«Introduction This paper examines the fiscal condition of school districts in Nebraska. Our methodology is a new one that has been applied to ...»
Moreover, export ratios vary widely among the smaller districts, but not among the largest districts.
Table 4 also reveals that export ratios are especially high in Lincoln and Omaha (classes 4 and 5) and in district classes 3 and 6.
Revenue-raising capacity from the property tax, expressed in index form, is presented in Table 5. For convenience, the average district has an index of 100. A school district with a capacity of 150 has 50 percent more capacity than the average school district. Capacity through the property tax is highest in the smallest and largest districts. It is also higher in classes 4 to 6 than in classes 1 to 3. As before, however, capacity varies widely within each enrollment group or district class. Note that class 1 districts, with an average size of only 26 students, have a relatively low average capacity, whereas districts with fewer than 10 students have relatively high average capacity; among the class 1 districts, the smallest districts apparently have the greatest capacity per student.
Revenue -Raising Capacity From Other Sources School districts also collect revenue from charges and fees. According to Table 2, charges and miscellaneous revenues account for 18.6 percent of total own source revenue.
The average burden across the state for charges and miscellaneous income equals total revenue from these sources divided by aggregate income. Because miscellaneous charges and fees apply mainly to residents, we assume that these revenue sources have no export potential. One exception to this rule arises when non-resident students pay charges for services in the receiving district, such as school lunch or activity fees, or when non-resident adults pay to attend activities such as athletic events. This export potential, however, is extremely small and is unlikely to significantly affect the revenue-raising capacity of a district.
As a result, a district's revenue-raising capacity per student from charges and miscellaneous income equals the average burden multiplied by the income per student in that district.
Because it depends only on income per student, revenue-raising capacity from other sources has the pattern of variation shown in Table 3. This type of capacity ranges from $11 to $1,072 per pupil.
On average, the smaller the district, the larger its capacity from these sources. In addition, this capacity is highest in district classes 1 and 6. As in the case of income, however, this capacity also varies widely within each enrollment group and district class.
Overall Revenue -Raising Capacity To find overall revenue-raising capacity, we first add capacity from the property tax and capacity from other sources. We then adjust this sum to reflect capacity gained or lost through interschool district payments of tuition and transportation fees, which are commonly made in Nebraska. Many school districts in the state either send some of their students to other districts or receive students from other districts; in some cases they both send and receive students. Although the reasons for these student exchanges vary, a district's capacity is affected by each student sent or received. When a sending district makes a tuition payment the dollar amount of that payment directly reduces the capacity available to that district to educate the remaining students. Likewise, when a district receives a student, its overall capacity is enhanced by the dollar amount of the payment.
The overall revenue-raising capacity of Nebraska's school districts, expressed in dollars per student, is summarized in Table 6. In Table 6 the average capacity per student is set equal to the average current spending per student in Nebraska. This standardization does not change the relative position of one school district to another; it simply facilitates comparison of capacities and needs later in our analysis.
The variation in overall revenue-raising capacity for school districts is quite substantial. The range is from zero up to $29,967 per student. Six school districts, very small in population and income, have capacities of zero. The correct interpretation of this result is that these districts would not have any revenue left over to educate their own students if they imposed the baseline tax burden upon their residents and paid the tuition and transportation charges they owe other districts. In a few cases, a district makes payments to other districts, which are in excess of their calculated revenue-raising capacity from property taxes and charges. No district has a calculated capacity very far below zero, however, so for ease of interpretation we set minimum capacity to zero.
In contrast, some districts have dramatically high revenue-raising capacity. Because of a very high income per student, one school district has a revenue-raising capacity per student of almost $30,000.
The average capacity per student is $3,504, so this district could raise over 8-1/2 times as much revenue per student at the baseline tax burden as the average district.
For districts with enrollments greater than 20, capacity per student tends to increase as enrollment rises.
The enrollment size groups with the highest overall revenue-raising capacity are districts with more than 9,999 students and districts with fewer than 10 students. Districts with enrollments between 20 and 29 students have, on average, the least ability to generate revenue. Overall capacity is also highest in district classes 4 to 6. Within each enrollment size group and district class, however, there is substantial variation in capacity.
A school district's expenditure need is the amount it must spend to provide an education of average quality to each of the children enrolled in its schools; in other words, a district's expenditure need is the inverse of its costs for providing public services. The concept of expenditure need, like the concept of revenue-raising capacity is designed to facilitate comparisons across school districts by holding constant educational quality and focusing on factors outside the control of local officials. Comparisons based on actual education spending can be misleading. Some high-spending districts may be poorly managed or may face relatively high costs and may therefore receive very little education for their money.
Moreover, actual spending is controlled by school officials and therefore cannot be an objective basis for comparison.
Our method for estimating expenditure need was developed by Bradbury et al. (1984), Yinger (1986), and Ladd and Yinger (1989). Although this method was designed to estimate the cost of municipal services, it can readily be applied to the cost of education. At least since the work of Bradford, Malt, and Oates (1969), scholars have recognized that the cost of public services depends on the characteristics of the jurisdiction as well as on the wages of public employees. The problem is that the relationship between community characteristics and public service costs is difficult to estimate. This method provides one way to solve this difficulty.
Some school district characteristics outside the control of school officials influence school costs because they are part of the technology of producing school outputs; schools with harsh or unfavorable characteristics must spend more to obtain the same output. Thus, one way to determine the role of district characteristics is to examine the technology of public production, that is, to look at the relationship between some public output, such as school test scores, and both inputs, such as teachers, and school district characteristics. This is the approach of the literature on educational production functions, which is reviewed by Hanushek (1979, 1986). Many of the studies in this literature find support for the important role of district characteristics on the technology of local schools. See, for example, Summers and Wolfe (1978) and Henderson, Mieszkowski, and Sauvageau (1978).
The problem with this approach for our purposes is that it requires a measure of public output, again such as school test scores, that inevitably leaves out many of the services provided by a school district.
As is widely recognized in the literature, neither test scores nor any of the other output measures that have been used adequately summarize all the services a school provides. To measure fiscal condition, however, we must be comprehensive, that is, we must estimate the cost of providing all school outputs.
As a result, we use an alternative approach that is less precise but more comprehensive.
Our approach starts by recognizing that, by definition, school spending equals the level of school output multiplied by the cost per unit of output. According to an extensive literature on local public choice (see Inman, 1979, or Rubinfeld, 1986), the level of school output selected by voters depends on a school district's income, tax-price, intergovernmental grants and a variety of other factors. As just explained, the cost per unit of output depends on teacher wages and various school district characteristics. As a result, one can estimate the impact of wages and school district characteristics on costs by estimating the relationship between the relationship between these variables and school spending, holding constant income and other determinants of the level of school output. For a detailed explanation of this method, see Ladd and Yinger (1989, Chs. 4 and 10). The set of variables we use to hold the level of school output constant is presented in the appendix.
Several studies have examined economies of scale in public education, which is one aspect of public school costs. These studies are reviewed in Cohn (forthcoming) and Kenney (1982). Some of these studies explain school spending per pupil as a function of the number of students in the school, holding school quality constant. School quality is held constant by including an output measure, such as a test score, as an explanatory variable in the analysis. In contrast, we control for school quality indirectly by including variables that influence voters' decisions about school quality. Although our controls are indirect, our approach has the advantage that it builds on the extensive literature on public expenditure determination and that it can readily be extended to aspects of school costs other than economies of scale.
We find five factors that are largely outside the control of school officials and that influence the cost of public education in Nebraska. Because of data limitations, we were unable to estimate the impact of local area wages on the cost of providing education.
Our measure of expenditure need reflects the impact of each of these factors on the cost of education.
The first factor is the number of handicapped students the district educates. The available data indicate the number of handicapped children who live in each school district. Because not all districts educate and transport the handicapped children that live within their boundaries, however, we make adjustments in the number of handicapped children by district to reflect the shifting of responsibilities of these students from district to district. We find that for every increase of 1 percentage point in the ratio of net handicapped students to total students the total operating costs per student of a district rise by $9. This cost factor does not appear to operate in tiny school districts (fewer than 10 pupils), perhaps because they rarely educate severely handicapped students themselves.
The second factor that influences expenditure need is transportation cost, as determined by the number of students that the district is required to transport and the number of miles these students must be transported. We do not have an exact measure for the latter variable but we are able to closely estimate its impact by including the total route miles traveled in each district into our analysis and by controlling for other factors. See the appendix.
Our analysis reveals that, controlling for other factors, a district must spend an additional $84 dollars for an increase of 1 percentage point in the ratio of students eligible for transportation to total students.
Districts that either are not required to transport students, such as Class 6 districts, or districts that are not spread out and have few students to transport do not have to spend as much as other districts to achieve average-quality education.
The third factor is the proportion of secondary students to elementary students in the district. Not surprisingly, we find that, on average, it costs more for school districts to provide an average quality education to secondary students than to elementary grade students. Normally, additional services and activities are provided for secondary students, such as more extensive curriculum and sport programs.
To be specific, we find that for every increase of 1 percentage point in the ratio of secondary to elementary students a district must spend $8 more per enrolled student.
The fourth cost factor is that larger school districts in Nebraska apparently can take advantage of significant economies of scale. We find that the cost of education per pupil declines as the total number of students in a district increases. This relationship is not linear, however. In the case of small changes in enrollment, for every increase of 1 percent in enrollment a district needs to spend roughly $5.39 less per student. Large increases in enrollment can have an even more dramatic impact on costs per pupil.
If district enrollment is increased by 10 times, for example by combining 10 districts with 10 students each into a single district, then the cost per student will decline by $1,241. These economies to scale reflect the fact that larger districts can spread out the cost of administration, libraries, and other systemwide activities over a larger number of students. As noted earlier, several other studies have examined economies of scale using a different methodology. Some, but not all of these studies also find that economies of scale exist. See Cohen (forthcoming) and Kenney (1982).