Journal Issue: Financing Schools Volume 7 Number 3 Winter 1997
Fifty Years of Spending Increases
The National Center for Education Statistics (NCES) calculates that the average per-pupil expenditure (excluding capital expenditures), in constant 1993–94 dollars, has risen from $1,299 in 1949–50 to $5,734 in 1993–94.16 Figure 2 illustrates the changes in enrollment, number of teachers employed, average teacher salary, and cost per pupil from 1949–50 to 1993–94, with projections of enrollment to fall 2003. Here, three time periods are discussed separately: the baby boom (years of increasing enrollment) from 1949–50 to 1971–72, the baby bust (years of declining enrollment) from 1971–72 to 1984–85, and the baby boom echo (increasing enrollment) from 1985–86 onward.
While the baby boom is a well-known phenomenon, its magnitude is still impressive: in just 22 years, the nation's school population almost doubled (from 25.1 million to 46 million students, an increase of 83%), and the number of teachers more than doubled (from 913,000 to 2,016,000 teachers, an increase of 121%). Because teachers were hired at an even more rapid rate than the growth in enrollment, the student-teacher ratio dropped from 27.5:1 to 22.3:1, contributing to the increasing per-pupil cost.
Another major factor in the increasing per-pupil cost was the 86% real increase in teachers' salaries (from the equivalent of $18,580 in 1949–50 to $34,490 in fall 1971, both figures expressed in 1993–94 dollars). Taken together, the increase in salary levels and the dropping student-teacher ratio account for a major share of the 171% increase in per-pupil costs (from $1,299 to $3,517 in 1993-94 dollars) during the baby boom years.
During the baby bust years, student enrollment fell by 15%, but the number of teachers employed actually rose by 3%. Again, this meant that the student-teacher ratio fell further, to 18.3:1, contributing to rising per-pupil cost. However, offsetting the change in student-teacher ratio was a decline in teachers' salaries. During the high inflation, declining enrollment years of the 1970s, teachers' salaries failed to keep pace with inflation, and teachers saw their real (after-inflation) wages erode by 8.6% (from $34,490 in 1971–72 to $31,530 in 1983–84, expressed in 1993–94 dollars). Despite the wages lost by teachers during the baby bust years, the per-pupil cost rose by 30% (from $3,517 in 1971–72 to $4,561 in 1983–84, expressed in 1993–94 dollars).
This rise in per-pupil costs continues through the baby boom echo, with an additional 28% increase in per-pupil costs in the 10 years from 1983–84 to 1993–94. During the echo years, teachers' salaries made up for the ground they had lost to inflation, but not much more. Teachers' salaries rose by 14% during the 10 years from 1983–84 to 1993–94, but they still averaged only 4% higher in 1993–94 ($35,820) than in 1971–72 ($34,490). Student-teacher ratios continued to drop, to 17.4:1, as the number of teachers employed rose even faster in the echo years (rising by 17%) than the increase in student enrollment (which rose by 11%).
While this picture reflects long-term trends, it should be noted that schools have made progress in controlling cost increases in recent years. Between 1989–90 and 1993–94, per-pupil costs rose a barely perceptible one-half of 1% (from $5,715 to $5,741).
It is also important to keep in mind that, while the cost of education per pupil has risen, the American economy has also expanded significantly. Total spending, expressed as a percentage of gross domestic product (GDP), has followed enrollment closely. The peaks of spending—4.6% of GDP in 1971–72 and 4.1% in 1991–92—represent high periods of K–12 enrollments (46 million and 42 million students, respectively). Again, the pattern during the baby boom period differs from that in subsequent years, as illustrated in Figure 3.
During the baby boom years from 1949–50 to 1971–72, while the number of students increased by a phenomenal 83%, per-capita personal income also grew by 77% (from $8,873 to $15,739 in 1995 dollars),17 and spending per pupil, as noted earlier, increased by 171%. Although the increase in personal income shows very strong economic growth in this period, spending per pupil increased even faster, as illustrated by the increasing percentage of the GDP devoted to education (rising from 2.1% to 4.6% of GDP).
During the baby bust and baby boom echo, the story changes. In the 21 years from 1971–72 to 1991–92, per-capita income increased by 37% (from $15,739 to $21,620 in 1995 dollars), enrollment dipped and climbed again, for a total drop of 8.7%, but per-pupil expenditures (as shown in Figure 2) rose by 63%. In other words, a rising standard of living coupled with a temporary decrease in school population has afforded an opportunity to educate more children for more years and to spend increasing amounts of money on each child's education.What Did the Added Money Buy?
In brief, increases in school spending have enabled schools to employ more people with more advanced academic credentials providing a wider range of services. Also, compared with 50 years ago, teachers are much better paid, though most of the real increase in pay came before 1971–72. In addition, schools are serving more students with disabilities, awarding more high school diplomas, and providing more services to children ages three to six. Supplementary services are provided for students with lower incomes and with limited English proficiency. And court cases requiring more equitable school funding have led to greater total funding for schools.
Where feasible, the following discussion estimates costs of changes in schooling. However, it would be misleading simply to add these amounts together. Many categories of students are overlapping. For example, individual students may receive supplementary school services for their low-income status and also because they have a disability or have limited proficiency in English. These same students may receive free or reduced-price meals and fall into a category of schooling—such as preschool, kindergarten, or secondary—which was either unavailable or underutilized a half century ago. These overlapping conditions suggest that estimating a total amount of incremental school spending is risky.More and Better-Paid Employees
- More School Employees. Principally, education's added resources purchase more people and pay them more money. An explanation for the largest part of the added spending comes from the sustained nationwide enhancement of pupil-employee ratios which has taken place over the past 50 years. For example, whereas, in 1949–50 there was one school employee for every 19.3 pupils, that figure has now changed to one for every 9.1 students.18
However, a reader should not confuse such changing employee ratios as necessarily reducing average class size. There has been a class size reduction, or at least a favorable pupil-teacher ratio change, for the time period involved, from 27.5 to 17.4. This is a 37% reduction. In addition, schools increasingly employ large numbers of nonteaching personnel. These are primarily teacher aides and support staff (which include school secretaries, bus drivers, cooks, janitors, health and recreation staff, and "psychological personnel").18 Vedder's analyses suggest that 25% of U.S. education budgets are allocated for nonteaching personnel. Other industrial nations expend only 15% to 20% in a comparable manner.19
Interestingly, school-district administrative staffs, taken as a national average, have not grown and, in fact, appear to have shrunk relative to the number of other school employees. Between 1949–50 and 1980, district administrative staffs shrunk from 2.6% to 1.9% of all school personnel. In 1993, the figure was 1.7%.18
Compared with other large enterprises, a relatively small number of school administrators supervise a large number of staff members. An association of school administrators analyzed data from the U.S. Bureau of Labor Statistics and concluded that elementary and secondary schools employ 14.5 persons per executive, administrator, or manager, compared with 9.3 in the transportation industry, 8.4 in food products, 7.1 in manufacturing, 6.3 in construction, 5.4 in mining, and 3.6 in public administration.20
- More Experienced and Better-Educated Teachers. As discussed above, educators' salaries rose dramatically (86% after adjusting for inflation) from 1949–50 to 1971–72.21 Presumably, a major factor was the enormous demand for new teachers to serve the rapidly rising school population.
During the years of declining enrollments in the 1970s and early 1980s, it appears that few teachers lost their jobs (at least, the total number of teachers employed stayed constant), but their salaries lost ground to inflation as real salaries dropped 13.8% between their 1972–73 peak and their low point in 1980–81. Thereafter, annual increases exceeding the rate of inflation brought teachers by 1986–87 to parity with their salaries in 1972–73. Since 1986–87, teachers' salaries after inflation have risen a total of 2.3%.
How do the salaries paid to teachers compare with those paid to other employed persons? In 1949–50, professional educators' compensation was 3% higher than that of the average of all full-time employees in all industries. (Figures are adjusted to account for educators' typically shorter work year.) In other words, teachers were paid more, by a small amount, than the average of the entire U.S. salaried workforce. By 1992–93, they were paid 26% more than their fellow workers.22 In slightly more than 40 years, the comparative pay of professional educators had improved substantially.
When compared with other professionals who have comparable levels of schooling, educators are paid close to the average for women. In 1994, the mean salary for employed women with bachelor's degrees was $31,741 and for women with master's degrees it was $39,457,23 while the average teacher was paid $37, 868.21 Currently, 42% of teachers nationally have obtained a master's degree.24 Men with similar levels of education earn significantly more ($43,663 with bachelor's degrees and $53,500 with master's degrees).
Today's teacher is not completely comparable to a counterpart from an earlier era because she has more years of teaching experience and is more likely to have a master's degree, as illustrated in Figure 4.
- Productivity Increases and Class Size. Many industries have become more productive. Why has education not made similar leaps in productivity? Because productivity almost always means that fewer employees are needed to produce the same amount and quality of goods. In education, this would mean that a single teacher should be able to teach an increasing number of students each year, that is, that average class sizes would grow every year. Looking at "productivity" another way, a single teacher, if not teaching more students, should be able to induce higher achievement, on average, from her students.
Schooling is more labor intensive than most enterprises. Education has not proven amenable to the substitution of capital for labor which has taken place in agriculture, manufacturing, and communication. Indeed, over the 50 years examined here, schooling has become more, not less, labor intensive. However, even if school employee-pupil ratios had remained constant over the past five decades, it is likely that school costs would have increased because schools must compete in a labor pool with other employers for the talents of professionals. If these other employers offer higher remuneration, affordable through productivity gains, education must elevate pay to attract comparable talent.
Expansions in Schools' Functions
Why do schools today employ more personnel? Simply ascribing added costs to added personnel and their higher salaries may contribute to a misperception. These higher-paid professionals have often been posted to the payroll because of public demands for added services, even if they are not always classroom services. In a nutshell, today's schools are not the same institution known to our parents and grandparents.
- Expansions in the Number of Students Served. In the half century since the conclusion of World War II, America's schools have come to serve a far larger population. This is true principally because there has been a far larger population to serve. The nation's overall population has increased 80% during this period of time.5 However, population growth alone is not the explanation. Several categories of students now attend America's schools who were seldom a presence a half century ago. This impacts per-pupil costs because, for the most part, these students require additional, or more expensive, services than the population served in 1950.
Expanded Secondary School Attendance. High school graduation has come to be the norm for America. In the past half century, median years of schooling for the U.S. population have increased three and a half years, from 9.3 in 1950 to 12.9 in 1993.25 Slightly over six million more students are to be found in America's secondary schools (grades 9–12) today than a half century ago. Assuming the mean per-pupil secondary spending level to be $7,000, these added six million students result in approximately $42 billion in added expenses in today's dollars.
Part of this added cost is due to the fact that high school students are more expensive to educate than are elementary students. Their specialized classes have lower student-teacher ratios, and some of their courses (biology, vocational education) require specialized classrooms and equipment. Per-pupil expenditures in secondary school are typically 1.2 to 1.5 times the amount spent per elementary pupil. Using this rule of thumb, one can estimate that between one-sixth and one-third of the $42 billion increase is attributable to the fact that these are secondary rather than elementary students.
Students with Disabilities. Before World War II, schools in most states had little or no legal obligation to serve students with disabilities. By 1973, 45 states had passed some form of legislation encouraging, funding, or mandating the education of students with disabilities.26 Between 1971 and 1973, more than 30 federal court decisions made it clear that the equal rights clause guaranteed a free, appropriate public education without discrimination on the basis of disability. Congress, in response, enacted the Education for All Handicapped Children Act in 1976.27 This legislation has subsequently undergone revision, but it continues as a civil rights landmark. It has also resulted in significant increases in school costs.
About 12% of the pupil population takes part in special education.28 The most widely cited study concluded that students in special education cost, on average, 2.3 times as much to educate as other students.29 A slightly more conservative estimate comes from Ladd, who speculates that the amount spent to educate students in special education is closer to twice what is spent for students without disabilities.30
Extrapolating from these numbers, costs of special services to students with disabilities might be taken to add approximately $30 to $35 billion to the overall costs of today's schooling. However, this figure is too high. Some youngsters with disabilities were schooled before the 1976 legislation. Also, not all students classified as disabled today are receiving high-cost treatment. Because of the difficulty in disentangling such factors, it is impossible to determine the precise amount that instructing students with disabilities has added to the total education bill. However, the cost assuredly is large, and the federal government's fiscal appropriations just as assuredly do not cover such costs.31
Kindergarten Students. At the end of World War II, half-day kindergarten was an available option in only a few wealthy school districts. By 1965, almost half of America's five-year-olds were enrolled. Kindergarten, presently, is well on its way to being offered universally in America's public schools and extended to a full day, usually defined as about five hours of daily time in school. Kindergarten class sizes are typically smaller than other classes, leading to a higher per-pupil cost, although neither the federal government nor the states collect data on kindergarten per-pupil cost, separate from the cost of other classes. There is also no accounting convention for assigning non-classroom costs to kindergarten services. Moreover, even though full-time kindergarten is becoming the national norm, many school systems offer only half-time (three- to four-hour) kindergarten. These complexities render it impossible to estimate the added cost of this service accurately.
Preschool Students. Service for children three to four years of age, while by no means universal, has rapidly expanded. In 1965, some 10.6% of three- and four-year-olds were registered in preschool (including private, public, and religious preschools). By 1995, this number had grown to 48.7%.32 As with kindergarten, however, estimating the share of this cost in the public school budget is not possible.
- Expansions in Services Provided by Schools. There have been many systematic efforts to improve the quality of students' lives through schooling. The efforts discussed here include compensatory education for low-income students, nutrition, the impact of lengthening the school year, and racial desegregation of schools.
Compensatory Education: Title I. President Lyndon B. Johnson's mid-1960s war on poverty featured education as a central campaign theme. Foremost among the education bills enacted during this period was the 1965 Elementary and Secondary Education Act (ESEA) authorizing federal funds for supplementary education for students from low-income households. Today, ESEA costs approximately $6.8 billion per year. More than 20 states have followed suit with additional state appropriations.
Since enactment of the ESEA more than three decades ago, the number and percentage of children living in poverty have increased. School districts and child advocates claim that available federal funding is, therefore, insufficient to cover the growing need. However, regulations surrounding use of federal compensatory education funding now permit more efficient deployment of these resources. For example, local districts are encouraged to concentrate compensatory funds in schools with the highest concentrations of disadvantaged students so as to derive maximum impact. Also, schools are allowed to combine this federal funding with local resources, leading to the opportunity to use the money more effectively.
Nutrition. Since World War II, the United States has vastly increased government subsidies for the feeding of schoolchildren. Free and reduced-price lunches and breakfasts, which were virtually nonexistent 50 years ago, are now commonplace in America's public schools. The U.S. Department of Agriculture in 1995 spent $8.4 billion providing subsidies and surplus foodstuffs to public schools.33 (This figure is a component of the total federal expenditure for K–12 education.) Many local school districts supplement this amount.
Longer School Year. The NCES reports approximately a three-day extension of the mean U.S. public school year since 1949–50. Through an extrapolation of current daily costs, it is possible to calculate that this extension adds approximately $4.2 billion to the nation's annual school expenditures.
Racial Desegregation. America's immediate post-World War II schools were racially segregated. The U.S. Supreme Court's 1954 decision in Brown v. Board of Education34 began to dismantle the legal wall separating black and white students, and on many practical fronts the process continues still. But why does eliminating a dual school system cost added money? Did not the previous condition of "separate but equal" result in the same costs?
There is little systematic evidence regarding pre-Brown spending on black students. Court testimony and anecdotal historic references suggest that black schools and teachers were not treated equally. However, the 1966 U.S. federal report on Equality of Educational Opportunity,35 the so-called Coleman report, did not find significant resource differences in 1966 between black and white student populations within regions of the United States. It could be that the more egregious forms of financial discrimination had been alleviated by this time.
Regardless of such complexity, racial desegregation adds to school costs, at least through the added expense of transportation imposed. Also, steps had to be taken to ensure that the salary schedules for black and white teachers were uniform.
In many districts, efforts have been made to intensify instructional opportunity to compensate for past desegregation ills. Kansas City, Missouri, is perhaps the most dramatic of such compensatory efforts. Approximately $1.2 billion has been spent by this system over the past five years to provide new and better facilities, establish magnet schools, and enhance curricula offerings in an effort to elevate black student achievement and render urban public schools competitive with private and suburban alternatives.36,37
California also allocates more than $500 million annually in state aid for court-ordered and voluntary desegregation efforts, not including transportation costs.38 These funds are used in a wide variety of ways to assist students, for example, employing teacher aides, reducing class size, and purchasing computers.
Other. While this article lists some of the services added in the public schools during the past two generations, other added services are far from inconsequential. A partial list of these services would include special instruction for students with limited English proficiency, services to migrant students, asbestos removal, increased security measures in schools, and subsidies of special education and textbook materials in private schools.
Pursuing Equity Has Raised Total Spending
Since the mid 1960s, lawsuits have been filed in 43 states seeking a reduction in intrastate, district-to-district, per-pupil spending disparities. Such spending differences can be substantial: Texas and California once had conditions under which the highest-spending districts had per-pupil expenditures that were 10 times greater than those in the lowest-spending districts.
Plaintiffs have lost the majority of these finance reform cases. Even so, cases have been decided in favor of plaintiffs in at least 16 states, and legislatures have had to redraw school financing statutes as a consequence. (See the article by Augenblick, Myers, and Anderson in this journal issue.) Additional decisions are pending which could increase this number. These decisions have indirectly affected other states; often, the simple threat of litigation has prompted state legislatures to equalize funding.
Murray, Evans, and Schwab39 have undertaken a set of sophisticated analyses regarding the results of legislative equal protection remedies. They conclude that state efforts to improve funding equity have resulted in a 19% to 37% reduction of district-to-district funding inequality, few spending reductions for students in high-spending districts, an average 12% increase in per-pupil spending in the low-spending districts, an average 8% increase in per-pupil spending in the school districts spending at the state's median, and few reductions in state spending for other activities. This last feature suggests that elevated school spending comes not from reallocating existing resources but from expanding the pie of total resources through higher taxation. By this reasoning, the efforts of states to assure equitable spending across school districts have contributed approximately $16 billion to the nation's annual school spending total.40Increases Not Accounted For
Although the discussion above covers many important, legitimate reasons for increasing costs, these factors do not fully explain the fourfold increase in per-pupil expenditures since 1950. It appears that unmeasured, possibly political, factors also are at work. What follows is a brief consideration of two opposing theories.
The first of these theories is that schools have improved in their performance: Increased numbers of school employees and higher pay have resulted in higher student achievement spread over a wider base of students and, thus, the voting public approves and has been willing to pay for it.
Berliner and Biddle41 and Bracey42 all contend that current criticisms regarding school performance are politically motivated and incorrect in their assertions that schools are performing poorly.43 When compared to a half century ago, higher proportions of students are in poverty or have not grown up with English as a native language. Yet more individuals now attend school, stay longer in school, graduate from school, sit for college entrance examinations, attend college, and graduate from college than ever before.
Rothstein and colleagues44 also point to the above described costly education system alterations and additions, rely upon a labor-friendly set of inflationary assumptions to adjust spending figures, acknowledge the panoply of expectations imposed on education by the political system, and conclude that increases in spending for the regular education program have been modest.
The second, and opposing, theory is that an underlying political dynamic, favoring the interests of school employees over those of students and taxpayers, facilitates an increase in school costs regardless of pupil achievement and school performance. Proponents of this analysis note first that public school budgets are products of political processes. Clearly, pupils, parents, and the public generally benefit from having public schools. However, the principal or immediate beneficiaries of increased school spending are professional educators. School spending means more and higher-paying jobs.
Teachers and other professional educators have a more intense incentive than taxpayers to engage in collective political activity, spend money on lobbyists, run for office, serve as delegates to nominating conventions, or personally vote for higher financing for schools. Added dollars for schools will cost each individual taxpayer only a marginal amount. Conversely, added spending for schools will likely benefit professional educators a great deal more.45
This latter perspective motivates free-market-oriented education reformers to argue that the political playing field can be leveled only through reforms such as vouchers. In this way, contend free-market proponents, a better balance would be struck between producers of education services and their clients (parents, pupils, and the public). See the article by Koppich in this journal issue for further discussion of vouchers.
Those who contend that America's schools display unacceptably low productivity point to Consumer Price Index-adjusted figures, view spending increases through the lens of academic achievement, point to improved demographic characteristics (for example, parents' education levels have increased and family size has declined, which some argue counterbalances the impact of higher poverty rates), and conclude that sweeping reforms are justified. For example, Hanushek46 concludes that, after controlling for all that is reasonable, per-pupil school spending has increased 3% annually since the beginning of the twentieth century. Odden, noting the rise in per-pupil spending, contends that improved productivity is definitely in order.47
These two, and dozens of other analysts, may advocate different reform strategies. What they share, however, is a contention that, if the goal is enhancing student achievement and gaining greater productivity, then the structural arrangements of American education must be altered.