Where Has the Money Gone?
In this belt-tightening era, how money for public education is spent has become a centerpiece of highly rhetorical debate. It will become increasingly contentious because public school enrollments are increasing for the first time in two decades, causing education budgets to rise further. Meanwhile, highly publicized claims that there is little relationship between spending increases and student achievement have created a hostile environment for educators seeking new funds.
Economic pressures and rhetoric that "money doesn't matter" will require educators to have clearer explanations of the purposes for which money has been spent and the specific outcomes expenditures are designed to enhance. Academic achievement of regular students is only one of many intended outcomes of today's public schools. Since the 1960s, different student populations bringing new challenges, new federal and state legislative mandates, and court rulings have created a host of new responsibilities. In many cases, these mandates have required reallocation of existing resources or created funds which come with strict stipulations regarding their use. Schools must prepare handicapped children for more independent living (special education), fulfill welfare needs (the school-lunch program), promote racial integration (desegregation), teach English to immigrant children, and provide education even to the most unruly, unmotivated students. To evaluate the productivity of school spending and hold schools responsible, we must be able to link expenditures to programs. If, for example, schools have funneled much of the new money to diagnose, monitor, instruct, and provide services to special-education students, it makes no sense to judge the effectiveness of this spending by whether Scholastic Assessment Test scores improved for the college-bound.
Unfortunately, school-finance reports almost never detail expenditures by "program," so meaningful assessment of school spending and productivity is next to impossible. Instead, districts, states, and the federal government almost always report school expenditure by function (instruction, support services, etc.) or by "object" (salaries, supplies, etc.). To show how it might be done differently, the Economic Policy Institute issued a report last week that details the programmatic uses of education money from 1967 to 1991, and how the growth of expenditures for different purposes compared during that period. (See story, page 1.) With support from the Metropolitan Life Foundation, we examined detailed expenditure records from nine districts, chosen to mirror national experience. The nine included districts that were urban, suburban, and rural; large, middle-sized, and small; from affluent, middle-class, and poor communities; with high, average, and low percentages of minority and non-English-speaking students.
We found that these nine districts increased their per-pupil spending by an average of 73 percent from 1967 to 1991, but barely one-fourth of the increase went to support regular education. Regular-education spending dropped from 80 percent of all spending in 1967 to 59 percent in 1991. About 60 percent of the new dollars went to support more diagnosis, instruction, and other services for special populations (including special, compensatory, and bilingual education, desegregation, and programs to keep students "at risk" from dropping out of school).
Special-education programs represented the largest component of new spending on special populations. More than one-third of the net new dollars went to support special-education students; this program rose from 4 percent of all expenditures in 1967 to 17 percent in 1991. Every category of per-pupil special-education spending grew substantially, but some grew faster than others. Teachers declined from 44 percent of special-education spending in 1967 to 37 percent in 1991. This decline was largely offset by growth in special-education paraprofessionals, which grew from an insignificant cost in 1967 to nearly 8 percent of all special-education spending. Special-education support (administration, contracted services, and supplies and equipment) stayed at near 20 percent of special-education spending. But within this support category, administration took a smaller share while private school tuitions for students with disabilities too severe for local districts to accommodate took a larger share. Finally, transportation increased from near zero to 6 percent of all special-education costs in 1991.
In other spending on special populations, about 8 percent of schools' net new money went to attendance, dropout prevention, alternative instruction, and counseling aimed at keeping youths in school. Programs to deter dropouts did not exist in 1967. In 1991, our sample districts operated academic programs, as well as provided support services, for pregnant and parenting teens; they operated alternative schools for students who worked or cared for siblings during regular school hours; and even operated separate classes for students suspended for disciplinary reasons. In 1967, schools had no commitment to any of these youngsters. About 7 percent of the new money went for bilingual and compensatory education. Contrary to conventional wisdom, bilingual students in 1991 did not have 1967 counterparts who "made it" in regular English classes. In 1967, the only bilingual program in our sample districts was operated in Fall River, Mass., in Portuguese. Since that time, not only has immigration increased, but non-English-speaking students increasingly come from families having limited literacy in any language.
Another 8 percent of the net new money went to expand school-lunch programs. In our poorest sample district (Claiborne County, Tenn.), 17 percent of all net new money went for lunches. In the urban districts, school-lunch spending also consumed more than 10 percent of net new funds. In 1970, 15 percent of American children came from poor families. By 1991, 21 percent did so. Thus, the overall growth in school-lunch spending may only attempt to compensate for the fact that many children came to school more poorly nourished in 1991 than in 1967.
In regular education, per-pupil spending increased by 28 percent during the quarter-century we studied. Both average teacher compensation and the number of teachers per student (staffing intensity) grew. Higher average teacher salaries were mainly due to teachers' greater experience (age) and training (for example, with master's degrees) in 1991 compared with 1967. In districts we studied, real salaries for teachers of similar experience and training did not significantly increase during this period and declined for many salary-schedule steps. Because teacher-salary levels did not significantly catch up with salaries of other college-educated professionals, districts may have had a difficult time attracting more highly qualified college graduates to the teaching profession. This apparent trade-off, between increasing absolute salary levels and paying more for experience and course-taking, may not have been the most cost-effective. Leading researchers have yet to demonstrate an empirical link between student achievement and either teachers' course-taking or teaching experience after the initial five years.
Growth in regular-education staffing intensity was more marked at the elementary than the secondary level, representing perhaps a response to research which suggests that investment in early grades of education can have positive returns. About half the reduction in elementary school pupil-teacher ratios reflects smaller class sizes. The other half reflects the addition of more subject specialists supporting increased planning time for classroom teachers and more resource teachers working outside the regular classroom.
In the urban districts we studied, per-pupil regular-education spending dropped most sharply as a percentage of all spending. In Los Angeles, the one urban megadistrict in our sample, regular education fell even further, from 87 percent of spending in 1967 to 51 percent in 1991. This share decline is nearly twice the average share decline of regular education in the remaining eight districts in our sample. This disparity, if typical, may help explain why there is such great concern for academic outcomes in large urban districts.
Thus, the majority of the spending increase has gone to support new services and increased instruction to special populations of students, some of whom may not have been in schools in the 1960s (such as severely handicapped children, immigrants, students who formerly dropped out of school, and so forth). While school districts control some of these resource-allocation decisions, state, federal, and court regulations often influence the assignment of staff and services to special populations. Thus, the changing allocation of the educational dollar results from a combination of district and outside initiatives. These data provide a foundation for re-examining the allocation of educational dollars and the priorities reflected by this distribution of resources.
The relatively modest growth of regular education also suggests that drastic pronouncements regarding the ineffectiveness of school spending to improve achievement may be overstated. If claims made by others of modest improvement in regular-education outcomes are accurate, then it may be that both regular-education resources and outcomes have grown modestly. Increased spending in the future may or may not produce further outcome gains, but based on the experience of the past quarter-century, the possibility cannot be easily dismissed.