A strict delineation between general and categorical revenues is, by definition, somewhat imperfect because flexibility in allowable use varies somewhat across general and categorical revenues.6 However, this type of breakout provides a perspective on the amount of revenues that different types of districts have for general purposes over which they have a high degree of discretion and control, and on the levels of revenues that are generally earmarked for specific purposes. For categorical funds, discretion in use is often severely limited. This type of analysis provides a different perspective on the levels of resources received by differing types of students and districts. For example, two districts receiving comparable levels of total revenue per student may face very different circumstances in terms of discretionary buying power. Separating general from categorical resources allows more in-depth analysis of the true spending power for general education purposes of districts receiving comparable total revenues per student. For example, while two districts may be very comparable in terms of total revenues per student, substantial differences in the extent to which their revenues are comprised of categorical versus general revenues will considerably impact the degree of control they have in deciding how these funds should be spent.
Summary of Findings
How do general, categorical, and total revenues available for public education vary for different types of school districts and communities?
Analysis and Structure of Tables
The tables in this section of the report present average (columns 3 through 6) general, categorical, and total revenues per student (groups of rows) for different categories of district and community characteristics (rows within groups).7 The average dollar values are shown in actual, cost-adjusted, need-adjusted, and cost- and need-adjusted forms using the cost- and need-adjustment factors described in chapter 1.
In assessing the relationship between two listed variables, it is important to examine all four of the alternative sets of results shown in each table (columns 3 through 6). Any single set of numbers shown in isolation from the others may present a very different set of interpretations than viewing the full set of actual and adjusted findings.
The tables in this section also show the percentage of students represented in each of the district and community characteristic categories (column 1) and the percentage that each of the revenue groups represent of the total revenue (column 2).
School-Age Children in Poverty
Column 3, table II-1, shows the average actual revenues per student for each of the four percentage of children in poverty categories under each of the three revenue groupings (general, categorical, and total). In the general revenue grouping, per student, actual revenues are highest for students in the lowest poverty category of districts (less than 8 percent poverty) compared to other districts ($5,555 compared to $4,458, $4,079 and $4,193). The next revenue grouping under column 3, shows that actual categorical revenues per student increase with the percentage of school-age children in poverty. Federal and state categorical programs provide 98 percent more revenues for students in high poverty school districts than for those in low poverty districts. The highest poverty districts receive $695 more per student than districts in the lowest poverty category ($1,406 minus $711). The last grouping under column 3 shows that actual total revenues per student are highest in the lowest poverty districts, which is not surprising given that education is largely funded through local sources. High poverty districts are likely to have greater difficulty than their lower poverty counterparts raising money, because high poverty districts tend to have relatively small tax bases. Although categorical programs are not providing enough additional revenues to supplement the education of the highest poverty districts, in districts with students who are most in need of supplemental education, these revenues act to decrease the revenue differential between the lowest and highest poverty districts from 32 percent ($5,555 versus $4,193) for general revenues to 12 percent ($6,266 versus $5,600) for total revenues.
On adjusted bases (columns 4, 5, and 6), the patterns of general, categorical, and total revenue allocation are similar to the actual pattern, except that with adjustments, overall revenue values are expectedly lower. Analysis of all four forms of revenues (columns 3 through 6) show that Federal and state categorical programs provide about twice as much revenues for students in high poverty school districts than low poverty districts.
Although there is not a large difference in the total revenues per student across categories of children in poverty, there is a large difference in the amount of discretion that districts have in allocating these funds. Column 2 shows that for districts in the highest poverty category, about 25 percent of their total revenues comes from federal and state categorical programs, while categorical funding represents only 11 percent of total revenues in the lowest poverty districts. Since categorical funds come attached with regulations on how the district must spend the money, it means that high poverty districts have discretion over only 75 percent of their budget, while low poverty districts have discretion over 90 percent of their education resources. Low poverty districts have much more flexibility in deciding how to allocate education resources.
When actual revenues are cost- and need-adjusted (column 6), the same pattern for general, categorical, and total revenues exists, although the categorical revenue difference between the highest and lowest minority districts drops from $802 per student in actual dollars to $575 per student ($1,172 minus $597).
table II-3), in cost-adjusted and cost- and need-adjusted terms, more resources go to the students in the nation's smallest school districts (columns 4 and 6). In cost- and need-adjusted terms (column 6), districts in the smallest category of district enrollment receive $4,948 per student, compared to $4,558 per student in districts with the highest enrollments (figure II-3). The smallest districts also tend to have higher levels of general revenues across all actual and adjusted measures. This may be due to higher costs resulting from diseconomies of scale, a cost factor not accounted for in this analysis, or it may be due to districts with higher resource levels preferring to stay small.
In actual and adjusted dollars (columns 3 through 6), categorical revenues increase with district size. For example, in actual dollars the largest districts receive 49 percent more categorical dollars than the smallest districts ($1,289 versus $865) and in cost- and need-adjusted terms this differential is 35 percent ($1,028 versus $764).
District type refers to whether it is an elementary, secondary, or unified district. While the vast majority of students (97 percent) are enrolled in unified districts, which serve both elementary and secondary students, a small percentage of students are enrolled in districts serving only elementary or only secondary students. Because it is more costly to serve high school students (Hertert, Busch, and Odden 1994), is it not surprising to see, as shown in column 3 of table II-4 and figure II-4, that districts serving only high school students have higher total revenues per student than other districts ($7,192 compared to $6,122 and $5,509). Also interesting to note is that elementary district and unified district total revenues are essentially the same in cost- and need-adjusted terms ($4,870 and $4,635, respectively). This is surprising in that it is more costly to serve secondary school students, yet unified districts serve secondary, as well as elementary, students. One possible reason for this finding is that unified districts may be able to spend less by sharing administrative costs across all grade levels.
table II-5 and figure II-5. This pattern also holds in terms of adjusted averages (columns 4 through 6).
In actual and all three adjusted terms, districts in the Midwest receive significantly lower levels of categorical revenues than other geographic regions ($697 compared to $1,204, $1,113, and $1,093, in actual dollars).
Districts in the Northeast receive the most total revenues. For example, in the Northeast, districts receive $2,303 per student more than districts in the Midwest ($7,769 minus $5,466). Districts in the South and West receive the lowest total revenues at $4,890 per student and $4,992 per student, respectively. There is a $2,879 per student (or 59 percent) difference between districts in the Northeast and districts in the South ($7,769 minus $4,890). When these values are cost- and need-adjusted, Northeast districts still have the highest revenues ($5,846) and districts in the West have the lowest total revenue ($4,116), a difference of $1,730 (or 42 percent).
table II-6, although actual general revenues per student are substantially lower in rural districts (column 3) than in other districts ($3,963 compared to $4,476 and $4,833), this differential is reduced substantially when expressed in cost-adjusted and cost- and need-adjusted terms (columns 4 and 6). This pattern is also evident in the results of total revenues per student. Actual total revenues per student in rural districts are $4,894 compared to urban/central cities and suburban/metropolitan areas at $5,781 and $5,748 per student, respectively (figure II-6). When expressed in cost- and need-adjusted terms (column 6), the differential in the total revenues per student is reduced ($4,597 versus $4,593 and $4,730). These reductions in general and total revenue variations may be largely due to the lower costs exhibited in rural areas.
Urban districts receive more actual categorical revenues per student than suburban and rural districts ($1,305 compared to $914 and $932, respectively). This general pattern also holds true in need-adjusted items.
figure II-7) and 14 percent in cost- and need-adjusted terms ($4,677 and $5,321).
Although there are no clear patterns between categorical revenues and housing values, the pattern of total revenues per student by housing value is similar to that found for general education revenues. In actual terms there is a $1,556 or 31 percent differential between the lowest and highest housing value categories ($5,018 versus $6,574, in figure II-8). In cost- and need-adjusted terms (column 6), the difference between the highest and lowest housing categories is $424 or 9 percent ($4,988 versus $4,564). In cost- and need-adjusted terms (column 6), the total revenues per student are about the same across all housing categories.
 For example, most impact aid funds may be spent on children who are not federally connected (e.g., do not live on a military installation or Indian reservation). In addition, in some states the funds districts receive for students with disabilities may not be restricted for use on this population.
 District and community characteristics were broken down into approximate quartile or quintile categories while still making logical breaks. Some characteristics had their own logical sub-categories (e.g., metropolitan status was broken into urban/central city, suburban/metropolitan, and rural). See appendix D for more information.
 See, for example, the arguments of Taylor and Piche, 1990.