an instrument of reform lacks refine- ment; it cuts away the most visible, surface forms of inequity and discrimination only to reveal more intricate forms beneath. This is, perhaps, characteristic of adjudication in any area of pressing social or economic concern. Granted that when major societal problems have not been resolved in the political process the judiciary customarily fills the void, yet the judiciary is often.re1uctant to enmesh itself in cases demanding creative improvisation, preferring instead to stay its hand until no other avenue is available. Once commitment is unavoidable, the judiciary is more apt as not to limit the breadth of its decisions so as to defer again to the legislative-political process. Judicial activism is 21. John E. Coons, "Fiscal Neutrality After Rodriquez," Law and Contemporary Problems, Vol. 38, 197A, at 300. 22. Ibid. 26 not typical of the Supreme Court; even the most judicially active court Of the century, the Warren court, tempered its desegregation mandate in Brown v. Board of Education with the order Of "all deliberate speed." Thus arises the need for a judicially manageable standard like fiscal neutrality which allows courts to address themselves to issues which Obtain generally with clear perimeters, leaving the framing of a comprehensive remedy to the legislative branch. This reluctance surfaced in San Antonio v. Rodriguez23 where, by a five to four margin, the U.S. Supreme Court ruled that it was inapprOpriate for the Court to intervene on con- stitutional grounds in school finance. The Court was asked to overturn the Texas system because, plaintiffs argued, it discriminated against a "suspect" class, depriving them Of a fundamental right to education. The Court, however, found that the Texas system did not discriminate against any class Of persons considered "suspect" since the case dealt with prOperty poor districts, not poor persons. Moreover, the Court pointed out that the deprivation complained Of was relative, not absolute. Arguing that "the marketplace of ideas is an empty forum for those lacking basic communicative skills" and "a voter cannot cast his ballot intelligently unless his read- ing skills and thought processes have been adequately developed,"24 the plaintiffs held that a fundamental 23. San Antonio Independent School District v. Rodriguez, 411 U.S. 1, 36 L.Ed. 2d 16, 93 S.Rptr. 1278 (1973). 24. Ibid., at 35-36. 27 constitutional right to education was implicit in the right of free speech and in the right to vote. The Court answered this claim that education is the sinepqua non of an effective practice of these other guaranteed rights by stating that "we (the Court) have never presumed to possess either the ability or the authority to guarantee to the citizenry the most ef- fective speech or the most informed electoral choice."25 Education is neither explicitly nor implicitly mentioned in the Constitution, and though it is an important service, its importance does not determine whether it is a fundamental right. With neither suspect class (i.e. a party believed to be rightly chargeable with something wrong) nor a fundamental right involved the Court found no reason to apply strict scrutiny, which demands that the state show a compelling interest in maintaining the practice in question. The Court, therefore, applied the less exacting rational basis test and decided that the Texas system, though "chaotic and unjust,"26 was nonetheless rational and furthered a legitimate state pur— pose--that is, to encourage "a large measure of participation in the control of each district's schools at the local level."27 Explicitly symptomatic of the Court's reluctance to tackle a problem with such policy implications, the Court deferred action to the legislative process: "the ultimate solutions must come from the lawmakers and from the demo- 25. Ibid., at 35. 26. Ibid., at 58. 27. Ibid., at 28. EC C1: 28 cratic pressures of those who elect them." Even the minor- ity justices seemed reluctant to embark on a constitutional condemnation of the system. Of the four dissenters, only two, Justices Marshall and Douglas, accepted the plaintiff's constitutional argument. The other two, Justices Brennan and White, dissented on the ground that there was no rational relationship between the statutory scheme and the goal of local control. "It is worth observing," wrote N. R. Anderson in the Washington Law Review,29 that in Rodriguez the Court was asked to enunciate principles of educational finance which would be binding on all fifty states. Given the vast differ- ences in the educational finance systems of the various states, it is perhaps not surprising that a majority of the Court was reluctant to confront these complexities. Justice Powell, indeed, stated as much. Obviously, a much more manageable range of issues is presented in a case before a state court. The Rodriquez decision effectively closed the doors to challenges on a federal level of state finance schema on the basis of the fourteenth amendment's equal protection clause. Less than a month after the Texas decision, however, the New Jersey Supreme Court overturned that state's school finance system on the ground that it violated the state's constitutional command to the legislature to provide a "thorough and efficient system of free public schools."30 28. Ibid., at 59. 29. N. R. Anderson, "School Finance in Washington," Washington Law Review, Vol. 50, 197A, at 853-900. 30. New Jersey Constitution, Article VIII, Section A, Paragraph 1. :9 This marked a significant tactical change -- the resort to state constitutional provisions conveyed the message that even though the federal courts were closed, the state courts Offered an alternative instrument for school finance reformers. In addition to "thorough and efficient" clauses, state equivalents to the federal equal protection clause provided another means whereby reformers could challenge inequitable state finance systems. In Serrano II the California Supreme Court upheld a state trial court which, on remand of Serrano I had held that education was indeed a fundamental right under the state constitution, thus requiring the state to satisfy the requirements Of strict scrutiny. The California court rejected the claim that the Rodriguez decision affected the trial court's determination of a fundamental right, replying:31 Decisions of the United States Supreme Court defining fundamental right are persuasive authority to be afforded respectful considera- tion, but are to be followed by California courts only when they provide no less individual protection than is guaranteed by California law. In Connecticut the Supreme Court of that state held that education was indeed a fundamental right both explicitly and implicitly, thus meeting the standards for strict scru- tiny Of both the U.S. Supreme Court and the California court. Their decision in the case, Horton v. Meskill,32 struck down the state's finance scheme as unconstitutional. 31. Serrano v. Priest, 18 Cal.3d 728, at 764. 32.172 Conn. 615, 376 A.2d 359. 5’0 In summary, though the U.S. Supreme Court's decision in the Rodriguez case caused a slackening in school finance litigation, reformers soon discovered the state courts to be an amenable forum for their views. Another very important result Of the Supreme Court decision was the prevention of a blanket acceptance on federal constitutional grounds of the fiscal neutrality principle. By 1973 fiscal neutrality was still not yet subject to the kind Of discussion and de- bate which would ultimately point up its shortcomings. 'Phe variety Of state cases and the arguments and discussions generated by Rodriguez revealed the inherent restrictions Of fiscal neutrality. Critiques in law and education journals argued that, as a negative standard, fiscal neutrality lacked the sephistication to deal with the more complex questions of inequity which were coming to judicial attention post- Rodriguez. Its simplistic approach allowed for an easily understood and readily grasped standard for measuring the constitutional inadequacies of state school finance laws, but ignored the variations of state conditions. The most pointed criticism leveled at fiscal neutrality and the finance plans born Of it point to its inherent limi- tations -- its inability to account for variations in cost and educational need among districts. As Berke and Callahan argue, this is most striking with respect to cities:33 33. Joel S. Berke and Callahan, "Serrano v. Priest: Mile- stone Or Millstone," Journal of Public Law, Vol 23, 1972. ...cities have higher costs for providing education in terms Of the prices paid for educational items. Land acquisition costs, insurance rates, vandalism expenses, and non- professional personnel costs all reflect higher costs of living in central cities. But bulking largest in school budgets are costs Of instructional personnel, and here a combination Of factors has pushed a central city costs well above those in the suburbs. It is no longer accepted that the crude formulation, equal dollars equal services, holds true. Therefore, in response to such criticisms, the school finance reform move- ment has moved creatively in new directions. Having gone as far as it could with the fiscal neutrality approach, the courts have given renewed life to educational need after re- jecting need as tOO complex and nebulous in the late 60's. Indicative Of the trend toward a special dispensation for cities in school finance systems is the New York case, Levittown v.Nyquist.34 The Supreme Court of Nassau County affirmed the concept Of municipal overburden and ruled that, under a formula with equal tax yield for equal effort, the capacity Of large city school districts to finance education with local resources had been overstated. The city school districts had charged that the state aid formula discrimi- nates against them by overstating their fiscal capacity and not accounting for their special needs and higher costs. The Court upheld the claims that there is a drain on city tax dollars caused by the need to provide excessive amounts of non-educational services, that the purchasing power of the 34. NO. 8208/74 (Nassau County Supreme Court, New York). ..L) \) dollar in metropolitan areas is diminished, and that large cities have the greatest proportion of low-income and edu- cationally disadvantaged students. The court said:35 The asserted basis for the contention is that the extraordinary problems stemming from urban poverty and its associated ills in turn create an array of educational overburdens that the large cities have been unable to copy with by resort to their own fiscal resources and which are essentially disregarded by the state aid formula...a1though they have the largest con- centration of disadvantaged children requiring compensatory education, they are unable to supply that help, thus resulting in a denial to those children of an equal educational opportunity. The court then concluded that:36 Where it is demonstrated that there is a greater need, it follows that a greater amount of aid must be furnished in equal educational opportunity is to be available. Something more than average aid to school districts must be furnished to accomplish this goal. Reading the New York Constitution as demanding that all children of the state be given "the opportunity to acquire at least these basic skills necessary to function as a citi- zens in a democratic society,"37 the court found that the state finance scheme did not meet the rational basis test. The court argued instead that the system failed to remedy inequalities in educational opportunity and, in fact, per- Petuated them.38 By use of the rational basis test the court was able to strike down the finance scheme on the basis of both the New York and the federal constitutions in compliance 35. Ibid., at 65. 36. Ibid., at 99. 37. Ibid., at 100. 38. Jean Scheuer, "Levittown v. Nyquist: A Dual Challenge" Phi Delta Kappan, Vol. 20, Feb., 1979, at A35. with the Supreme Court's ruling in Rodriquez. The same outcome, that of adjusting for higher costs and municipal overburden in cities, was reached in an Ohio case involving the city of Cincinnati. Again, on the surface at least, Cincinnati seemed to be an anomalous choice as plain- tiff when one considers its high assessed property values. Here an Ohio trial court agreed that "the large urban inner city districts have unique and special problems and substantial extra costs which the non-urban districts do not have."39 The court, therefore, determined that in order to comply with the state constitution, any state finance system must include as One of its criteria adequate compensation for the special costs of urban school districts. Plaintiffs' arguments in both the New York and Ohio cases stress the impact which greater educational need has on the capacity of urban school districts to provide serv— ices. Recalling the earlier difficulty courts had with the concept of educational need, Betsy Levin cautions that cur- rent concepts of educational need should be careful to avoid the pitfalls of ambiguity and judicial manageability evi- denced in McInnis and Burruss.“0 She goes on to say:ul The Levittown intervenors have attempted to avoid this problem by showing that the high con— centration of urban poverty in the inner cities results in high proportion of children suffer- ing from such education overburdens as impaired learning readiness, impaired mental, emotional and 39. Cincinnati, slip opinion at 117. A0. Betsy Levin, "Current Trends in School Finance Reform Litigation: A Commentary," Duke Law Journal, Vol. 1977 at 1119. Al. Ibid. physical health, mental retardation and other learning disabilities, and English language dif- ficulties. The intervenors then showed that specific programs or services have been success- ful in overcoming these problems, but that because Of inadequate resources these programs and services have to be curtailed. Levin questions whether courts will find constitutional imperatives to take into account the relative educational needs Of students. Perhaps courts will recognize relative need only in relation to constitutional violations of a dif- ferent sort, such as segregation or finance inequity, while not giving it a constitutional niche of its own. Whatever the difficulty in assessing educational need or Of determining constitutional status, the fact remains that urban school districts are faced with increasingly high costs, educational and non-educational, which affect their ability to buy educational services. This is only now being recognized by the courts. The trend away from the simple fiscal neutrality approach is very much a part of this new awareness. The negative standard which said that educational Opportunities could not be a function of district property wealth has been broadened to read 'educational Opportunities may not be a function Of district fiscal capacity which, in the case of cities, means that fiscal capacity must reflect the additional costs and municipal overburdens characteristic of cities as well as property wealth. Distribution Of Resources and Fiscal Need The preceding analysis of the judicial response to school finance has attempted to highlight the limitations Of state school finance plans attempting to implement the Serrano principle. The equal protection clause encompasses but one aspect of the injustice inherent in the way elemen- tary and secondary schools are financed in most states Oftflua nation. The judicial concept Of wealth discrimination, so central to the holdings Of Serrano and Rodriguez, outlines the problem of wealth discrimination but fails to deal with the whole concept. Vfilflithe exception Of Levittown v. Nyquist the judicial focus fails, on the one hand, to provide an appropriate comparative measure Of fiscal cost and, on the other hand, fails to examine the total tax responsibilities (municipal and educational) Of various jurisdictions. While the courts have been successful in creating the conditions for change in school finance patterns, they have not assured that the form will inevitably channel more resources to school systems faced with greater fiscal need. The chief cause Of the pattern Of inequitable distri- bution Of funds is state policies which create school districts responsible for supporting education but endowed with vastly different abilities to raise those revenues. As a result we find different levels Of education existing in neighboring school districts independent of the degree Of effort the communities expend in raising revenues for their schools. H. Thomas James made the following Observation on the implications of this phenomenon:42 42. H. Thomas James, "Diagnosis & Treatment for Ailing School finance, " A paper presented for the annual meeting of the Council Of Chief State School Officers at Louisville, Kentucky, Nov. 15, 1971, at 1. 3O When I completed the series Of school finance studies that occupied my attention at Stanford through much Of the last decade I announced what I thought was the most significant generalization to come out Of those studies. That generalization was that local property tax paying ability was the major determinant Of social policy for public edu- cation, and that unless we could reverse that equation and let social policy determine what should be spent for education we would have deep trouble in our schools. An equitable system has been described by Joel Berke and John Callahan as one in which "greater educational resources are allocated to those students who come to school with the greatest learning problems and the greatest social disadvantage." Further, they described equal educational Opportunity as "as allocation of educational services which enables pupils from low socio-economic backgrounds to compete equally for higher educational and job Opportunities with those who come from more advantaged walks Of life."43 In addition to equitable distribution Of educational services, Berke and Callahan contend also that an equitable system must recognize that such services take a greater toll from those jurisdictions and individuals less able to pay than from those who can better afford such costs. State financial assistance, consequently, needs to be adjusted to recognize cost differentials and ability to pay}4 A study conducted by the Urban Institute for the Presi- dent's Commission on School Finance documented a strong regression pattern in six of the eight states it studied 43. Joel S. Berke and John J. Callahan, "Serrano v. Priest: Milestone or Millstone for School Finance, "Journal Of Public Law, Vol. 21, 1972, at 34. 44. Ibid., at 39. 37 relative to tax burden and income levels. For example, Michigan families earning less than $2,999 per annum paid 8.7 percent of income in state and local taxes. For those earning over $15,000, the figure was 3.5 percent. Just as there is a mismatch in educational need and' ability to pay, so too there seems to be a mismatch in educa- tional costs. Betsy Levin, in her study "Public School Finance: Present Disparities and Fiscal Alternatives," notes that cities have higher costs for providing education in terms of the prices paid for educational items. Insurance rates, vandalism costs and, most important, personnel costs all reflect the higher cost of living in central cities.“6 The higher non- educational expenditures of central cities also have had the effect of creating total tax disparities between central city and the suburb. Urban education costs have had to respond to suburban cost pressures. According to Dr. Seymour Sacks, "the common metropolitan environment, which in past years had a salutary effect on central city education, now has a debilitating effect on central city education as the metropolitan areas determine the level of costs without providing the resource for meeting these costs."u7 This is especially true regard- ing salaries, which account for a full 80 percent of a A5. Betsy Levin, Public School Finance: Present Dispari- ties and Fiscal Alternatives, The Urban Institute, Washington, D.C., 1971 at Al. op cit., at 59 A7. Seymour Sacks, David Ramsey and Ralph Andrew, City Schools - Suburban Schools: A History of Fiscal Conflict, Syracuse University Press, Syracuse, New York,l972 at 61. 38 school system's operating budget. Bothwell notes that subur- ban districts also exert cost pressures as a result of their lowering of class size, which translates into hiring more teachers and other instructional personnel to teach the same number of students, which further increases total salary A8 costs. Compounding the problem has been the relative decrease in income which cities have suffered vis-a-vis their metro- politan suburban neighbors. By 1959 the proportion of fami- lies with incomes under $3,000 was almost 50 percent greater in central cities than in their outlying areas, and the dis- parity even greater for Standard Metropolitan Statistical A9 Areas over one million. By 1967, median family income was only $7,813 in central cities and $9,367 in outlying metro- politan areas.50 Thus, when considering income alone, cities would have to levy an additional 20 percent tax burden on their incomes in order to offer educational standards on a par with their suburban neighbors. State educational formu- las, however, generally have not recognized or compensated for income differences among communities. J. Allan Thomas made the following comments in his study "School Finance and Educational Opportunity in Michigan:"51 A8. Robert Bothwell et al., "State Funding of Urban Edu- cation under the Modern School Finance," Manual, Department of Health, Education and Welfare, Washington, D.C., Jan. 1976, at 5. A9. Ibid., at 65. 50. Ibid., at 61. 51. J. Allan Thomas, School Finance and Educational Oppor- tunity in Michigan, Michigan Department of Education, Lansing, Michigan, 1968 at 20. It is necessary to include in a state aid formula a procedure for defining the share of the costs which will be paid by the locality. This involves the development of some type of "index of local tax—paying ability." This is a much-debated issue which cannot be resolved on any "scientific" basis. Rather, the values of the citizens of the state will determine the manner in which the local contribution to the foundation program will be determined. The following possibilities exist: a. The local contribution should be based on a given yield on the system of taxation from which the local contribution is obtained. If the property tax remains the basis of this contribution, a given millage on the property would continue to provide the local share. b. On the argument that all taxes eventually are paid from income, the local contribution would be determined on the basis of the average income (rather than the average property value) per pub- 1ic student in the taxing area. If the local contribution is to be obtained from a property tax levy, the millage required for the local share of the foundation program would need to be weighted, on the basis of the per capita in- come in the community. 0. The composite index of income and property might form the basis of calculating the local share of the foundation program. However, some means must be found to permit low valua- tion districts to improve their programs so that the present variations in opportunity may be narrowed. In the central cities of the 38 largest SMSA's in 1966-67, per capita taxes for non-educational services amounted to $1AA, whereas in the surrounding suburban areas the per capita taxes for these services were only $76.52 Seymour Sacks notes in his report, "Central City and Suburban Public Education," that during l96A-65 the central cities in thirty-four metro- politan areas spent on the average $50 more per citizen in total tax dollars (municipal and educational services) than 52. Sacks, op cit., at 66. AO did the surrounding suburbs. However, these central cities 53 spent an average of $50 less per person on education. As schools are in keen competition with other municipal services for tax dollars, the level of school support is affected. Alvin Townsel observed while studying the school dis— tributional patterns in sixteen states that reforms in school finance legislation most frequently have been effected through indirect means. Economic and sociological concomitants which impinge on the fiscal problems of urban schools most ipfggf guently have been of major consideration.5n Adjustment in state basic aid formulae, while. responding to variations in property wealth throughout the states, often have failed to realize that the central city fiscal difficul- ties are not attributable to this factor. They are, by and large, attributable to higher costs for both educational and non-educational public services. Speaking to the additional burdens of central cities and disparities in property wealth, J. Allan Thomas said:55 State aid in Michigan is sizable in total quan- tity. It does not, however, serve its basic pur- pose, which is to equalize educational opportunity. Even after state aid, there are wide variations in the expenditures of local districts. The low valuation districts tend, on the average, to have higher tax rates for education and lower expendi- tures than the high valuation districts. 53. Seymour Sacks, "Central City and Suburban Public Edu- cation: Fiscal Resources and Fiscal Realities," Metropolitan- ism: Its Challenge, The Sixth-Seventh Yearbook of National Society for the Study of Education, University of Chicago Press, Chicago, Illinois, 1960. 5A. Alvin H. Townsel, "Adjustment for Urban Fiscal Prob- lems in State School Finance Systems," Journal of Educational Finance, Vol. 2, Summer 197A, at 70. 55. Thomas, op cit., at 207. A1 He continues: Michigan's major urban centers have additional revenue problems. Faced with a declining tax base, the need to finance a wide array of govern- mental services, mounting costs, and the presence or large numbers of students who have special educational difficulties, Michigan's cities are in dire need of special consideration. Federal funds are helping to some extent, but the state of Michigan has a responsibility to help meet these urgent needs. One of the objectives of this study was to develop a measureable standard that recognizes expenditure variations in the cost of educational inputs. If school finance systems are ultimately to move toward the goal of equality inputs, then what is needed are adjustment indices that reflect price differences. Hopefully, this study can shed some light on the policy question of whether states can and should adjust their formula aid to achieve both dollar and input equity. Cost Differences and Educational Opportunity Several studies were reviewed which look at the features of school districts which account for differences in the prices paid for the goods and services purchased. The under- lying assumption in all these studies is that one dollar in state aid does not necessarily purchase the same amount of educational services. The studies emphasize features which determine differences in the salaries paid for these services. Essentially, the problem is to select variables that account for differences in salaries paid for teachers of like quality in various districts. These studies demonstrate A2 both the possibilities and limitations of statistical analysis as a means of estimating appropriate adjustments designed to neutralize local cost differences for state aid purposes. In the study "Financing Education in Michigan,"56 Dr. C. Robert Muth cites several variables not controllable by board policy that are important determinants of salary levels, in- cluding training and experience. Another such variable cited which influences salary levels is cost of living which varies from area to area and may be reflected in professional sala- ries. Schedule history, market location and stability of the system influence salary costs, as do community.differ- ences as they relate to job desirability for teachers, and all are beyond the control of the district. Thus, teacher salary costs, which represent nearly 90 percent of instruc- tional costs, are subject to minimal control by local boards and should, therefore, be recognized in state funding. How- ever, differing salary costs across school districts create different input prices for personnel services for children. Muth concludes that:57 If equity in financing education is to be achieved, districts facing excessive costs over which they have little or no control must receive help from the state in meeting those costs so that an equit- able proportion of funds will be available in each district to meet the program needs of its students. 56. C. Robert Muth and Paul R. Stuhmer, "Financing Educa- tion in Michigan: A Report," a study funded by the Ford Foundation through the Middle Cities Education Association, Michigan State University, East Lansing, Michigan, 1977. 57. Muth, op cit., at 67. J. Alan Thomas, in his study Of school finance and educational Opportunity in Michigan, sought to measure edu- cational Opportunity. The Thomas staff looked at "quality" programs by district size, geographical region, district valuation, and expenditure per pupil. He isolated five determinants Of educational equality:58 1. Quality in education is related to the development Of programs apprOpriate to the economic, social and political context of the society. 2. The number of children with whom a teacher works is an important element in the total education environment, particularly at the primary and pre- school levels. 3. The most important single element in the educa- tional process is the teacher - the quality and efficiency of teacher services are important determinants Of educational excellence. 4. Variations in the quality of school buildings may be realted to a community's view of education as a form of consumption. 5. The usefulness Of television, teaching machines, and language laboratories has not yet been adequately evaluated. The Thomas study indicates that availability of programs and services was more related to the number Of students than either valuation or cost per student. Thomas made a very comprehensive study of the Michigan teaching force. He looked at age, experience and training differences as well as supply and demand factors, average salaries and pupil-teacher ratios. The study reported that regional characteristics are more significant than wealth in determining these teacher characteristics. Thomas found that there is also a high degree of correlation between 58. Thomas, Op cit. at 18-19 AA school district size and the average teacher's salary. The larger the system, the higher the mean salary paid to its teachers.59 Indicating that 8A percent of school district expendi- tures were for salaries, Thomas emphasized the relationship of salaries to wealth, size, location and urbanization. His study concludes that "the quality and quantity of the teach- ing staff is by far the most important single consideration in determining how youngsters will be taught."60 According to Thomas, then, the criteria for quality is number of teachers per student, not dollars per student. The dollar represents both differing services and costs for the same services. The Task Force on Public School Finance of the Michigan Association of Professors of Educational Administration pub- lished a suggested plan for financing K-12 education in Michigan in 1969.61 The proposal, known as the "equal quality plan" advocates assuring and equalizing education opportunities appropriate to each child within the state and emphasized not equal money but equal opportunity. Vocational and special education would be assured for all those students requiring or desiring such programs wherever they lived and educationally deprived urban or rural children would have access to special programs needed to overcome their depriva- tion. 59. Ibid., at 108. 60. Ibid., at 106. 61. Wilson, op cit., at A6. TO accomplish equal Opportunity the group suggested a classroom unit approach to financing education. The basic assumption involved here is that teacher-pupil ratio 5A; the major factor determining quality of education. Given this assumption the task force proposed varying teacher-pupil ratios for different groups of students with different needs. For example, they suggested a ratio CHE 1:40 for kindergarten, 1:25 for regular pupils, 1:15 for vocational education, and 1:15 for educationally disadvantaged pupils. Under the equal quality plan funding was to be based on classroom units which were to be arrived at by claSsifying and counting pupils and then determining the number of teacher positions needed to arrive at the specified pupil- teacher ratios. Funding was to be based on the number of teacher positions with each district receiving an amount derived by multiplying the number of teacher positions needed by a state established professional salary allowance. Funds for other services such as administration, inservice, main- tenance and non-professional salaries were to be based on a percentage of the professional salary level and the number of classroom units. The plan also recognizes a second quality variable - teacher experience and training - which is directly related to cost. A teacher experience and training index was de- veloped in recognition Of the variations in cost Of teacher salaries in various districts. The index was based on an average teacher having a bachelors degree and eight years Of A6 teaching experience. Such a teacher was assigned a value Of 1.0 and teachers with more experience and training were assigned higher valuations and vice versa. From these values a district average index value was to be derived and applied as an adjustment tO the total number of teaching positions for state aid calculation. This kind Of index assumed that teaching quality is related to teacher training and experience and the index was designed "to make it pos- sible for all school districts to compete on a more nearly equal basis for well qualified teachers."62 Under the "equal quality plan," no attempt was made to identify or adjust for cost differences existing in differ- ent regions or types Of districts for similar programs. The authors suggest, however, that a factor dealing with salary variations by educational region might be added to the professional staff index if creditable data could be derived. In a 1974 study in Michigan,63 Harvey Brazer pointed out that while Observed differences among school districts in levels Of expenditure per pupil may or may not be asso- ciated with similar differences in such measures Of output as achievement test scores or dropout rates they clearly differ "as a consequence Of differences in prices paid for major inputs such as teachers (salaries) and/or differences 62. Wilson, Op cit. at 48 63. Brazer, "A Study of Expenditures Levels and Student Output Measures Among School Districts," Selected Papers in School Finance: 1974, U.S. Office of Education, Washington, D.C., 1974 at 108. in the quantities of inputs used (pupil-teacher ratio)." He notes that to the extent these prices vary, neither an equal dollar distribution Of funds among school districts nor an equal yield per unit Of tax effort will result in equality in terms of inputs. "If such approaches are to attain . . . (equality) . . . some means must be found for allowing appropriately for price differences among districts." Brazer also suggests that traditional sparsity or den- sity corrections and adjustments which weight certain grade levels more heavily than others are "aimed rather vaguely at educational costs as a whole rather than at prices Of inputs as variables of cost element." Brazer concludes that cost of living adjustments are only one element entering the supply function for teachers. Brazer's study was based on the assumption that since teacher's salaries account for a high prOpertion of district current Operating expenditures, if measures Of relevant in— fluences on teachers' salaries can be derived an adjustment index for this crucially important input element can be developed. Brazer was able to account for 72 per cent of the variance in average teachers' salaries in the districts studied through the use Of 15 a priori identified supply and demand factors. The methodology is dependent on satisfactory a priori judgments of variables which appear to be related to differences in price. The achievement of greater pre- cision in estimating elements related to price is a chal- lenge Offered by Brazer's study. In a follow-up study conducted a year later, Drs. Brazer and Anderson used multiple regression analysis to measure the influence Of selected factors in local price differences for educational services or local costs for school districts. The study is divided into two parts. In the first, factors which affect instructional costs are analyzed, and in the second part non-instructional cost dif- ferences are studied.64 Since teachers' salaries are the major component Of local instructional costs, much of the analysis focuses on those features of school districts which affect the level Of . teachers' salaries. Brazer emphasizes,however, that state aid to a school district should not reflect adjustments based on a district's willingness to pay higher salaries for teachers. Any state aid adjustments to school districts relating to cost must be based on those district features which affect the willingness Of teachers to work there (i.e. supply condition). Since it is unclear as to what "quality" teachers are or what measures are relevant to teacher quality, Brazer excludes considerations of quality in his analysis. As a result of an analysis Of a number Of variables, Brazer is able to account for about 70 per cent of the variation in average teacher salaries among districts. Having completed his analysis of the factors which influence 64. Harvey E. Brazer and Ann P. Anderson, "A Cost Adjust- ment Index for Michigan School Districts", University of Michigan, Ann Arbor, 1975. “9 the willingness of teachers to work in a school district, an adjustment index for teacher salaries was constructed. The index indicates to what extent the combined effect of the actual values Of those district features taken into account which do bear on teacher salaries in any district differ from the combined effect of the average values for all the sample districts. If state aid were to be adjusted according to a salary index based upon district features over which it had no con- trol, i.e., features affecting the supply conditions or the willingness Of teachers to work in a district, the following pattern emerges. All central cities would gain state aid; all but one rural district would lose state aid; and among independent cities and suburbs no distinct pattern emerges. Brazer next develOps an index for non-instructional expenditures using the same methodology. He adds, however, that the two indices are independent of each other and that using a teacher salary adjustment index does not require the use Of an index for non-instructional expenditures. W. Norton Grubb, when given the same problem Of study- ing the applicability Of adjustment in state education aid for local price differences, wrote his paper entitled, "Con- structing Teacher Cost Indices: Methodological Explorations with California Unified School Districts."65 Grubb states 65. Norton W. Grubb, "Constructing Teacher Indices: Method- ological Explorations with California Unified School Districts," mimeo, Education Finance Reform Project, Los Angeles and the Childhood and Government Project, University Of California, Berkeley, 1975. there are three factors which comprise instructional cost for any school district: (1) salaries paid teachers of a given quality; (2) teacher quality (as expressed by training and experience), and (3) teacher-pupil ratio. Grubb first develops a theoretical model identifying the essential nature of the teacher supply and district demand curves for teachers but concludes that such a model is not useful because it does not enable him to isolate the teacher supply curve for any district. He concludes that his only recourse is to use statisti- cal analysis of reduced form equations to analyze the impact of teacher supply and demand on district wages, teacher quality, teacher-pupil ratios and expenditures. This multiple regres- sion analysis was the necessary first step in order to elimi- nate cost adjustments reflecting district demand variables. Grubb proceeds to a consideration of the variables in district features which have an impact on the demand and sup- ply for teachers. The variables are essentially those used by Brazer. The impact of these independent variables are studies separately for each of the three major areas of instructional costs, namely, on teacher salaries, on teacher quality and on teacher-pupil ratios. What particularly concerns Grubb is the small effect that the various independent variables have on the available measures of teacher quality, i.e., on teacher degree status or graduate credit hours and years of teaching experience. He states, "the abysmally low explanatory power of these regressions indicates that variations in these measures of 'quality' are largely unexplained by any of the independent variables available... It appears, therefore, that these ...measures do not reflect teacher quality at least in the sense of measuring teacher attributes which school districts attempt to increase."66 What should also be emphasized is that these attributes are precisely the ones over which a school district has no control once a teacher is hired. Teacher staffs inevitably acquire years of experience so long as they retain their positions, and individual teachers will also accumulate graduate credits if the incentives to do so exist as in the form of salary differentials attached to graduate credit hours or to degree status. In his concluding section, Grubb notes that statistical analysis does Offer the means of calculating local cost ad- justments. However, he is concerned with some of the prob- lems associated with his methodology. As does Brazer, he notes that the independent variables account for 70 percent of the variations in teachers' salaries. And he also ob- serves that not all of the relevant variables such as union strength may be desirable for inclusion in an index. Poli- tical decisions would be required with respect to the in- clusion of variables such as district's unemployment rate and the proportion of school-aged children to a district's population. In the end, Grubb emphasizes that the redis- tributive effects of such aid adjustments should be carefully 66. Ibid., at 36. weighted especially if the net effect is to take away from the poorer districts to benefit the wealthier ones. In the paper entitled "Measuring Differences Among Florida School Districts in the Cost of Education: An Alter- native Approach,"67 Lawrence Kenny, Irving Goffman and David Denslow reported the results Of their statistical analysis based on Florida data. The "alternative approach" in their title refers to the use Of statistical analysis to adjust state aid for local cost differences. The current practice in Florida is to price a list of commodities and services in various parts of the state and to adjust state aid for local price differences based on the difference between the local prices and state average prices. Florida is the only state which adjusts state aid to school districts on the basis Of estimates Of local consumer price levels. The assumption this study is predicated upon is that teachers Offered the same wage in different districts do not respond solely to living costs in choosing jobs but do in fact take into account other features Of school systems. The authors of the Florida study diverge markedly from Grubb and Brazer in the variables which they believe affect the supply and demand of teachers. In their view district demand for teachers or willingness to pay differing teacher rates is influences solely by the degree status and years of experience of teachers. All other variables are viewed as 67. Lawrence W. Kenny, David Denslow and. Irving Goffman, "Measuring Differences Among Florida School Districts in the Cost of Education: An Alternative Approach," University Of Florida, Gainesville, 1975. hr“) \.'7 affecting the supply price of teachers, i.e., the wage at which teachers are willing to work in any district. In the analysis, the authors conclude that such factors as a district's cost of living, population density, pupil test scores and a measure of the crime rate all influence the willingness of teachers to work in school districts. As a surrogate measure for the cost of living, the authors use the price of land in any district plus the degree of urbani- zation of that district. The final section of their paper contains a critique of the present Florida price of living index. The Florida index concentrates on the cost of living and ignores differ- ences in local property wealth. Measuring4Equity The measure of equity most commonly used is normally represented by revenues or expenditures, usually on a per pupil basis. Thus, to measure the equity of a number of com- ponents related to school budgets comparison are made on the inequity of per pupil revenues or expenditures across dis- tricts or students. As Robert Berne points out in his study, "Alternative Equity and Equality Measures: Does the Measure Make A Differ— 68 ence?" one difficulty with state school budget measures is its failure to recognize that schools purchase different / 68. Robert Berne, "Alternative Equity and Equality Mea- sures: Does the Measure Make A Difference?" Selected Papers In School Finance, U.S. Department of Education, Washington, D.C., 1978 at 6. I‘M amounts of educational inputs (teachers, textbooks, etc.) in different areas due to price variations. As a result, expenditures or revenues from different districts may not be comparable.69 One way to overcome this difficulty is tO adjust the school budgets according to different input prices. The problem with such a method is that actual price differences reflect supply and demand factors. In theory only supply factors should enter a price adjustment index, but the supply and demand factors are difficult to untangle. Furthermore, prices should be compared for inputs of similar quality and valid quality measure of inputs, such as teachers, is not currently available.70 The recent research on price indices indicates that there are two categories of adjustments that can be used in an attempt to translate dollars into equal buying power. (hue technique is to estimate a price index for each location -- a local"basket of goods" approach. Kenny discusses this practice for Florida where it is used to adjust the state aid formula.71 Berne notes, however, "that there are technical problems encountered in the computation of such an index and its questionable whether it is the apprOpriate index for educational inputs."72 69. Ibid., p. 6. 70. Ibid., p. 6 l 71. Kenny, Op cit. at 197. 72. Berne, Op cit. at 21. 55 The second approach is to adjust the salaries or costs according to factors which influence the supply of inputs as determined through econometric analysis. The primary prob- 1em here is the identification of the relevant factors and the adequate measure of controls. Kenny compares the Florida "basket of goods" index and Brazer's econometric analysis approach applied to Florida and found "noticeable differences among (SIC) between them."73 Grubb and Hyman display the correlation among differenct in— dices based on a range of supply/demand assumptions and find that "alternative assumptions can yield substantially different indices."7u A third approach considered—~and the one used in this study-~18 the direct measurement of educational resource inputs such as teachers, textbooks, supplies, etc. Berne notes that there are two problems that must be overcome in order to measure the inequity of a number of distributions using resource inputs as the measurement of the equity function:75 First, it is very difficult to assess the differences in quality for a type of resource input such as teach- ers. This measurement problem makes the task of con— verting the resource inputs of a given type to a common measure rather complex. In order to measure resource inputs we should find a way to take qualit1g4 differences into account. For example, all teachers are not alike and a simple teacher count could possibly be misleading. The second problem is the conversion of different types of resource inputs such as teachers, administrators, supplies, etc., to a common measure 73. Kenny, op cit., at 200. 7A. Grubb, op cit., at 116—118. 75. Berne, op cit., at 8. r4 (4 in order to obtain a per pupil measure of re- sources for a school district. Since human inputs form such a large percentage of educa- tional resource inputs, we may be content to only measure these but even here we must find a common denominator for teachers, teacher aides, administrators, etc." In order to overcome the problems noted by Berne in measuring equity vis-a-vis a direct assessment of resource inputs, two procedural methods have been incorporated into this study. Relative to the problem of assessing differences in the quality of resource inputs, a system of weighting per- sonnel input units has been developed to account for variances in a teacher's education and years of experience. The second problem Berne notes relative to converting different types of resource inputs to a common measure has been avoided by this researcher by isolating the equity func- tion in terms of building level instructional personnel inputs only (teachers, auxiliary staff and building administrators). Other, less significant cost variations have been avoided. In order to obtain a common per pupil measure the total building level personnel inputs (as weighted) will be divided by a common denominator—-the number of students in regular membership in that particular building. The resulting common measure of input equity is called "students per personnel input unit." Finally, it should be pointed Out that the equity function is sometimes measured in terms of educational out- puts. Such as, for example, achievement test scores, func- tional literacy tests, or educational attainment measured in years of education or degrees. In the context of school DY finance reform, it seems questionable whether such an ap- proach should be used. The question which is always at the center of such research is whether there are any inputs which make a difference in determining outputs and, if there are, how "important" are they. Recent summaries of the literature on this question do not point to any con- clusions any different from those Of a few years ago. In short, there are no consistent relationships between inputs and learning outcomes.76 Leonard Stavisky in his article, "Dilemma for Decision Makers: Contradictory Education Research,"77 com- piled a listing Of seventeen variables that frequently have been cited as possible determinants of education outcomes. From other literature in the field Stavisky selected from representative survey reports. Some of the disparate find- ings are summarized below: Level of Teachers' Education Study 1. In 83 percent of the studies, the more highly educated the teachers, the more impressive the students performance. Study 2. Only 12 percent Of the studies showed a significant relationship between teachers' education and pupil performance. Study 3. The percentage Of teachers having some graduate training had the most consistent relation- ship with pupil achievement. 76. Byron Brown and Daniel Saks, "Production Technologies and Resource Allocations Within Classrooms and Schools: Theory and Measurement." Paper developed under contract from the National Institute of Education, Michigan State University East Lansing, 1978, at 15. 77. Leonard P. Stavisky "Dilemma for Decision Makers: Con- tradictory Education Research," Compact, Summer 1979 at 24. Study A. Extra education credits among teachers do not appear to boost student learning. Years of Teaching Experience Study 1. Teacher experience was found to be related to student performance only 57 percent of the time. Study 2. More than a dozen studies examined the relationship of teachers' experience to pupil per- formance--about half indicated a direct association. Study 3. The Regression Studies revealed no relation- ship; the Outliers Study, on the other hand, showed a positive relationship between teacher experience and student performance. Study A. Differences in teaching experience have an impact on achievement growth. Class Size Study 1. Overall, class size was found to be sig— nificantly related to student performance on only 37 percent of 19 studies. Study 2. The finding in five studies showed that smaller classes matter for young children but pro- duced little evidence that they matter at higher levels. Study 3. The effect of class size may be dependent on the type of student and the subject taught. Study A. Elementary pupils below grade level gain in classes with fewer than 28 students, but other students can be in classes of over 28 pupils. The contradictory findings are understandable for sev- eral reasons. First, the objective systems for measuring performance of students are not sufficiently comprehensive to cover all these complex variables. Second, such systems for measuring educational quality are often interpreted in- correctly and, consequently, face resistance. Third, the methodology for linking educational outputs of students to financial resource allocation is not sufficiently well 78 developed to apply to school finance reform. Development of Comparable School Cost Data: Rationale for Isolating Instructional Expenditures As members of the Chicago public school staff worked on the development of comparable per pupil expenditures, two problems immediately became apparent:79 (1) It technically was not possible to isolate and charge to an individual school all of the direct costs involved in educating the children in that school. (2) Numerous variables existed in individual schools which distort the per pupil expenditures and re- sulted in figures which were not comparable from school to school. . In addition to the problem of being unable to isolate certain expenditures so that they could be charged to indivi- dual schools, it also became apparent that numerous variables exist in individual schools which distort the per pupil ex- penditures and result in figures which are not comparable from school to school.80 For example: (1) The more special education classes there are in a school, the greater the per pupil expenditure be— cause there are more teachers for a given number of children. (2) When a school services as a resource center or of- fice for teachers who serve a large number of schools, the per pupil expenditure level for the school will be abnormally high if the salaries of the itinerant teachers are charged to the school. 78. Stanley A. Rumbaugh, "Cost of Education Differentials: The Search for the Missing Link." Presented at the American Education Finance Conference, San Antonio, Texas, 1977; Lansing, Michigan, mimeographed, at 15. 79. Jean M. Raymond, "Financing Equity Among Schools in Large Cities, Chicago Public Schools: A Case Study." Paper presented at the American Educational Finance Conference, Nashville, Tennessee, March 1979, at 5. 80. Ibid., at 7. (3) High school branches which share building with elementary schools are likely to show low per pupil expenditures because the ele- mentary school rather than the high school branch is charged for all of the custodial and maintenance salaries and supplies. (4) Schools with government funded programs which Operate within the school day are likely to have inflated per pupil expenditures in rela- tion to the expenditure level for other schools. (5) The inclusion Of monies spent for such major permanent improvement items as toilet or electrical rehabilitation also distort per pupil expenditures since these are one-time expenditures and are not related to the prOgram Of education in the same way that teacher salaries or textbooks are related. Thus, variables, such as described above which exist from school to school make it difficult to secure per pupil expenditure figures which may be compared with one another to determine whether or not some schools are receiving less or some schools more Of the funds available for education. As Raymond points out:81 It became apparent that an improved instrument for de- cision making in relation to per pupil expenditures has to be develOped. Since over 80 percent of the total Operating fund is expended for salaries, and in working with the data it became Obvious that, in spite Of many variables, it is the differences in the salary levels in a school which determine whether the per pupil expenditure is high, low, or average, it was determined that salaries could be used as the basis for calculating per pupil costs which would be comparable from school to school. Further, a study of the Urban Institute in Washington, D.C.82 shows that there is little difference between the instructional expenditures of one district or another, 81. Ibid., at 8. 82. Betsy Levin, Thomas Muller and Corazon Sanavol, "The High Cost of Education in Cities," The Urban Institute, Wash- ington, D.C., 1973. Cl whether urban, suburban or rural (except for the very high transportation costs of rural districts, which in turn is offset by the very high building and maintenance costs of cen- tral cities). The main point regarding cost differentials-~as pointed out by Robert Bothwell in his study of the Texas school finance system--"has to center around instructional salaries, which 83 amount to 80 percent of operating budgets of school systems." Summary The review of the court cases dealing with educational equity and the literature review of educational cost equity research, together with the writer's educational finance ex- perience and legislative involvement, lead this writer to the following observations: 1. The courts have not addressed the equity question in terms of input costs and programmatic opportunity and only recently have addressed the cost-need question of educational equity. The legal question of equity has dealt with equity of dollar expenditures, and disparities in the cost of educational inputs have not been factored into the equity questions. The assumption that dollars equate to programs has, for the most part, prevailed. 2. The data generally available and usedfor educa- tional finance research begin with gross per student costs. 83. Robert O. Bothwell, "The Inequities of the Texas School Finance System." Paper presented at the HIU—Minority Tax Re- form Project Training Institute, Houston, Texas, September, 1973 at 17. ()2 The problem with this cost measure is that it includes cate- gorical expenditures such as special education and compensa- tory education. Thus a district with a high percentage of categorical funding will have its per capita costs skewed so that gross costs do not reflect costs across comparable populations. There is an urgent need for truly comparative data for educational finance research. This need is illustrated by the U.S. Office of Educa- tion's funded research to develOp a method for determining the added cost of vocational education. It was found that contamination Of gross cost data by categorically funded programs made it impossible to compute the cost Of regular secondary education and,therefore,impossible to compute the added costs of vocational education. Computing the specific costs of vocational education was not difficult but comput- ing the cost Of the regular program across districts was impossible because of the inclusion of the costs for widely differing categorical programs. The same problem pervades educational finance research. An adjustment formula is needed to compute the cost Of education free Of categorical costs that serve specifically qualified individual students and do not have an impact on the general program of the school. 3. School districts vary in the school and community climate they present to teachers and these factors affect the personnel selection of the district. But today this is a qualitative factor and not a quantitative factor affecting price. Salary schedule differences do not reflect today's conditions but rather historical conditions. Teacher sala- ries are determined by market conditions in unionized states (such as Michigan), and these markets were generally estab- lished during the decade Of the sixties when there was com- petition for teachers and when working conditions, ability to pay, and school board policies affected salary schedules and average teacher salaries. Dr. C. Robert Muth showed in his Ford Foundation study that variations in teacher salary schedule differences across districts in Michigan were rela- tively unaffected by professional negotiations because school districts, teacher unions, mediators and arbitrators used comparable markets on which to base their decisions. Union- ization has increased salary levels generally but not eliminated salary schedule differences between groups Of districts. There is a lack Of knowledge about or an unwil- lingness to accept the concept of educational labor markets outside of the control of board policy, and so the reality Of educational unionization has not been adequately built into educational finance research. 4. Because educational salaries are lower in rural and non-urban areas, the political strategy Of outside areas is to Oppose any cost difference consideration; after all, the state politics of educational finance is as much how the pie is cut as it is the overall size. Any new factor must give to some districts in disproportion to others. Resear- chers themselves are prone to negate cost differences because 6A Of their fear that more funds for high cost districts will increase the disparities. NO acceptable formula for cost recoqnition has been develOped and no general climate of re- ceptivity exists for cost differential recognition in educational finance. 5. The research on the significance of quality factors in education is very mixed. Teacher variables as they relate to experience, training and class size show no consistent or substantial correlation with student achievement. The human qualities for success in teaching have not been sufficiently isolated to provide for adequate measurement either4for selec- tion or evaluation. The research on class size as a quality factor is mixed although lower class size for the disadvan- taged students seems to be supported. School climate and building leadership seem to bear a positive relationship to student attitude and achievement, but as yet the unique in- dividual, organizational and community variables that contri- bute to the environment have not been adequately isolated and certainly not priced. The limitations of present re- search emphasize the need for better research in first iso- lating and then integrating the personal and organizational quality factors that are essential to success in the educa- tional enterprise. 6. There are standards of quality in terms Of program- matic input which are generally recognized and can be quantified. Qualified teachers, professional training beyond the B.A. and lower teacher-pupil ratios are generally accepted quantitative measures of quality. These factors have been built into categorically funded programs in vocational edu- cation and special education. If such standards of quality are recognized for individual programs such as those men- tioned above, they can be accepted as research assumptions applicable to all educational programs. An overall summary of the literature indicates that if school districts pay different salaries to teachers and other personnel with similar credentials, as well as different prices for non-personnel inputs, then it only holds that the ratio of expenditures to quality of inputs will vary. It follows, therefore, that expenditures per pupil do not serve as an accurate measure of educational equity. There is need for research which addresses the question of how dollars equate to services. It will be the purpose of this study to establish such a base for understanding the differences in cost and educational inputs. CHAPTER III DESIGN OF THE STUDY Introduction The design of the study and the order of research are described in this chapter under the following headings: Population, Sample Selection, Instrument Development and Use, Isolating Basic Educational Expenditures, Modifying State Per Capita Cost Data, Identification of Need Variables and Col- lection of Data. Population The population under study is comprised of the twelve school districts in Ingham County. They are: East Lansing Public Schools, Lansing Public Schools, Dansville Public Schools, Haslett Public Schools, Holt Public Schools, Leslie Public Schools, Mason Public Schools, Okemos Public Schools, Stockbridge Public Schools, Waverly School District, Webber- ville Public Schools and Williamston Community Schools. The sample school districts provide a meaningful grouping in terms of size, geographic structure, community and expendi— ture levels. The population used to respond to the instrument is broadly distributed among the fourteen student membership divisions used by the state in grouping Michigan's 530 K-l2 school districts (see Appendix B-l). The Ingham County School Districts also provide a meaningful grouping in terms of other classifications 66 67 relevant to this study. Relative to the geographic struc- ture, the population consists of one central city, five suburban cities and six rural districts. Millage rates for operating the schools vary from 27.0 mills in Leslie and Mason to 38.8 in East Lansing. Operating expenditures range from $1,187 to $1,710 per pupil, thus pro- viding a range well above and below the state average of $1,AlO. The twelve communities represented are also diverse in terms of social and economic characteristics and in the range of public services provided to its constituents. The 1978-79 basic school district data (see Appendix B42) further exemplify the variations in the twelve communities relative to SEV, membership and tax rates. Sample Selection The author deals specifically with elementary buildings in the stated school districts. A sample of the elementary school buildings within the districts has been selected to provide for full representation of the uniqueness of each district. It should be noted from the following schedule that in those districts where less than the total number of elementary buildings are included in the sample, the largest units are selected as representative of the district. It is assumed that the sample selection will provide for optimum comparability across schools and school districts. The sample was selected as follows: 68 Number of Cumulative School Elementary Sample No. of Schools District Schools Selection In Sample Dansville 1 Select only choice 1 Webberville 1 Select only choice 2 Leslie 2 Select largest 3 Williamston 2 Select largest A Haslett 3 Select largest 5 Stockbridge 3 Select largest 6 Okemos A Select two largest 8 Holt 5 Select two largest 10 Mason 5 Select two largest 12 Waverly 5 Select two largest 1A East Lansing 9 Select three largest 17 Lansing 36 Select seven largest 25 Total: 86 2A This research is exploratory and developmental in nature. The author attenmtedto study cost in relationship to a compara- ble measure of programmatic equity. If the findings can con- tribute to a better understanding of cost variations for com- parable services, it will enhance and provide new direction for further research in program equity as a goal of educational finance. Instrument Development and Use As noted in the review of the literature, a principal source of confusion in previous analyses of school expendi- tures has been the failure to make a clear distinction be- tween spending differences among school districts which are due to variations in cost and those which result from dif- ferences in the level of service. In order to avoid this problem, the survey instrument was designed to determine the level of programmatic services provided to elementary grade students and the cost of providing such services. 69 The single greatest factor for predicting educational cost is personnel inputs which account for 80 percent of build- ing level expenditures. In light of this, the instrument is designed to measure programmatic services in terms of person— nel input units per student at the building level. Personnel inputs include classroom teachers, principals and supervisors, and other instructional and instructional-support personnel such as guidance counselors, librarians, teacher aides, etc. Differences in other instructional expenditures, such as those for supplies and textbooks, can be assumed to be due exclu— sively to the quantity of those items, as books, pencils, lab equipment and projectors and vary little in price nationally.8u The problem encountered in using personnel factors for determining comparability of programmatic services, is one of how to quantify differences in personnel resources in terms of programmatic comparability. The research clearly points to three major factors affecting the purchasing power of the educational dollar: pupil-teacher ratios, teacher experience and teacher education.85 While differences in pupil-teacher ratios, where they exist, are easy to quantify, differences in education and experience levels of teachers are not so clear cut. As stated earlier, the literature is mixed relative to whether students in schools with more experienced and better 8A. Betsy Levin, et al., The High Cost of Education in Cities, The Urban Institute, Washington, D.C., 1973, at 22. 85. Ibid., at 1A. 7O educated teachers receive a qualitatively better education. On the other hand, these are fixed costs in the sense that the tenure system leaves districts little choice as to whether such teachers are returned or replaced. Even though there are disparate findings regarding the relationship between educa- tion and experience of teachers and pupil performance, these two factors need to be recognized due to the realities of the present tenure system, district salary schedules and general perceptions. Districts with higher proportions of teachers at the upper level of the salary schedules can do little to reduce their costs regardless of whether it can be demon- strated that such teachers, overall, make a difference in educational quality. The research on the quality (student achievement) value of lower pupil-teacher ratios is also mixed. There is evi— dence that lower pupil-teacher ratios make a difference if ratios are lowered substantially (15:1) but little or no dif- ference around a mean of 27:1. However, teacher contracts continue to stress lower ratios as an improvement in teach- ing environment and an implied value to students. Further, these factors are recognized and built into categorically funded programs, such as special education and vocational education. To what extent, then, should differences in teacher characteristics be recognized in the development of models for determining a comparable measure of programmatic services? On the other hand, there is not consistent and clear evidence 71 to date that teachers with additional education or experience provide a better quality education than teachers with less experience and training. Yet, both labor and management generally accept as a rational factor in salary schedule con- struction that the more experienced teacher with an advanced degree, overall, does make a difference in educational quali- ty. This is recognized through the automatic award of salary increments for additional years of service and education beyond a Bachelor's Degree. A generally accepted quantitative measure was desired for reconciling the discrepancy between performance value for and professional recognition of advanced training and experience. Several leaders in school finance and instruc- tion were asked what maximum weight they would recognize for training and experience as a measure of teacher performance and program quality. Based on the interviews, a measure of 30 percent was arrived at for recognizing the maximum combined value of teacher experience and training. A teacher with a B.A. and less than one year experience was defined as one in- put unit and all other instructional inputs were weighted in relationship to the base input unit up to a maximum of 1.3 for a classroom teacher with 13 or more years experience and a M.A. degree or better. This approach considers the common sense belief that professional people grow, the lack of supportive research and the discussion with practitioners in the field. Fewer years of experience and hours of educa- tion were prorated as follows: 72 Teacher Experience Teacher Training Over a B.A. Years Weight Quarter Hours Weight 1-3 .03 1—11 .03 A-6 .06 12-22 .06 7-9 .09 23-33 .09 10-12 .12 3A-AA .12 13-15 .15 A5+ .15 Auxiliary instructional staff members (i.e., art, music, physical education, librarian, etc.) will be defined as one input unit if he or she serves the particular building in question on a full-time basis. Anything less will be pro- rated accordingly. Instructional aides financed through the general fund budget and who work full—time will receive a maximum weight of .5, and likewise, will be prorated accord— ingly for less time served in the building sampled. In order to recognize the value of the building princi— pal, who... "far more than any other person, shapes and articu- lates the prevailing ambiance and creates a sense of mission,"86 a quantitative measure of 2.0 has been established for a full-time building administrator. Research over the last decade has increasingly pointed to the leadership of the principal as a quality factor in the school building. The most recent statement of this point is found in an article by John GoOdlad in the January, 1979, Phi Delta Kappan. In his article, "Can Our Schools Get Better?," Goodlad notes.87 In recent studies of schools affecting integration with some success, almost invariably the principal was identified as strategic. In the successful 86. John I. Goodlad, "Can Our Schools Get Better?," Phi Delta Kappan, Vol. 60, No. 5, January, 1979, at 3A6. 87. Ibid., at 3A6. 73 black schools referred to earlier, again the signifi— cance of the principal—-his or her values, dedication and strength--eame to the surface. Almost invariably, too, the principal is a person with a strong sense of personal worth and potency, one who takes a position on issues and is not regarded as a pawn of the super- intendent or of strong individuals or groups within the community. In order to recognize pupil-teacher ratios as a quan- titative measure of programmatic services provided, the sum total of personnel input units arrived at for each of the sample schools was divided by the number of students in each representative building to obtain the number of students per personnel input unit. The lower number of students per input unit indicating a high quality program. This measure then considers both the quality and the quantity of building level instructional personnel input as a composite quality or equity measure of educational programs. The cost per input unit for each sample school was com- puted by dividing the total salary of identified personnel (determined by means of direct review of payroll records) by the number of personnel input units. In similar fashion, cost figures can be computed to determine the cost of personnel input units per student. The reliability of the instrument was discussed in terms of content, clarity, conciseness and purpose with select school business officials from two educational groups, the Middle Cities Education Association and the Michigan School Business Officials. Based on their recommendation, minor changes were made in the format and structure of the questionnaire. 7A The reliability of the author's approach is further enhanced in that the concept of adjusting for experience and training is currently receiving considerable discussion as part of a new state aid formula concept sponsored by Michigan State Senator Jack Faxon.88 Contingency tables were used to rank, order and corre- late the data (i.e., students per personnel input units and cost per input unit) by school and school district. If cost differences exist for providing comparable services, the data will indicate a variance in the cost per personnel input unit among the schools sampled. If differences exist rela- tive to the programmatic measures of educational equity used in the study, the data will indicate a variance in students per personnel input units across the sample schools. Isolating Basic Educational Expenditures There are four major sources of revenue for K-12 educa- tion in Michigan. The first is formula state aid which is paid to districts on the basis of property value in the dis- trict and taxing effort. The second major source of funds is local taxes which are based on local wealth and tax rate. These two revenue sources comprise the bulk of general fund resources and are used, at the discretion of the school board, to provide the regular education program of the dis- 88. Jack Faxon, "The Faxon School Aid Plan for 1979—80: A New Concept in Michigan School Finance," mimeo, Lansing, Michigan, 1979. 4,) trict. The two other fund sources are state and federal cate- goricals: funds provided to finance added services for spec- ial students and support specific programs. There are, then, two distinct types of revenue tied to two clearly identifiable expenditure categories. Funds pro- vided by the general state aid formula and the locally levied school taxes are used at the discretion of the local school board to provide general education for all the district's children. Categorical funds support special and sometimes mandated programs for specific individually identified chil- dren. Such funds may not be used to subsidize general pro- grams for all children. For the purpose of this study, "basic education" expenditures are those expenditures for K-12 education that serve all students excluding expendi- tures reimbursed through categorical program revenues. Because of the method of accounting used by Michigan school districts, figures for educational expenditures in— clude categorical expenditures. But students receiving categorical funds are not evenly distributed across all districts. Conclusions drawn therefrom become suspect. In order to avoid this problem, the sample elementary schools' budgets were purged of categorical expenditures and the units of educational input they represent. Thus,personnel inputs and building costs represent only those supported by general state aid and local tax sources, for these are the sources of revenue which provide general education for a district's student population. 76 Modifying State Per Capita Cost Data Michigan Department of Education Bulletin 1012 presents per capita cost data for all school districts. It is from this bulletin that the news media often compare per capita costs within a particular area. These per capita cost com- parisons are usually detrimental to central city districts because their costs are generally the highest or nearly the highest among school districts in their respective counties. There are several reasons for such cost differences which are not explained fully in the Michigan Department of Educa- tion reports or understood by the users of the data. First of all, categorical expenditures are included in computing per capita costs as reported in Bulletin 1012. Therefore, a district with a high percentage of categorical funding will have its per capita costs skewed for all students. In other words then, the categorical funds that serve speci- fic children are added to the total cost, and when those costs are divided by the total Fourth Friday membership, per capita costs for all students are overstated. Secondly, many districts operate areawide programs in special education and vocational education. The total costs of these operations are included when arriving at the dis- trict's per student costs, but no recognition is given for the additional non-resident students associated with these costly services. 77 Thirdly, many districts operate quasi-municipal serv— ices such as libraries or museums. Expenditures for these types of operation are also included in a district's total expenditures and escalate costs. In response to this problem of overstated costs, an adjustment formula has been developed. In the first step of the formula, a basic current operating expenditure is computed by taking per pupil current operating expenditures (shown in the MDE Bulletin 1012) and subtracting from this figure the following categorical funding sources: (1) federal direct funding; (2) redistribution of federal funds; (3) intermediate district funding, and (A) state indirect appro— priations. These first three figures are all found in Bulletin 1012, and the fourth amount is obtained by taking State Direct Appropriations (also contained in the 1012) and subtracting out-formula state aid. Finally, the County Special Education Tax along with other incoming transfers (items 0860 and 0870 of the 1976-77 Form B report) was sub- tracted from current operating expenditures. The basic operating expenditure, then, is simply current operating costs less those categorical expenditures described above. The basic operating expenditure when computed for all dis- tricts provides comparable cost data for the general educa- tion services of each district. The Lansing School District serves as a case in point and should clarify the meaning of the adjustment formula. The Bulletin 1012 lists its 1976-77 current operating expenditures 78 at just over $1,661 per student, second highest among the twelve districts in Ingham County. Using the data from the 1012 and the 1976-77 Form B report, the basic current operat- ing expenditures are computed as shown below: Source Current Operating Expenditures = $1,661.37 1012 Categorical Funding Sources: (a) Federal Direct Funds = 80.0A 1012 (b) State Redistribution of Federal Funds = 72.A8 1012 (c) Intermediate District Funds = 2.81 1012 (d) State Indirect Appropriations = 137.03 1012 (State Aid Formula) (e) Incoming Transfers minus County Special Education Tax 67.86 State Form B, 1976-77 Item 0870 (f) Incoming Transfers minus Others (Voc.Educ.,Adult Educ.,etc.) = .59 Same as above Basic Current Operating Costs: (Current Operating Expenditures) minus (Federal Direct Funds) minus (State Redistribution of Funds) minus (Intermediate District Funds) minus (State Indirect Appropriations) minus (County Special Education Tax) minus (Other Incoming Transfers) = $1,661.37 - (80.0A + 72.u7 + 2.81 + 137.0u +67.86 + .59) Basic Current Operating Costs = $1,300.57 In summary, educational expenditures reported by the state represent the total expenditures of the school district and per pupil expenditures represent total expenditures divided by resident pupils. They include categorically funded programs within the general divisions. But students receiving categorical funds are not evenly distributed across all districts. Districts with large numbers of handicapped 79 and disadvantaged students receive federal and state cate— gorical resources to support special and individual services to such children, and when these expenditures are averaged over all students in the district, comparative data are unreliable. Research efforts in the past have used gross cost data without trying to isolate "basic education" costs. The conclusions drawn using gross cost data, therefore, be- come suspect. The basic assumption is that categorical revenues equal categorical expenditures——that revenues designated for special education are spent totally on special education programs. This seems a safe assumption in light of continual auditing of categorical programs and the requirement of general expendi- ture comparabilities across all students within a district. The stipulated basic operating formula allows the writer to remove categoricals from total operating expenditures in order to have a better common denominator for comparing costs. The basic education costs as computed were then converted to per student costs. The results are basic education costs per student purged of categorical expenditures. This figure along with state published cost statistics per student were analyzed to determine which, if any, of these figures most accurately reflect elementary program expenditures per stu- dent as reported by each district on their annual school district financial report (Form B). Admittedly, the cost figures computed by the state and modified by the aforemen— tioned formula will not reflect elementary school expenditures. 80 But it can indicate whether statistics such as these can be a more realistic indicator of basic program costs. This modified cost statistic was examined in terms of how accurately it reflects the cost statistics computed for the sample schools. Hopefully, this study will highlight the need for a more comprehensive look into the development and application of a detailed program cost analysis model at all levels (elementary, middle school, secondary, special education, adult education, etc.) which could then be examined in relationship to personnel inputs as a comparable measure of programmatic equity. Identification of Need Variables If the research data indicates significant differences in the cost per input unit between schools and districts, the author will then attempt to study the large question of what characteristics high cost districts have in common. The set of need variables to be analyzed are factors over which a dis- trict has little or no control. In an effort to explore and more fully detect a pattern for cost variations, the dependent variables (cost and per input unit) were correlated with the eight identified independ- ent variables (factors suspected of causing the behavior of the dependent variable). The following variables have been identified to allow this writer to examine whether some underlying pattern exists in accounting for observed interrelations in the data. :3 Size of district (as measured by number of pupils). Type of community (central city, suburb, rural). State equalized valuation per student. Average per capita family income. Total K-l2 operational millage rate. District average teacher salary. Percent of students from families receiving AFDC payments. Percent of student attaining less than 75 percent of objectives on the Michigan Educational Assess- ment Program tests. AAA/\AAA CD \lmU'l 1‘:me V VVVVVVV A Collection of Data The data for the study were collected by means of per- sonal interviews with superintendents and directors of per- sonnel from each of the twelve school districts which, col- lectively, comprise the Ingham Intermediate School District. The personal interviews, however, were preceded by a number of preparatory interviews and correspondence. The author, on the basis of a consultation with Dr. C. Robert Muth, committee chairman, arranged a meeting with Dr. Donald Shebuski, Superintendent of the Ingham Intermediate School District to discuss the nature and purpose of the contemplated study in each of the twelve Ingham County districts. To this Dr. Shebuski offered his commitment to cooperate and support the study as much as possible. At this point it was necessary to contact each of the twelve superintendents of the Ingham County school districts. The author composed a letter in which the nature and purpose of the study were explained and in which the superintendents cooperation was asked. A facsimile of the letter sent to Dr. Matthew Prophet is included as Appendix C. 82 Realizing that correspondence by mail was the least effective means of garnering the superintendents' support, the author arranged to appear before the Ingham County Super— intendents at their monthly meeting in September. The author presented his study to the superintendents, answered any ques- tions they had, and asked them to decide whether they would allow the author to conduct his study in their schools. The superintendents voted unanimously to permit the author to visit their school district offices and collect the necessary data. From the time of the superintendents' meeting to the middle of June, the author arranged with each superintendent a personal interview session at which the data could be col- 1ected and analyzed. Prior to each interview the question— naire instrument and a document further explaining the study were sent to the superintendents. In this manner, each super- intendent knew in advance precisely what data, from what source, was being requested of him. In one case this prior correspondence, coupled with all that had preceded it, served to eliminate the need for an interview entirely; the super- intendent in that case knowing exactly what to provide the author. In all other cases, however, an interview was necessary to go over with the superintendent the question- naire instrument and the data. Either the superintendent provided the data at the interview or completed the ques- tionnaire at his leisure and forwarded it to the author. In the case of the Lansing School District, Dr. Gary Wegenke 83 provided the author with computer printouts of the pertinent data, explained the Lansing coding system, and left it to the author to pull out the necessary data. A few problems were encountered by the author during the interview sessions and subsequent reviews of the data provided thereby. The most frequent problem was the inclu- sion of special education personnel and salaries,therefore, with the data. To ensure that such data were not included, it proved necessary to review directly the personnel sheets of the schools, or to consult with the superintendent directly, and eliminate those for whom funding was provided through special education or some other categorical program like Title I of ESEA. Another problem arose due to the forms in which the districts maintain salary, experience, and education infor- mation on their teachers. It was found that some schools, most notably the Lansing School District, do not include in their records the number of years worked by teachers outside of the school district. Thus no measure was possible for overall work experience for those schools. In the case of the Lansing School District, the computer printout listed the number of years worked within the Lansing School District, but to find the number of years worked outside the district (if any) one would have to consult the individual teacher's personnel fine, something which the author did not have the legal right to view and the Lansing personnel director did not care to do. HA Once the data were collected, either by means of the interview or by personal analysis of the data (as in the case of the Lansing and Williamston school districts), the author again analyzed the data to ensure that all special education personnel had been excluded. In many cases, this necessitated a follow-up phone call to the superintendent so as to double check the results. Another area in which problems were encountered was the category of librarian. In some instances the school did not have a certified librarian but still assigned some- one full- or part-time to the library. It had to be ascer- tained whether the librarian was certified or non-certified. In those cases in which the librarian was non-certified, that person was then classified as an aide and was awarded a weighting of 0.5 rather than 1.0. The determination of fringe benefits for the average elementary teacher also proved to be next to impossible to accomplish. In all instances the fringe benefit package was not available on an individual basis. On a district- wide basis, the average cost per teacher across the twelve districts was not significantly different and was, there- fore, not computed into input unit costs for the sample buildings. CHAPTER IV PRESENTATION AND ANALYSIS OF DATA Introduction The analyses of data relevant to this study are presented in this chapter under the following headings: Relationship Between Programmatic and Cost Inputs, Cost Per Input Unit and Program Equity, Cost and a District's Wealth, District's Abil- ity to Pay, District's Need Variables, Comparison of Per Capi— ta Cost Data and Summary of Findings. Relationship Between Programmatic and Cost Inputs Chapter III described the procedural steps and methods utilized in obtaining the input variables (personnel and cost inputs) for each sample school by school district. As was indicated in Chapter I, much of the school finance research does not take into consideration the district-by-district variation in the input costs of education. Up to now, per student cost input differences have been the basis for inter- preting program or education quality disparities; even though average per student costs include special categorical and community program costs that are not uniform across districts, and even though it is generally knowledged that input costs, primarily salaries, vary geographically and by district size. The driving mechanisms for educational cost and local tax effort are personnel input costs and the teacher/pupil ratios. All other cost variables are relatively insignificant com- 85 86 pared to the salary costs of education and the student units over which they are spread. The following analysis focuses upon the question of whether equal dollars can provide equal program diversity and opportunity when the same services have different costs across school districts. In order to isolate costs related to direct programma- tic benefits to students, the author constructed a uniform method of weighting and pricing input units (personnel factors) at the building level. Categorical expenditures and those services acquired through categorical funds are eliminated from the computation to insure comparability among sample districts. Having programmatic inputs and cost per student as the indicators of educational programmatic equity, the author surveyed (via on-site visitations) a representative sample of 2A elementary schools from the twelve Ingham County school districts. The sample of elementary school buildings has been selected to provide for full representation and optimum comparability across schools and school districts. As can be seen from Tables 1 and 2 following, the Ingham County school districts provide a meaningful grouping in terms of size, geographic structure, community and expenditure levels. mm._m ep.mm eo.mm oo.Pm cm.—m mm.mm mm.mm om.mv om.Fm mo.om om.~m om.me mo.om «P.0e mp.me oms_aaz Passe oe.~ w~.~ mw.m 00.x om.¢ mo.m cm.“ on.“ mo.~ om.m mm.~ .erasme ammo -.wm ~n.mm mm.~m oo.pm om.m~ ow.om oo.wm o~.mm oo.n~ oo.- ov.om ca.em oo.w~ m~.mm oo.mm Nc.mp mm.m~ mn.¢~ oo.~m cm.o~ om.- oo.mp o~.- oo.m— oo.wp o¢.~N om.m~ oo.mp m~.e~ om.om oo.m oo.m oo.m zo~p.xm .oeppa ------ezsemm omImNmp www.— mm~.m om~.m~ >m >mHznou z.m.m pagan Fauna coacaasm posse concanam ~ncsm —uc:m censsaam censsazm Forum coat: ceassesm :ovueo uvwwmmmpu Parse censsnsm seaweed . cent: ecumEMVvas oppr>soanoe xpco>es amassexooum mesoxn some; m_PmoL ape? bumpmm+ op—_>mcefl mcmmcm- mcwmceb “we“ mEez uuprumwfi mms.ma so.oas.a ems.mso.m mmm.em 0mm ..oeeom hoe Heeoe ssm.aa mo.emm.a msm.HH mam.os mm com soaom ssm.mfi so.mmm.a esmwfle smm.am om mam op com oHHassooooz Hme.ma Hm.mma.a mom.mm mmm.sm we mas.a o» ooo.a oaaasmeeo Has.ma Hm.mHN.H mam.mma ame.mm ms mmm.H4 op oom.H coonsmaaaaz .oaamoq sas.ma ao.amm.a moo.0sa Hme.sm me ems.m op ooo.m omeasoxooem .eooanem ome.ma ma.mam.a mso.mm mms.mm mm mam.m o» oem.m mmm.sa sm.mmm.a osm.om mmm.mm mm mas.m oo ooo.m nosoxo mmm.sa os.am~.fi ess.mm Hmm.am em mmm.m oe oom.m some: was.ma mm.mmm.a mem.sHH see.sm em mmm.s o» ooo.= Assess: ess.ma ss.mss.a =ms.ms mes.sm m mmm.= o» oom.s use: .w2amcwq pmmm ssm.mfi se.oas.a smm.mas ess.em He mmm.m op ooo.m was.ea Hm.mmm.a Hme.som mae.sm mm amm.mH os ooo.oa cam.ea sm.mmm.a mmm.mem was.sm a www.ms oo ooo.em meaneeq emm.aa mo.mom.a mam.asm msm.om H peso a ooo.om awrnwfiwm mwhzufincmmxm QHEmLmDEmZ LODEOE UH< mpwpm wuoahuwwa QSOLO mUOHLumfiQ Lmrcmee wcfipmnedo e>fipe< Lem cofipmsam> no dfinmpebsmz eHdEmm mmwpw>< Hausa pmm UmNHHNSUm mumum LwDEZZ m>Huo< omlmsmfi memmmmEmE wm ammaomo mBonBmHQ goozom mfilx 0mm m.z-»5 indicates relatively low correlation. Computations for the variables indicated 98 were computed and placed in the matrix table which follows. Each combination of variables will be discussed separately. so» oesou—Lu Lame.. .Lee~.. Less. . 93. 2mm. 33. p.25 3ch sum umoo .. .mem.. A~nm. . 23. o.— mom—i noon. 3:: 3ch pom nucecsum .. Ape... ..aea.. . 22. e; 3.2: H 339 a .. Lamp. . .Lmsn.. seep. mike. e.. . eea~:. sateen dam: Lena v Lee... “Nee.. L.me.. n.~.. . ass..- sees}. seam. seen. new secessm Lame . Aaea.. Leas. v Lame. . Lace.. 3:. $2.- LEN. :8. 8.3... 2505 338 tom .92 Lesa c Laee.. ._ee. v Lace.. Laap. . A_ae. . BS. 2: - e._ 38. was: 38. .22. as .98 .mots sons Aaee.c .ae. . Amee.v Amie. . Lace.. Laos. . L_ee.. was. as; - 8E. as: $8. 88. as. as .ea :88 Lama . L.e_ . L_ee.v Amse.v Lane. a Lepa.v seem. epme.. same. e.. Peon. case. mane. as .axm .soao cause a. Lame . Am_m a LeeN. . 9. Fees. amo~.. o._ ~mmm. sunfish success smasose Laee.c Lees S Apes.. Lama. . slam.- same. meee:. e._ Lesa. - suspends no case Leas a Lens . Amee.. Laee.. Leae.. Leee.v Lamp. . Ame..v seem. mmFe_. seas. seam. saws. ease:. sees. ene~. omeaaaz measetoao .. .mmm.. Lame. v .een.. Lama.. 28. 82.- 28. SS. e.. 38. as in ..Ne.. Lama.. Lsae.v Leee.. Leee.v Lama.c See. 88. :8. $3. $3. £8. c; 32835: accuser. O fldS I: S OV 3&3 308 SLV Cl S S 7A a mam i av as use as“ was an n mm an an? a p T? ..u d o o o B m o J TD 8 O a n4 a u. p a T7 I a q a I d a o o a o 0 Tu q o d d 8 TL 8 d U u T. 8 u a d d .A O 0 .d . U 8 G. 8 d1 3 OJ J J 3 I J W 8 . a meBS‘ onefimmmoo ooeso.c_easm e use.mm Hesse emefisnxeos me.efi NH.N= NH.OH oo.>m www.me emm.am Hesse efiame om.efi mm.~s ma.oH oo.mm www.ma sme.m~ Hmcsm eHHH>hce efi.mfi mm.mm mm.oH om.mm sem.me asa.mm Hesse eHHH>seene Hm.mm *em.mm mo.HH wo.~m mmm.m» Nae.mm muse meshes H:.om os.om mo.m oo.~m mem.mh mm~.~m Hmcsm comm em.mm mm.mm om.oa m~.om mmm.me HH~.Hm Hesse cesmEmHHHH mw.om mm.:o mm.oa so.mm =ma.me ~m=.mm cencsnsm Became em.eH mm.wm Hm.w Hm.om mmm.m» mmm.om emhcspzm ado ms.sm om.:e HH.HH om.mm Hmz.me on.ss ceeasnsm meshes; chm mm.mH mm.oo mo.m om.om sow.m« Hem.oo censspsm sace>m mm.mH om.mm H:.m so.om =mm.:w mmh.wm :mopsnsm mome whoocH comm mEooCH no mpmm mEoocH wuHQmo MMMMMW mmmmmmmm mmwmmmw he ooo.as ewmflfluz ooo.~a sea emeHHz sea emmse>< do ease ate them .EEoo.soe seem .Hfisz Hoocem Hmsoe .MHE.E00 Hoocom BUHmBmHQ wm MZOOZH mo ooo.Hfi mmm m9< .mBmm OH mqmm mumszozwaxm uz~pmcmo eme_eaxeoem ape: mppp>ceanez comemvp—m: oeepmex moemxo apcm>m3 newmcma mcmmcmd ummu seaasmaa 12? The basic operating costs computed for Lansing ($1305.37) represent a decrease of $533.40 from its reported operating expenditures (per MDE Bulletin 1012). When this comparison ' is made to all districts in Ingham County, the magnitude of the difference between state published data is not nearly as great-~the average for all districts being $245.24-- Williamston shows the smallest difference of $59.62. While the Lansing School District ranks second in the county in terms of the Department of Education's presentation of per capita costs, its rank drops to seventh when costs are ad— Justed to more accurately represent basic education pro— gram expenditures. J Average salaries, by observation, correlate with per capita cost of education and are the significant factor in determining per capita cost. Although this is well known and an obvious fact, school statistical reports do not index per capita cost with average salary data. It should be emphasized that even the basic operating costs do not give a true comparison of programmatic equity in terms of dollars. In Ingham County, for example, there is a county vocational education millage for all constituent districts with the exception of the Lansing Public Schools. The Ingham County Intermediate District operates a skill center for all students outside of Lansing. The cost of vocational education for those students attending the skill center never appears as an expenditure in their respective school districts even though the full memberships of such students remain with the constituent districts. Lansing operates its own vocational education program with its own 123 resources, and the full cost appears in their cost data. The net effect is a lower average cost per student among the Ingham County school districts using the skill center as com- pared to Lansing. Also, it should be noted that such complexities are not dissimilar to problems encountered by urban cities in other counties in computing and comparing operating costs for dis— tricts. In order to ascertain which of the two cost statistics is a better predictor of elementary program expenditures, the figures were correlated with the district's elementary program expenditure per student as indicated on the school's annual financial report (Form B). The classification "ele- mentary program expenditure" is the cleanest statistic of instructional educational cost currently available as it is devoid of categorical expenditures and district overhead. Correlation coefficients were computed and are presented below in Table 15. TABLE 15 CORRELATION COEFFICIENTS FOR SELECT PER STUDENT COST DATA Elementary Program Ex- penditure Per Student Operating Expenditure Per Student .7509,¢k = .005 (MDE Bulletin 1012) Basic Expenditure Per Student (As .8354,0A = .001 Modified) As one would expect, there exists a strong and signifi— 124 cant positive relationship between Elementary Basic Program Expenditures and both alternative per pupil expenditure meas- ures. That is, both the Department of Education Bulletin 1012 computation of operating expenditures and the author's modi— fied figure are accurate indicators of a district's instruc- tional costs. The author's formula, however, is more strongly related to instructional costs. A comparison of the relative strengths of those rela- tionships can be made with the use of the respective r2 values. The r2 statistic is a measure of the proportion of variance in one variable "explained" by the other. Thus, the author's measure explains about 69 percent (r = .8354, r2 = .6895) of the variation in per student program expenditures while the MDE measure explains about 56 percent (r = .75, r2 = .5625) of this variation. In summary, educational expenditures reported by the state represent the total expenditures of the school dis- trict and per pupil expenditure represent total expenditures divided by resident pupils. They include categorically funded programs within the general divisions. But students receiv- ing categorical funds are not evenly distributed across all districts. Districts with concentrations of handicapped and disadvantaged students receive federal and state categorical resources to support special and individual services to such children, and when these expenditures are averaged over all students in the district, comparative data are unreliable. 125 Research efforts in the past have used gross cost data with- out trying to isolate "basic education" costs. The conclu- sions drawn using gross cost data, therefore, become suspect. The author's adjusted formula tends to more accurately repre- sent basic programmatic costs as identified by every school district on their annual financial report and is not diffi— cult to compute. Summary of Findings In this study, the author sets out to analyze programma- tic equity in terms of instructional support personnel among the select buildings and districts of Ingham County and the difference among school districts in the prices of educational services (isolating on instructional staff) among these same school districts. Equity was measured in terms of inputs--providing per— sonnel and cost inputs--to determine whether services or resources are distributed "equitably." In summary, equal dollars cannot buy equal services when the same services have different costs across school districts. The data reveal, for instance, that while the Lansing School District has the highest input cost per pupil, it also has one of the least favorable programmatic input measures per student (e.g. staff programs) of the 12 districts surveyed. The most obvious explanation for the relationship between dollars and programmatic inputs is staff compensation. Lansing ranks the highest in average teacher salaries among the 12 sample districts. The author also concludes from further analysis of the data that disparities exist in the cost of education and the resultant tax effort. The high cost districts also levy the highest millage rates. These districts with lower staff costs are able to provide comparable staffing ratios while levying less millage. Virtually all studies which seek to answer whether a system of school finance is an equitable one select local district wealth, and generally property wealth, as the dis- criminatory criterion of interest. That is, the studies ask "are educational resources (e.g. revenues, staff, pro- grams) distributed among the local districts according to local property wealth?" The equity principle in such studies is one of "equal opportunity" equity which is perhaps best stated in the following negative form: differences in the treatment of pupils should not stem from differences among pupils which are discriminatory such as property wealth.. This principle provides the basis upon which the state's general membership aid formula is constructed. The formula uses a district's SEV per student to determine state aid reimbursement. The "equal treatment of equals," that is, equal dollars for equal efforts, fails, however, to acknow- ledge equity in terms of ability to pay, cost differentials and educational needs. With regard to ability to pay, this study reveals that equity criterion is not met. While per capita income figures 127 (as a measure of ability to pay) for the rural districts is slightly lower than Lansing. The rural districts also levy significantly lower millages, indicating that the tax- payers of Lansing are making a relatively greater tax effort in relationship to their income. The data also reveal that the community variables charac— terizing the level of education need were highest in Lansing. Higher basic educational needs also suggest higher costs. Cost is the key determinant relative to how much a dis- trict must pay in order to provide comparable programs, staff and services. If equity is to be achieved, districts facing excessive costs over which they have little or no control must receive help from the state in meeting those costs so that an equitable proportion of funds will be available in each district to meet the proper needs of its students. The recommendations of this study which follow recog- nize and address that need. The author will set forth in the final chapter recommendations as to how a state system of financing can be implemented that achieves both "equal opportunity" equity and next the equity principle addressed to in this study (i.e. the appropriate unequal treatment of unequals). CHAPTER V SUMMARY OBSERVATIONS AND RECOMMENDATIONS Summary Observations Defining and measuring equity in state school finance structures is long overdue. Local, state and federal offi- cials are searching for clear definitions of the many dif— fercnt equity goals in school finance. Equity in school finance, particularly the term "equali- zation," is used loosely in policy discussions. "Equity" and "equalization," however, cover many diverse, and at times conflicting, goals of a school finance policy. There is need for a coherent explanation of what is implied by equity in school finance in order that the discussion of various alternative policies can be properly analyzed. Such a framework should be organized around possible answers to three questions that constitute the framework for this dissertation. The first answers the question: Equity for whom? The two groups of primary interest are children who attend the schools and taxpayers who pay for the schools. The second question is equity of what? For children that could mean equity in terms of cost differentials or equity in terms of programmatic opportunities whereby differences in needs among children are recognized. For taxpayers, it could mean fair property taxes according to ability to pay. For either taxpayers or children, the answer to the third question "what is to be fair?" is pivotal to conclusions that can be drawn about the equity of the system. 128 129 The problem addressed in this research was to demonstrate the relationship of cost input differences with programmatic differences and correlate these differences with other recog- nized variables of educational equity. The author developed and applied a formula for measuring programmatic equity and correlated these measures with educational cost data and other measures indicative of program need and other variables generally applied in school finance. The first stated purpose of this study was to answer the question of whether dollar expenditure equity equates to pro- grammatic educational opportunity for children. The analysis of the data reveals cost differences do not equate to student program differences. As it turns out, city and suburban dis— tricts tend to have the highest millage rates, the highest cost per input unit, but do not necessarily have the most favorable staffing ratios. Rather, as in the case of the Lansing School District, their staffing ratios are less favorable than those of the smaller districts. This is true despite the fact that small districts spend considera- bly less per pupil. But what is the reason for such an apparent paradox? This study has shown that the major variable is salaries and that staff compensation to a great extent is outside the con- trol of local school boards. While Michigan's state aid membership formula is clear- 1y "wealth-neutral" in terms of equal yield for equal effort, equity has not been achieved with regard to the substantial 130 variation which exists across school districts relative to distribution of general revenues, basic instructional ex- penditures and regularly funded instructional staff. Failure to make adjustments in the general membership aid formula will force high—cost, high-levy districts to reduce programs to children as needed high taxes continue to be rejected by taxpayers. Today this situation is pri- marily found in urban and metropolitan suburban districts where the difficulty of meeting higher costs with higher taxes is deepening and spreading. Salary costs will be determined by the regional market place or district size while per student costs must be controlled by the local board, even at the expense of educational equity. The second purpose of this research was to determine how current state cost data reflect differences in the cost of personnel input units across school districts. Educa~ tional expenditures reported by the state represent the total expenditures of the school district,and per pupil expenditures represent total expenditures divided by resi- dent pupils. They include categorically funded programs within the general divisions. But students receiving cate— gorical funds are not evenly distributed across all dis- tricts. Districts with concentrations of handicapped and disadvantaged students receive federal and state categorical resources to support special and individual services to such children, and when these expenditures are averaged over all students in the district, comparative data are unreliable. 131 The author's adjusted formula tends to more accurately repre- sent basic programmatic costs as identified by every school district on their annual financial report, and further, is not difficult to compute. The third purpose of this study was to determine the correlation of recognized school district and community variables with the computed cost of programmatic variables for the sample districts. The data presented in Chapter IV indicates that poverty indicators such as a district's socioeconomic status, aver— age per capita income and the percentage of Title I stu- dents are much higher in an urban city such as Lansing than in the suburban areas of East Lansing and Okemos. As a result of the almost exclusive reliance of schools on the property tax for local revenue, property value is an important indicator of fiscal capacity. However, the cen- tral cities, such as Lansing, are now experiencing trends that are depriving them of that base. These trends are the movement of industry and commerce to the suburbs and the decline of property value in cities while other areas property values are increasing at the rate of or in excess of the rate of inflation. This is supported by the effect the Headlee Tax Limita— tion Amendment has played on school districts during the 1979-80 school year. In these districts where equalized property values grew faster than the rate of inflation, mil— lages had to be rolled back to bring revenues into compliance 132 with the Headlee Amendment. In 1979-80, about two-thirds of the state's 530 districts faced millage rollbacks. For the most post, however, these millage rollbacks did not occur in urban areas of the state where tax bases are already Shrinking because of declining property value and the lack of new construction. While the causes of the movement of industry and com- merce to the suburbs are complex, it is sufficient to note their effect on per capita wealth. Of special significance is the situation where the income base has fallen faster than property valuations. What this means is that the apparent wealth of cities as reflected in property valua- tion and state aid criteria is relative greater than the ability to pay from current income. Consequently, SEV is not an accurate indicator of the ability to levy or pay high millage rates necessitated by high costs. The state aid formula needs to be designed to take tax burden and ability to pay into account. While current income is not the sole determinant of ability to pay, it needs to be heavily considered along with valuation. RECOMMENDATIONS Clearly high tax effort and high cost per input units do not necessarily insure favorable staffing ratios in Michi— gan school districts. On the contrary, high-tax and high-cost 133 districts may find themselves with below-average programma- tic opportunities. The question then is what would be an appropriate policy approach? The problem as this author sees it is not one of unequal school district expenditures per se as one of unequal cost of educational services. The existing disparities in the cost of school staff across school districts Show that the current policy of equalizing dollars by tax effort may not be advisable in and of itself. Such a policy, without equalizing staff compensation as well, only exacerbates existing inequities. What should be considered is a policy which would recognize cost differentials. Based on the data for the twelve districts in Ingham County, this author would recommend that an educational cost index be built into the state aid formula which recog- nizes salary differences. Cost differences have been con- sidered for many years in state reimbursement of transporta- tion and for the last four years in state added cost funding of special education. But there has never been in any state school aid provision, recognition and compensation for the fact that some school districts must pay considerably more than others to provide a basic education program. It is often contended that local instructional cost differences start from the "local control" tradition and are, therefore, not the state's responsibility. After all, salaries are determined by local collective bargaining and local districts make their own determination of the experience 134 and training qualifications they want when vacancies are filled. However, to an ever-increasing extent, these conten- tions are no longer valid. Declining enrollment experienced by the vast majority of Michigan schools has given many dis- tricts no choice but to retain a teaching staff composed al- most totally of persons at the top of the salary schedule and the result of the collective bargaining process practi- cally forces conformity to regional levels of expectations. The problems of how to finance education and of how to reward for cost differences without increasing cost dispari- ties must be addressed. The author has presented analyses of cost differences and has shown that cost differences do not bear a direct relationship to student program differences. It has been shown that the major cost variables are salaries and that those are to a great extent outside of the control of local boards. The present state aid power equalizing formula is pre- dicted on equal resources for equal effort. Yet from its inception it has rewarded minimal effort and lack of effort. Adjustments must be made and the author offers the following recommendations. RECOMMENDATION 1 Consideration should be given for equalizing all the mills over 30. As the data presented in Chapter V reveal, high expendi- 135 ture districts are not necessarily these districts with the more favorable staffing ratios. Rather, it turns out that, in general, large districts have relatively high millage rates and per pupil expenditures but relatively unfavorable staffing ratios. Smaller districts, on the other hand, en— joy relatively favorable staffing ratios while levying rela- tively low millage and spending considerably fewer dollars per pupil than their larger counterparts. The principal reason was found to be differences among the districts in staff compensation. In response to this paradox, two modifications in state school aid are recommended by the author. The first is the equalization of all mills above 30 levied by a school dis- trict for school operations, as opposed to the equalization of only one-half the mills over 30 as currently provided for by the general membership formula. By equalizing only one-half of district's operating mil- lage in excess of 30, for those 88 districts in Michigan which were in-formula and levied over 30 mills in 1978-79, their school revenues varied according to property wealth. This aspect of the aid formula is not compatable with the equal-yield concept that equal district millage rates should yield equal dollars for school operations. An example of this inequity is provided in the follow- ing table, which applies the current general membership for- mula to 1979-80 financial data for two Michigan school dis— tricts: 136 TABLE 16 CURRENT SCHOOL AID INEQUITY General Total Local State General Operating Revenue Aid Revenue District SEV PP Millage PP PP PP Flint $32,948 39.82 $1,312 $676 $1,988 East Lansing $50,336 39.60 $1.993 $ 70 $2,063 Thus, while Flint levied a slightly higher millage than East Lansing for school operations in 1979-80, they generated about $75 less per pupil than their wealthier counterpart under the present formula. ' The proposed change would break the relationship between district SEV and district revenue by assuring equal per pupil revenues for equal tax effort. In so doing, the proposal would provide increased state aid to virtually all district levying over 30 mills for school operations. These are generally ur- ban school districts, which typically need to offer higher salaries in order to attract staff. (Salaries in these dis- tricts are, in fact, about $3,500 higher, on the average, than salaries in districts levying 30 mills or less. (See Appendix G). In summary, these high-millage districts must pay rela— tively high salaries in order to compete for staff in urban markets, while their millage rates are about 25 percent higher than the statewide average. The equalization of all mills over 30 could assist these districts in limiting future millage increases and improving their staffing 137 so as to be more in line with their substantial tax effort. Past reluctance to equalize all mills over 30 was due pri— marily to a belief that this policy would encourage dis- tricts to raise millages substantially. Research by the Rand Corporation, however, has revealed little response by Michigan school districts to this incentive. Thus, the proposal to equalize all operating mills would: 1. increase the wealth neutrality of revenue distri— bution among school districts in Michigan. 2. provide additional state aid to districts which face staff compensation costs and levy high mil- lages for school operations. RECOMMENDATION 2 It is recommended that an educational cost index based on present teacher salary differences be incorporated into the formula. As noted earlier, the cost of inputs to education are not substantially controllable at the local 1evel--shortage of resources means less programs rather than lower input unit costs of education. Education is a labor (professional) intensive industry in which 80 to 85 percent of costs are salaries. When resources are scarce, average per input unit costs of education actually go up as lower-salaried staff are released first. Output unit costs go down by 138 spreading the input cost (teachers) over a greater number of output units (children) or by reducing the quantity of output units (reduced school day). In either case, it means fewer services for children. In spite of an accelerated equalization of state fund- ing since 1965, relative salary differences between markets have remained the same. The result has been that high—cost districts in high-cost markets have remained high cost. Thus, high-cost, high-effort districts have had to steadily increase their effort or reduce their programs while low-cost, low-effort districts have been able to maintain programs at relatively low effort. Input salary costs are not fully controllable because of a state mandatory negotiations act (Public Employee Rela- tion Act--PERA) and because of a fully institutionalized teacher salary structure that rewards training level and experience. Further, negotiating teams establish markets to support their positions and arbitrators establish markets to justify their recommendations; these markets reflect both ggeographic and size relationships. The present research also showed that teacher experience :and level of training is a cause of the variance in average ‘teacher salaries. Lansing, for example, ranked highest in Inigham County in staff experience and education and also had tale highest average teachers salary of the sample schools. It is recommended that salary differences be based on tfile average market of local district salary schedules at the 139 BA and MA levels and the actual distribution of teachers within the schedule. Market determinants include either geographical location or district size. The one serious and inescapable problem with a weighted index approach is that it is extremely expensive. For the past three years, Senator Jack Faxon of Detroit has spon- sored a proposal similar in concept to the one described above. Senator Faxon's approach would consider experience and training qualifications of teachers as a factor in deter— mining the state—local per pupil guarantee of "gross allow— ance" of the basic membership formula. A teacher with less than a BA degree and less than one year of experience would be assigned a factor of .60, a teacher with a PhD and 11 or more years of experience would be assigned a factor of 1.95, and all others would have factors somewhere in between. The average for each school district would be determined and would then be multiplied by the district's gross allowance for purposes of computing state formula aid. Under the Faxon proposal, the average district would have a "grid factor" of 1.12 which means that added funding would go to nearly all school districts. Had it been imple— mented for 1978-79, this proposal would have cost approxi— mately $700 million. Speaker of the House Bobby D. Crim in his 1980-81 State-of—the-State address recommended a somewhat similar but considerably less costly proposal. Speaker Crim recom— 140 mended a "cost-of—doing-business" categorical be considered based upon the average teachers salary and that only dis- tricts with above—average costs be funded. The Speaker's proposal as described by Gene Caesar of the House Democratic Education Office89 would appropriate $30 million to be allocated to school districts where in— structional costs exceed the state average. The computation is based upon basic program costs only, using prior year data, and cannot exceed 25 percent of the amount that a district would receive under the basic formula, if its gross allowance was increased by the same percentage that its instructional costs exceed the statewide average. If in 1978-79, all districts where average teachers salaries exceeded the statewide average, had been given gross allowance increases in direct proportion to their percentage of excess, the total cost would have been approxi— mately $111 million. Thus a $30 million cap would provide about a 25 percent proration. Accordingly, if a factor derived from a high average teacher salary in a district is applied to the entire gross allowance, no more than 25 per— cent of the resulting entitlement would be funded. A combination of the four recommendations cited above would achieve increases in state aid for all districts fac- ing above-average staff compensation costs and would increase wealth neutrality in the distribution of revenue among dis- tricts. The four proposals, therefore, are complementary, 89. Gene Caesar,"A Proposal for a Cost-of—Doing-Business State Aid Categorical," mimeo, Lansing, Michigan, 1979 at 1. 141 rather than mutually exclusive. RECOMMENDATION 3 It is recommended that an effort ability factor be added to the state school aid formula. A measure of ability built into the distribution for- mula would generate greater equity. It should relate the per capita tax levy to per capita income in dollars. The problem inherent in ability-to-pay measures is the unavaila— bility of current data on income by school district. This problem is not insurmountable if real desire to use such a measure exists because income by school district can be ob— tained from the Michigan individual state income tax returns. Adjustment factors in developing such a formula could include the following: 1. State Equalized Valuation. 2. District and state total property tax rates. 3. Local income tax rates adjusted to property tax rates. 4. Average family income by school district. 5. State municipal revenue sharing adjustment. An effort-ability formula could be devised and used as a possible substitute for the present municipal overburden section of the state aid act. (The municipal overburden section of the state aid act provides for an increase in the formula gross allowance for school districts which have 142 unusually high non-school property taxes.) Such a formula should consider all the aforementioned factors and provisions which offset tax equity in Michigan. As essentially all edu- cation taxes would be equalized under present law and the four preceding recommendations, the effort-ability formula could apply only to non-educational local variables. Since all school taxes eventually are paid from income, the local contribution could be determined on the basis of average income in addition to property value. The millage required for the local share of the formula guarantee could be weighted on the basis of the per capita income in the community. A composite index of income and property valua- tion could permit high-cost, high-need, low-income districts to improve their programs, thus narrowing present variations in education opportunity. Implications for Further Research 1. Equalization and equity in school finance have different meanings for different people. These two objectives are quite different, and it is not clear how policies and research implemented to produce our objective effect the other. 2. There is a need for a framework which can provide order to and understanding of the various policy objectives and equity goals in school finance. Who are the differ- ent groups for which school finance systems should be equitable? What should be distributed fairly for the 143 different groups? And, what are different principles for assessing whether the distribution is equitable? Finally, there needs to be better understanding and consensus on what statistical tools can be used in ob— taining an acceptable standard of programmatic equity. Further research such as undertaken in this study rela- tive to how input variables affect equity and quality of education needs to be considered. This author has sought to analyze the effect of personnel and cost inputs or programmatic opportunity to children. Up to now, per student cost output differences have been the basis for interpreting program or education equity disparities. Greater research in terms of input variations in cost and educational opportunity among districts needs to be undertaken. A fact that is often lost sight of in stu- dent cost output research is personnel input costs as the standard for the instructional program in terms of pupil/teacher ratios. All other variables are relatively insignificant comparied to the salary costs of education and the student units over which they are spread. It it recommended that a major statewide study be conducted based upon the procedures used in this study. BIBLIOGRAPHY 144 BIBLIOGRAPHY Anderson, N.R. "School Finance In Washington". Washington Law Review, Vol. 50, 1974. Berke, Joel S. and Callahan, "Serrano v. Priest: Milestone or Millstone." Journal of Public Law, Vol. 23, 1972. Berne, Robert., "Alternative Equity and Equality Measures: Does the Measure.” Selected Papers In School Finance Washington, D.C.: U.S. Department of Health, Education and Welfare, 1978. Bee, Jason. "State Legislators and Quality Education." Compact for Education, Vol. XIII, Summer 1979. Bothwell, Robert O., and other. "State Funding of Urban Educa- tion Under the Modern School Finance." Washington, D.C. U.S. Department of Health, Education and Welfare, January, 1976. Bothwell, Robert O. "The Inequities of the Texas School Finance System.” Paper presented in Houston, Texas, September, 1973. Brazer, Harvey E. and Anderson, Ann P. "A Cost Adjustment Index for Michigan School Districts." Ann Arbor, Mich: University of Michigan, 1975. Brazer, Harvey E. "A Study of Expenditure Levels and Student Output Measures Among School Districts." Selected Papers in School Finance: 1974, Washington, D.C: U.S. Office of Education, 1974. Brown, Byron and Saks, Daniel, "Production Technologies and Resource Allocations Within Classrooms and Schools: Theory and Measurement.” Paper develOped in the National Institute of Education; East Lansing, Michigan, Michigan State University, 1978. Caesar, Gene. "A Preposal for a Cost-of-Doing-Business Cate- gorical." Lansing, Michigan: House Democratic Education Office, 1979. Carroll, S.J. and Park, R.E., The Search For Equity in School Finance: Michigan School District Response to a Guaran- teed Tax Base. Santa Monica, California: The Rand Corporation, 1979. 145 Coons, John E., "Fiscal Neutrality After Rodriquez." Law and Contemporary Problems, Vol. 38, 1974. Coons, John E., and Sugarman, Stephen D., Private Wealth and Public Education. Cambridge, Mass: Harvard University Press, 1970. Coons, John E., "Recent Trends in Science Fiction: Serrano Among the People of Number." Journal of Law and Educa- tion, Vol. 6, 1977. Faxon, Jack, "The Faxon School Aid Plan for 1979—80: A New Concept in Michigan School Finance." Lansing, Michigan: State Senate, 1979. Goodlad, John I., "Can Our Schools Get Better?" Phi Delta Kappan, Vol. 60, No. 5, January, 1979. Grubb, Morton W., "Constructing Teacher Indices: Methodologi- cal Explorations With California Unified School Districts." Los Angeles, California: Education Finance Reform Proj- ect, 1975. James, Thomas H., "Diagnosis and Treatment in Ailing School Finance." Louisville, Kentucky: Council of Chief State School Officers, November, 1971. Jordan, Forbis K., "Cost Differentials in State Aid Programs in Selected States." Washington, D.C.: American Educa- tional Research Association, 1975. Kenny, Lawrence W., and others, "Measuring Differences Among Florida School Districts in the Cost of Education: An Alternative Approach." Gainesville, Florida: University of Florida, 1975. Levin, Betsy, "Current Trends in School Finance Reform Liti- gation: A Commentary." Duke Law Journal, 1977, Vol. 1977, Number 6. Levin, Betsy, Public School Finance: Present Disparities and Fiscal Alternatives, Washington, D.C.: The Urban Insti- tute, 1973. Levin, Betsy, and others. The High Cost of Education in the Cities. Washington, D.C.: The Urban Institute, 1973. Levin, H. M., "Equal Educational Opportunity and the Distri- bution of Educational Expenditures." Education and Urban Society, Vol. 5, 1973. IHC Muth, Robert C. and Stuhrman, Paul R. Financing Education in Michigan: A Report. East Lansing, MI: Ford Founda- tion Study, 1977. Phelps, James L. and others. "State Aid for Michigan Public Schools: An Analysis of Two PrOposals." Lansing, MI: Michigan Department of Education, 1979. Raymond, Joan M. "Financing Equity Among Schools in Large Cities - Chicago Public Schools: A Case Study." Nash- ville, TN: American Educational Finance Conference, March, 1979. Rumbaugh, Stanley A. "Cost of Education Differentials: The Search for the Missing Link." Lansing, MI: Michigan Department of Education, 1977. Sacks, Seymour. "Central City and Suburban Public Education: Fiscal Resources and Fiscal Realities." MetrOpolitanism: Its Challenges - Sixty Seventh Yearbook of the National Society of the Study of Education, Chicago, IL: Univer- sity of Chicago Press, 1968. Sacks, Seymour and others. City Schools - Suburban Schools: A History of Fiscal Conflict. Syracuse, NY: Syracuse University Press, 1972. Schever, Joan. "Levittown v. Nyquist: A Dual Challenge." Phi Delta Kgppan, Vol. 20, February 1979. Stavisky, Leonard P. "Dilemma for Decision Makers: Contra- dictory Education Research." Compact for Education, Vol. XIII, Summer 1979. Thomas, Allen J. School Finance and Educational Opportunity in Michigan. Lansing, MI: Michigan Department of Education, 1968. Townsel, Alvin H. "Adjustment for Urban Fiscal Problems in State School Finance Systems." Journal of Educational Finance, Vol. 2, Summer 1974. Wilson, Russell E. and others. The Equal Qualipy Plan. Lansing, MI: Task Force on Public School Finance, June, 1969. APPENDIX A SURVEY INSTRUMENT - -11 ""'l' Aq_v Anuv Aauv Aduv 'l"l'-ll ll AOLV 0"8l888'1 Amy as Aug on II, Any Acv l"l'.'" 1““ Any Amy ll": A~V 3533.. 3 an: we acuuuoaoua ou vououounv auouom unmaes verb Huuosoo ou tonnage huaawm N Anna nouuuav any usanuoua mo Ho>oa avg; o» vucuwoou oawu uo cowuuoaoum Anus :«v oucouuoaxm e—natoaeou o acwv.>ota so» «moo as» we «pas—o=< c< w¢~ua oocouuoaum 39:01.31; 543‘“. \ O u ucecauo you made: uaacu Heccouuod no uuou add: usaca nod umou u chcoouom vowuuucovn mo muondm Huuoh "Atenhe3 we.~l sm.mas.s mo~.mw ~me.e~ so .. ..aae.~ on ooo.~ eH~H>newa sq.mS sm.m_N.s «an.m~s amo.nm NA .. ..aam._ on can._ eonmaaaaawz .esfimma mu.m~ so._m~.s moo.oe_ seh.e~ no .. ..ame.~ on ooo.~ «meannxooum aunoflnem mh.mL ~_.~l~.~ meo.wm ~ma.m~ mm .. ..aaa.~ on eon.~ mm.qi am.m~m.~ oam.oa wmm.m~ mm .. ..aae.m on ooo.n mosexo ~m.el oh.se~.~ sea.mm _a~.- «N .. ..aaa.n on oom.n comm: m..ms ma.~mm.~ ~em.ess eee.- AN .. ..aaa.e on ooo.e saue>ez an.mx -.m~e.s e-.~e me~.a~ a .. ..aam.e ob oem.e use: .mesnema sham .w.ms “0.01:.s emn.mie see.e~ is .. ..aaa.m on ooo.n a”.el sm.~am.~ sna.hom n_o.em mN .. .aaa.m_ on ooo.o~ m. rm.el ea.nnn.~ mmm.me~ Nae.e~ a .. .aaa.oe on ooo.o~ madness .m mn.ai mo.~om.s mma.~e~ ~em.o~ _ .. ..te>o ecu ooo.on ”Mom mousuficcemxm manmumnee: pupae: muoauumfio msouo mnommw wcaumumdo o>uuu< vu< euwum no aficmuebsez muowuumfia wHQEmm mnm>< Hausa pom uwm coaumsam> Hepasz e>uuu< voufiamocm euwum naueam_ mom AmHmmmmmzmz an awesome .meonemHa Aooxom Ndlx 0mm m.znennoz mm.mm wa.~ om.om on.H~ om.a emm.oe Nao.e om~.0am.me~ maooeem sano>mz mN.mm m~.n oo.m~ oo.as oo.m Nea.e~ ems.~ eke.emm.em nHoonun aaoo emefitnxeoum H“ om.me oo.a om.mm eN.a~ oo.m Hma.mm emm.m mea.~mm.a~s «Hoonum change noaexo I om.~m om.e oo.a~ oo.ms oo.a asa.a~ mae.m em~.mam.ea mHoonem gasses some: me.om me.m oo.a~ oo.ms oo.a amm.s~ ema.s ma“.smm.mm mHoonem assess ehameg om.am on.“ oe.om oe.s~ oo.a mme.o~ mee.e mam.eoe.mm naeogem usages ”Hem om.me om.“ om.em om.m~ oo.a mm~.e~ Aao.~ ees.m~s.om nHoonom usansa beeswax mo.om mo.~ eo.m~ oo.ms oo.a eme.w~ was.~ mee.men.sm mHoonum change eflan>meaa e_.oe em.e m~.mm mh.e~ oo.a o-.m~ oe~.a~ Hes.s~m.sem amen Hooeom change weanemq mp.me mm.~ om.am oe.om oo.a s~w.ee mmm.e owe.ems.e- suspense Hooeom managed anew season: .emm meHm a.xm .uean< amazes mmmmzmz .>.m.m azaz HUHmemHa gases Emma +aausm --- ezHa.m.m ah-wkas mn-wamu nenne3 my Rom now a Non.H sane>m3 Hes SOH.H ae~.H emennnxoonm mNH eHe.H Hen.H noeexo mm eao.H mmH.H comm: mmH eds. HAN.H moaned HMS maH.H emm.H uses moH nm~.H mem.H humanem SSH Neo.H m-.H eHHn>meea oem Hom.H Hee.H wasmnea me n Nae.a n OHA.H a nuances nmem mmozmmmemHo mmmseHazmmxm ANN-eaaa .NHOH caneaasmv HonemHo ezHeammmo ezmmmso mmmseHazmmxm onam uzueammmo ezmmmso mum xficcedad APPENDIX C INTRODUCTORY LETTER 153 Appendix C MICHIGAN STATE UNIVERSITY EASTLANSING-KICHIGAN 48823 COLLEGE or EDUCATION . anucxsom HALL 25 March 1979 Dr. Matthew W. Prophet Administration Building Lansing School District 519 W. Kalamazoo Street Lansing, Michigan 48933 Dear Dr. PrOphet: The question of cost and its impact on prOgrammatic equity across school districts has been a major issue in current discussions of school finance plans. I am addressing this question in my doctoral dissertation, entitled “Programmatic Equity and the Cost of Education." I am asking your cooperation so that I might obtain the information needed.to complete my study. Dr. Donald Shebuski, Superintendent of Ingham Intermediate Schools, and Dr. Robert Muth, Assistant to the Dean, College of Education, agree that this is a major policy issue. Moreover, both are supportive of the study and are assisting me in the gathering Of information. The major objective of my study is to study cost in relationship to a comparable measure of programmatic equity. A survey instrument was develOped to weight and unitize personnel factors at the elementary building unit in an attempt to obtain a comparable measure of programmatic inputs across building units and districts. I wish to test my model on a representative sample of elementary schools in each of the twelve districts in Ingham County. Because of the develOpmental nature of the model, and because the type and nature of data to be collected are not easily attainable, on-site visits will be required (see enclosed survey instrument). In most instances I would need to meet with your business manager and the elementary principals of the buildings included in the sample. I will call on you within the next two to four weeks to explain in greater detail the purpose and nature of my study. I hOpe that at that time an appointment can be arranged for me to meet with your business official and possibly others from your district if the situation warrants. Your help will be most appreciated. A summary of the results will be mailed to you following completion of the study. Sincerely, Michael A. Boulus APPENDIX D CORRELATION MATRIX Lo» 0:3 .Lu a x3592 . $8.. 9; Ammmn ..Aoncow . 3:: £35 you ace... 0 a. .mmpw e; 88. 3.5 £35 to.— 3525 as; 22.. .8qu .omew e; H 32... u s . ~ . Anne—w aka 3.. 3.38 an: m. 0 22. :3... gum. mum 0:853 28.. 55 as; as w . as... .3..- . as. 28 - 8.85 338 .8.— a: 3:... 88¢ 38.. 28.. 23.. . . . 83. 2:. - 3 8.x. 38. N5. .3 as not 5a.. :3; :3. . $8.. $8.. :8; . . as. 2:. - 83. e. 38.- as as ea .88 :3. 2.2 :84 as; 83.. can. 23... 3.8. 38. a; 38. .8 .96 ...oeo 38m :1) anwow Ammwuw- 5.326 .3238. amuse: 88.. $3.. 88. . as. :8. 83. pontoon. co on»... $8.. $8.. 38.. :8. v 33.. $3.. :3. as... 3:. 28. as. 83. «33:. 9.3.28 3:... .33 :84 $3.. :23 23. 82.- 88. :8. as. as as $8.. 33.. $8.. 8 . $3.. :23 83. 88. :8. #8. e. a2...3§: £33m RIO as I! S SS 0V LV 0 S A an» an n «m an an Mmm mm mmm «um um .m mm «on: a. D. u D. .. .3! .J .47: O. 30 u. .4 IO 1 a H... d. . d. .u. 0 Wm. . do I... d u u I I u 0 .4 d .d O . u I o. 0 d 3 a 3 o a a J 3 I a m. 0 ’ .. x532: onac ccc.mm www.mH cmc.ma cem.aa c cccccfl - accma ace.mm csc.mH ccc.cH ccc.HH AH cccmd - Home ecc.flm cmm.mH mmm.ca Hec.HH a come - accm mmm.Hm ccH.mH ccH.mH Hm~.HH HA cccm - Hmam mcc.cm HcH.mH mmc.cd com.HH ma cmmm - accm mam.cm cam.HH csc.en mcm.~H ma cccm - acmm cmm.c~ ccH.NH ccc.ca mmm.a~ cm ccmm - Accm Hmm.mH acc.HH ccc.ea mcc.HH c cccm - Hmea ecm.ca acc.mfi mm~.ea ccH.HH ca cmsd - Hcmfl ccc.cfl cmc.HH mc>.cfl csc.cH ~H ccmH - HmmH ccm.ca ccc.HH mcm.cH mac.cH c cmma - Hccfl mcm.ca mam.HH mcm.c~ cce.cfl AH cccd - amp smc.ca mmm.HH acc.cfi smm.ca c cmm - Hcm mcc.ea mm=.HH mmc.cH cec.cfi ma ccm - Hmm mmc.mfi ccm.HH cmc.sfl ecm.HH m cmm - Hc scenes: a: esefixcz cm gasses: cm mmmmmmmmm nccccscm EsEwaz <2 . no tones: do penesz Aeufim uceEHHoscm may mem< m xficcedo< APPENDIX F PERCENT OF SUCCESSFUL SCHOOL MILLAGE ELECTIONS Accszs accEccS mcozoefi enema"). Loose summecoam co cceocea .m chcmam’w c\oo r c\c ON c\com excom case ‘5“ APPENDIX G STAFF COMPENSATION AND DISTRICT SIZE 157 Appendix G STAFF COMPENSATION AND DISTRICT SIZE Average Professional Benefits Membership Salary Per Staff Over 7,500 $16,073 $1,867 N=50 2,000-7,500 $14,211 $1,533 N=221 Under 2,000 $12,365 $1,266 N=259 " State Average $13,491 $1,434 (K-lZ) Source: James A. Phelps, Michael F. Addonizio and Thomas S. Nicol, State Aid For Michigan Public Schools: An Analysis of Two Proposals, 1979 (mimeo) APPENDIX H COMPARATIVE INCREASES IN TAX EFFORT camczcfiz co nHooccm caficsm cc» soc ncacnacccm Hetecec "eonsom Amc.ccmm.=m Amc.cccc.cm Amc.ccfim.mm Amm.mccm.m~ cc.~m secs Amc.cccc.mm AmH.HHccc.mm Amc.cvam.~m Amc.ccam.- cm.mm cums Amc.vae.mm Amm.HHccm.mm Amm.HcAs.mm Amc.ccflm.mm Hm.- cccm Mm Aum.cccc.em Auc.mcdm.mm Amm.cccc.mm Amc.=ccc.c~ cm.mm cccm Amm.mccc.cm Aua.mvcc.~m Aum.=vc~.cm Ame.mccc.mm cm.cm cccc Amc.accc.mm Amc.mvcc.cm Amm.mccm.sm Ame.mcmc.cm -.mm cccm Amc.cccc.cm Auc.mccc.cm Amc.~cmm.cm Am~.mv=c.em ca.em cccc Ams.flcce.am Amm.=ccm.flm Amm.chc.cm Amm.accm.cm c=.c~ cccs Amc.H-cmm.mm Amc.mcmw.mm Amm.cccm.mm Amm.dcmc.cm mc.cm cccc Aum.H-V2m.mm Aua.mvmc.cm Amm.mvzs.mm Aum.ov:~.om sw.om cumm Amm.m-cmc.cm Amm.cccc.mm Amm.cccm.sm Amc.mvce.mm mc.Hm nccm mmnmmwm mmummmm Hmnmwww mmuwwwm “Hummmm eflccccccca Hmommm x<8 2H mmmHB