Disparities in Public School District Spending

February 1995

The following contains an excerpt from Disparities in Public School District Spending. A full copy of this report is available in portable document format (Adobe Acrobat PDF).
You need the Acrobat Reader software to view these files.

Download the Acrobat file now. (1.16 MB)

For questions about the content of this report, please contact Frank Johnson (frank.johnson@ed.gov).

Executive Summary

The Purpose of This Report

Finance issues are among the most fundamental to public education. Questions relating to "who pays," "how much," and "for whom" are central to the concepts of equity and adequacy, which have long been at the heart of public education fiscal policy. Equity issues focus on the fairness of the overall public education allocation system. Given our decentralized system of public education, it is not surprising that more public dollars are spent on the education of some school children as opposed to others. These differences may not be undesirable or unwarranted. Given the variations in the cost of education resources and in the needs of students that are known to exist across districts, equal dollars per student may not result in equal education opportunities. Thus, a major fiscal policy question is: Where do expenditure differences occur and to what degree? If expenditure differences are simply related to the differing capacities of states and localities to purchase public education services, at what point do these differences result in inequities for different types of students? These expenditure differentials are especially of interest as they relate to children in particular categories of historical concern, such as minority status, poverty, and other at-risk factors.

To the extent that these types of questions pertain to unequal allocations of public education resources to students with comparable education needs, they are considered to be horizontal equity issues. Vertical equity relates to expenditure differences justified by the differing education needs of students. For example, all public education funding formulas allocate different amounts of revenue to districts to account for the differing education needs of some types of students (e.g., special education). Vertical equity questions relate to which kinds of students should be eligible for additional aid and the appropriate size for these supplements. Both of these sets of equity issues closely relate to the adequacy of education revenues. Are they sufficient for their intended purpose? As the exact purposes of education are not fully agreed upon and the technology of education is not well understood, objective determinations of whether given amounts of education resources are sufficient for their intended purpose generally cannot be made. For this reason, adequacy issues are most often expressed in terms that are relative to some specific standard, and all of these traditional fiscal policy issues revert to basic questions about who is receiving how much and for what purpose.

This report begins to answer these and other important school finance questions in ways that have not been previously reported. Measures of how much districts receive in public funds from local, state, and federal governments and how these resources are used to provide public education services are available for all of the school districts in the country, and these measures are matched to such important district characteristics as the percentage of children living in poverty, the percentage of minority children, and average wealth. Through the use of resource cost factors, alternative measures of district spending are expressed in terms of relative "buying power," and through the use of student-need adjustments, variations in the number of students with additional education requirements are also taken into account. Multivariate, as well as bivariate, analyses are used to isolate the impact of individual district and community factors on variations in expenditure. In addition to these adjusted forms, all of the resource measures included in this report are presented in their original (unadjusted) form for comparative purposes.

As an example, table A has been extracted from the main body of the report to illustrate the differing results that can be obtained through the use of bivariate and multivariate analyses. These data show that while the bivariate results indicate a positive relationship between student/teacher ratios and the percentage of minority enrollment, the multivariate data indicate the exact opposite relationship between these variables (columns 1 and 3). This difference results from the fact that while the bivariate results show the direct relationship between these two variables, the multivariate analyses also take simultaneously into account the effects of a number of variables believed to be relevant to variations in student/teacher ratios. Thus, while the bivariate analysis suggests a positive relationship between these two variables, the multivariate analysis indicates that this result is really an artifact of the relationship between the percentage of minority students and other related variables, such as the percentage of students in poverty and district urbanicity. Using the multivariate analysis, in which the relationship between all of these related variables and student/teacher ratios are considered simultaneously, the relationship between percentage minority enrollments and student/teacher ratios becomes negative.

This example illustrates the importance of adding multivariate analyses to gain a fuller understanding of the relationships among the variables presented in this report. The general trends documented in this report primarily focus on the resource-cost and student-need-adjusted multivariate (fully adjusted) results. Although multivariate results are emphasized, the report notes when these general trends differ substantially from the actual bivariate results.

Table A.-- Student/teacher ratios by percentage of minority enrollment

-----------------------------------------------------------------------------------------------
                                        Bivariate Results              Multivariate Results
                                        -----------------               --------------------
                                                                             Student-Need-
                         Percentage of            Student-Need-   Unadjusted      Adjusted
District Characteristic     Enrollment    Actual       Adjusted    Estimates     Estimates
                                            (1)         (2)          (3)            (4)
-----------------------------------------------------------------------------------------------
Minority Enrollment
  Less than 5%                    21.9      17.0        19.1         18.7           21.5
  5% - <20%                       26.5      17.8        20.2         18.0           20.7
  20% - <50%                      25.7      18.2        21.0         17.7           20.4
  50% or more                     25.9      18.7        22.2         17.5           20.3
-----------------------------------------------------------------------------------------------

SOURCE: U.S. Department of Education, National Center for Education Statistics, 1989-90 Common Core of Data, 1990 Census School District Special Tabulation (summary file set I).

NOTE: All results are weighted by district enrollment. Percentages may not add to 100 due to rounding and missing categorization information for some observations.

Data Sources and Procedures

This report addresses school finance policy issues through the analysis of school district revenue and expenditure data from the 1990 Survey of Local Government Finances-School Systems (F-33) collected by the U.S. Bureau of the Census, as the collection request for the National Center for Education Statistics (NCES). To increase the policy relevance of these analyses, these fiscal data were matched to other NCES databases that provide more descriptive information about the districts and the communities in which they are located.

Three types of procedures were used to analyze these data:

Comparisons of the actual and resource-cost and student-need-adjusted amounts of education resources received by different types of school districts and communities (bivariate analysis).
Comparisons of the actual and resource-cost and student-need-adjusted amounts of education resources received by different types of school districts and communities, holding other factors constant (multivariate analysis).
Measures of the degree of variation in resource quantities across districts (dispersion analysis).

Policy Questions

Four important policy questions that relate to the financing of public education are addressed in this report.

How do education resource measures, such as total expenditures per student, vary in different types of school districts and communities across the nation?
How do school districts serving different types of students and communities allocate resources across the categories of instruction, administration, and capital outlay?
How do local, state, and federal revenues vary for school districts serving different types of students and communities?
To what extent do education resource measures vary across the nation?

The first question is addressed by examining alternative measures of school district spending and the relationships of these measures to such district and community characteristics as the percentage of students in poverty, the percentage of minority students, and average property wealth in the district. Addressing the second question involves breaking out aggregate expenditure measures into the more detailed categories of instruction, administration, and capital outlay. To address the third question, revenues are examined to assess varying reliance on local, state, and federal sources to support education services in different types of districts. The final question represents a departure from the basic approach used for questions one through three. The section of this report that addresses this question considers all of the school districts across the nation to assess total disparity in resource measures such as total expenditures.

Summary of Findings

How do education resource measures, such as total expenditures per student, vary in different types of school districts and communities across the nation?

Of all the variables included in this report, the largest variations in average public education expenditures occur between regions. Fully adjusted expenditures per student are substantially higher in the northeastern region of the country and are lowest in the West ($5,293 versus $3,632).
Public education expenditures per student are higher in the nation's smallest districts whereas students receive an average fully adjusted expenditure of $4,862 versus $4,216 in the largest districts (10,000 students and above).
Greater education expenditures per student are associated with higher community socioeconomic status as measured by the value of owner-occupied housing ($4,401 versus $3,992, fully adjusted), or by education attainment ($4,515 versus $3,953, fully adjusted). However, this relationship is less pronounced when socioeconomic status is defined in terms of median household income. When the relationship between this variable and education expenditures is considered in isolation, only the wealthiest group is considerably different from the other groups.
More money is spent in districts with the highest percentages of minority students compared to districts with the lowest percentages of minority students ($4,514 versus $3,920). Although minority students in poverty are often viewed as those least served by current systems of public education funding, these findings suggest that while inequalities may remain for students in poverty, they do not appear to be driven by minority status.
Public education expenditures per student are highest in low poverty districts. The fully adjusted differential between the highest and lowest poverty districts is $309 per student ($4,219 versus $4,528). However, this relationship is not linear and affects only the 11 percent of students in the wealthiest districts. Among the other 89 percent of students, the variation is only $8 per student
Districts with the highest percentages of students in special education show higher overall actual expenditures than do districts with the lowest percentages of special education students ($5,447 versus $5,061). However, when differences in the cost of living and the added cost of serving students with supplemental needs are included, an opposite expenditure pattern is observed ($4,219 versus $4,510).

How do school districts serving different types of students and communities allocate resources across the categories of instruction, administration, and capital outlay?

Student/teacher ratios vary substantially by district size and region of the country. In both actual and adjusted terms, the average ratio is over 20 percent larger in the nation's largest districts (10,000 students or more) as opposed to the smallest (less than 1,000 students). In actual terms, the average student/teacher ratio in the smallest districts is 15.1 as compared to 18.8 in the largest districts. These differences are even more pronounced by geographic region of the country, with the Northeast showing an average student/teacher ratio of 15.6 as compared to 21.9 for the West. The Midwest and South show average student/teacher ratios falling between these two values, at 17.3 and 17.1, respectively.

[Note: The following expenditure results were obtained only through bivariate analysis and do not control for other district characteristics. As such, they should be interpreted with greater caution than the results obtained through multivariate analysis.]

Districts serving relatively high percentages of students in poverty, minority students, or limited English proficient (LEP) students allocate greater percentages of their funds to core instructional purposes than do districts serving lower percentages of students in these same groups.
Capital outlay is the area of expenditure found to be the most sensitive to variations in total district spending compared to the categories of instruction and administration Districts with less to spend tend to focus on direct instruction and administration at the expense of capital expenditures.
By region, districts in the northeastern section of the country spend more in the area of administration and support than the other regions (the Midwest, the South, and the West), and districts in the West spend appreciably less ($1,371 versus $831 per student).

How do local, state, and federal revenues vary for school districts serving different types of students and communities?

The amount of local support for public education rises with the wealth and socioeconomic condition of the community.
State funds are the primary equalizing force in public education resource allocations.
Although state and federal allocations are larger in districts with large numbers of special, compensatory, and LEP students, based on the student-need adjustments used in this study, these additional funds appear to be insufficient to offset the supplemental cost of these programs.

To what extent do education resource measures vary across the nation?

The distribution of public education resources is substantially more nearly equal than wealth measured by housing values, and somewhat less varied than wealth measured by household income.
State public education allocation systems are the primary equalizing factors of education resources, with some additional equalization resulting from the various federal funding programs.

Implications for Further Research

The results obtained through multiple regression demonstrate that school district spending substantially varies by geographic region, as does student/teacher ratio. This substantial difference in education funding by region may be a matter of concern-the implications of which need to be thoroughly explored by the school finance research community.

Two findings of this study run somewhat counter to initial expectations. First, spending by school district size is higher only for those school districts with under 1,000 enrollment, other factors being equal. This seems to suggest that the expected diseconomies of small scale only hold for the nation's smallest school districts, a finding that may elicit further study.

A second, somewhat surprising, finding is that more money is spent in districts with the highest percentages of minority students ($4,514 versus $3,920), holding other school district and community characteristics constant. This finding should be further explored by school finance researchers.

A less surprising finding is that greater total expenditures per student are associated with higher community socioeconomic status, measured by the value of owner-occupied housing (per student expenditures of $4,401 versus $3,992, other factors held constant), or by education attainment ($4,515 versus $3,953). Differences in student/teacher ratios also appear but are less than 1.5 students per teacher.

When socioeconomic status is measured by cost-adjusted median household income, however, and all other factors are held constant, the expenditures per student between the highest and lowest income groups differ by only $186 ($4,382 versus $4,196). These findings relating alternative measures of socioeconomic status to education spending provide a fertile field for further research.

Public education total expenditures per student are highest in low poverty districts, but unlike socioeconomic status, the relationship is not linear. Controlling for other factors, the differential between the highest and lowest poverty districts is $309 per student ($4,219 versus $4,528). Controlling for other school district characteristics, only school districts in the category with the fewest children in poverty spend substantially more per student.

It is also worthy of note that smaller dollar differences in per student expenditures are observed when core instructional, as opposed to total, expenditures are examined by socioeconomic status. Core instructional expenditures reflect the central purpose of the local education agency, which is to educate children. These findings suggest that lower wealth districts appear to be investing a larger percentage of their spending on core instruction, rather than on other areas. Although such findings mitigate total expenditure differences by socioeconomic status, they may also indicate that poor school districts are deferring needed school construction, renovation, and the purchase of instructional equipment. The current study is unable to completely explore the differences in spending for capital outlay, which includes both school construction and purchases of equipment. This also provides yet another opportunity for further research.