Myth: Doubling the money spent on public education hasn't improved it.

Fact: Public education has improved, even though the increased spending has gone to other vital services, not general education.


The U.S. has doubled the amount of money spent per pupil between 1965 and 1993, but the money has not gone to direct improvements for general education. Instead it has gone to special education for disabled students, school lunches, transportation and dropout prevention, all of which are indirect education benefits. Increased spending on direct benefits, like smaller classroom size and larger teacher salaries, have been insufficient to produce the desired results. Even so, the last 30 years have seen some general benefits, such as a lower dropout rate and increasingly better SAT scores for minorities.


Irving Kristol summarized the conservative position on public education thus: "Look at the spending on public schools. It goes up and up, and the results go down and down and down." (1)

It is true that between 1965 and 1990, the U.S. more than doubled the amount of money it spent on each student in public schools, from $2,611 to $5,521 in constant 1990 dollars. But it is not true that the results have been a failure. Between 1970 and 1993, the high-school dropout rate fell from 12.2 to 9.2 percent. (2) And SAT test scores have been rising for whites and minorities alike over the last 30 years. Paradoxically, the average SAT score has fallen, because more minorities and lower class whites have been taking the test, and they tend to score lower than advantaged whites. But this is a statistical fluke, masking the advances made by everybody.

Still, some critics would argue that these meager results are insufficient, considering that we've doubled the amount of money spent on each student. But in fact the money has gone not to general education, but to other vital or supporting services. Of all the increased spending since 1965, here are the programs that benefited:

Percent of increased spending on public education devoted to services
other than general education, 1965-1993

Smaller classes      33 percent (approximately)
Special education    30
School lunches       10
Teacher raises        8
Transportation        5
Dropout prevention    3

These programs account for well over 80 percent of all new educational spending since 1965. It's worth examining each one briefly, to show how they do not directly result in higher academic scores (and may even lower them):

Smaller classes: Between 1965 and 1993, the student/teacher ratio declined about 30 percent, and the average class size fell to 24 students. This effort has required building more classrooms and hiring more teachers. Although this has resulted in better classroom discipline and management, experts believe that class size must fall to about 15 students per teacher to achieve the kind of individualized instruction that truly benefits students.

Special education: Since 1975, federal law has required a "free appropriate education" to any child with a disability, no matter how serious. The Supreme Court has also ruled that cost is no object in providing such an education. By 1990, nearly 12 percent of all students were enrolled in special education. Obviously, providing disabled children with an education is praiseworthy, but it is expensive, and it would do nothing to raise the SAT scores of able-bodied students bound for college.

School lunches: Providing children with nutrition is an indirect, not direct, academic benefit. Originally it was intended to keep students healthy, avoid the distraction of hunger, and provide an incentive for poor children to attend school. However, these benefits have been counterbalanced by an alarming rise in child poverty since 1965. Today, nearly a quarter of all children live in poverty.

Teacher raises: Teacher salaries have grown less than 1 percent a year between 1965 and 1990, from $27,221 to $32,977 in constant 1990 dollars. In theory, raising teacher salaries should attract more competent teachers. However, the growth in teacher salaries has been less than the growth in other professional fields requiring a college degree, which draws the best talent away from teaching. Furthermore, women have greatly expanded their career choices since 1965, and are no longer bound to the teaching profession. Not surprisingly, public education has seen a drain of its best female teachers.

Transportation: Between 1965 and 1989, the share of students who took a bus to school rose from 40 to 59 percent. The cost per student also rose, from $214 to $390 in constant 1990 dollars. Ironically, this program would reduce the average SAT score, since upper class children (who normally score higher) already have reliable transportation to school. Busing allows more lower class children to attend and complete school -- which results in lower average scores.

Dropout prevention: Spending more money on truant officers and dropout prevention programs also reduces the average SAT score, since it encourages many poor students to remain in school who would have otherwise dropped out. The very success of this program ironically gives conservatives an argument that spending more money only results in falling test scores.

In conclusion, most of the increased spending on public education in the last 30 years has been for programs that only indirectly affect educational outcomes at best. (And in the direct cases, like smaller classroom size and larger teacher salaries, the funding has been insufficient.) Liberals do not claim that spending of this type is intended to improve SAT scores, and conservatives are mistaken in assuming that they do.

Return to Overview


1. Unless otherwise indicated, all facts and quotes in this essay come from Richard Rothstein, "The Myth of Public School Failure," The American Prospect, no. 13, Spring, 1993.

2. U.S. Bureau of the Census, Current Population Reports, P20-479 and earlier reports. A dropout is defined as a person aged 14 to 24 who is not in regular school and has not completed the 12th grade nor received a general equivalency degree.