[An updated version of this article can be found at Education in the 2nd edition.]
By most accounts America's schools are not performing very well. The average combined (verbal and mathematics) score on the Scholastic Aptitude Test was seventy-five points higher in 1963 than it was in 1990. A high school senior ranked at the 50th percentile on the SAT in 1990 would have ranked around the 33rd percentile in 1963. American students trail most of their international counterparts in mathematics and science achievement. Recent comparisons of industrialized countries place the United States between tenth and fifteenth in these economically vital fields. And excellence is not all that is missing. Performance among schools is quite inconsistent. Blacks score nearly two hundred points below whites on the SAT. Urban high schools, which serve disproportionately large numbers of poor families, fail to graduate nearly half of their students; high schools nationwide graduate about four-fifths of theirs.
To make matters worse, performance has stalled or fallen despite aggressive government efforts to turn it around. Since the launch of the Soviet Sputnik more than thirty years ago, school reform has been an ongoing enterprise. President Johnson signed the Elementary and Secondary Education Act, the cornerstone of his Great Society program and the beginning of aggressive federal efforts to upgrade the schools of children living in poverty. President Carter formed the Department of Education in order to raise the political profile of federal education policy. President Reagan's National Commission on Excellence in Education spurred nearly a decade of ambitious school reform when it warned, in a landmark report of the same name, that the United States is "A Nation at Risk." Most recently, President Bush introduced "America 2000," a comprehensive reform plan that, if implemented completely, would create an entirely new system of education.
All of this political attention has brought many changes. Annual inflation-adjusted expenditures per pupil have tripled since 1960, surpassing $5,500 in 1991. Over the same period class sizes were reduced by about 30 percent. Teachers have become much more experienced—fifteen years on average now versus eight years in 1971—and have acquired more formal education. Only a quarter of the nation's teachers held master's degrees in 1971, while more than half do today. Since 1980, virtually every state in the union has raised its high school graduation requirements, and students are now taking more academic courses than they did a decade ago.
To be sure, there have been some signs of improved performance. The average SAT score fell by ninety points between 1963 and 1980, but then rebounded by fifteen points during the early and mideighties. Black and other minority students improved their SAT scores during the seventies and eighties, slightly narrowing the chasm between their achievement and that of whites. During the eighties the percentage of students scoring at grade-appropriate levels on the National Assessment of Educational Progress—a test that is a better gauge of in-school learning than the SAT—was up several points. The dropout rate has improved by roughly 10 percentage points since the early seventies—if General Equivalency Diplomas are counted. Yet, as the National Assessment of Educational Progress reveals, most gains appear to have come in rote learning, computation, and basic skills, rather than in problem-solving and other higher-order skills. As of 1991, moreover, the nation's average SAT score had lost another ten points. By any standard, recent improvements have been exceedingly modest.
Why, then, despite great effort, are America's schools doing rather poorly? One possibility is that students have become more difficult and families less supportive. Yet scores of careful analyses have found little support for this hypothesis. The facts are these: drug use among schoolchildren has plummeted over the last decade; childhood poverty rates, though up and down, are the same today as in the late sixties; and student achievement began to decline in the early sixties—about ten years before divorce rates and female employment began their rapid ascents.
Hundreds of studies have also examined the relationship between major mainstream school reforms and student achievement, allowing for differences in the characteristics of students and families across schools and over time. The overwhelming majority of these studies have found the same thing. School performance is simply unrelated to conventional school "improvements," including higher expenditures per pupil, better educated or more experienced teachers, and smaller pupil—teacher ratios. School performance appears to be eroding or stagnating not because school problems are tougher (though this may be the explanation in some schools) but because the school "improvements" made by reformers do not have much influence on performance.
What, then, does influence school performance, and why haven't school reformers done the right things? Unfortunately, there is no simple recipe for school success. The "inputs" necessary to produce desirable educational "outputs" are not well understood, and this puts school reformers in a serious bind. Fortunately, some reform strategies do not depend so heavily on knowledge of which inputs really work. Rather, they rely on accountability for outputs (i.e., what students actually learn), incentives for performance, and ultimately, the market principles of school competition and parental choice. These strategies have given some reformers fresh hope of finally doing the right thing.
Experts increasingly agree that the qualities that distinguish America's best schools are very difficult to mandate. "Effective schools," as the best institutions are now often called, are distinguished by such attributes as a clear sense of purpose, strong leadership by principals toward shared educational goals, professionalism and teamwork on the part of teachers, and high academic expectations for all students. These attributes, which fall into the realm of attitudes and behavior, are beyond the direct control of school reformers. Educators can be told that they must work cooperatively, enthusiastically, creatively, and assiduously, but no one can guarantee that they will.
An alternative strategy that I and many others believe makes good educational as well as economic sense is for reformers to cease telling educators what to do and how to do it, and to start telling them what society expects from them as a finished product. This strategy would allow educators to exercise their professional judgment—in assembling a talented and dedicated teaching force, in selecting and designing interesting texts and curricular materials, in tailoring instructional techniques to fit the needs of the students and families being served, and in orienting the school around a theme or mission that captures the imagination of teachers and students alike. In exchange for such autonomy, the schools would be held accountable for results. Schools that use their autonomy productively would be rewarded. Nonproductive schools might even be punished.
Gradually, many educators and researchers are concluding that autonomy and accountability are the way to go. Various scholars and most of the organizations representing teachers and administrators claim that the system of public education has become too centralized and bureaucratized, and that the quality of schools has consequently suffered. As Terry M. Moe and I argue in Politics, Markets, and America's Schools, excessive regulation undermines the professionalism and vitality of teachers and principals, leading many good people to leave the schools. Those who remain tend to organize and lobby for regulatory protection from the growing number of authorities above them. Political conflict increases, and despite lots of good intentions, the system grows more bureaucratic, less manageable, and less successful.
The fact is, as school performance has eroded over the last several decades, bureaucracy has grown larger. Instructional expenditures account for a rapidly declining share—now less than 60 percent—of local school expenditures. Full-time classroom teachers account for less than half of local school employment, while administrators represent about 15 percent. The number of nonclassroom personnel is growing at seven times the rate of classroom teachers.
The system has also become more centralized, with the states surpassing local governments as the major source of school funds. An average state today pays for 50 percent of public elementary and secondary education; an average school district pays for 45 percent. In 1960, school districts averaged 60 percent, states only 35 percent. Although the federal share of school funding has actually fallen 3 percentage points since 1980, the number of special programs that the federal government implements has remained constant at roughly eighty, and the volume of regulation it promulgates has grown.
These developments have come under increasing attack. Numerous reformers are now calling for radical reductions in bureaucratic control. The chancellor of the New York City Public Schools, Joseph Fernandez, is the nation's leading advocate of a form of decentralization known as "school-based management," a strategy that he pioneered in Dade County, Florida. The Chicago Public School System was dramatically decentralized in 1989 by shifting much of the control over schools from a central board of education to hundreds of boards, each responsible for a single school. All across America, school-based management and variations on that theme have become the most popular reform strategies of the nineties.
It is too early to judge these recent reforms, and not surprisingly, the academic gains produced by them so far have been small or nonexistent. Indeed, there is good reason to believe that decentralization alone will not promote major improvement. The difficulty is that while decentralization attacks the problem of bureaucracy and affords schools professional freedom, it does not necessarily provide society with accountability for results. Unless there is some means for society to express its interest in schooling—to specify what results it wants—a decentralized system of schools may not produce the educational improvements society values.
Reformers are aware of the need for accountability. But except in very small public school systems, reformers have been unable to devise accountability mechanisms that do not entail yet another layer of bureaucracy. Most mechanisms involve much formality and detail about such indicators of performance as test scores, dropout rates, teacher attendance records, personnel evaluations, and so on, along with many specifications of the conditions under which schools may be rewarded or punished for changes in these indicators.
The dangers in such new bureaucracy are many. Schools may produce formal improvements on the indicators, but little genuine improvement in education. For example, schools may "baby-sit" students in order to reduce dropout rates, or teach students to take standardized tests rather than to think. The new bureaucracy may also interfere as much as the old bureaucracy with the legitimate needs of educators for professional autonomy. It is little wonder that one of the school reforms that is most popular with the general public—merit pay for teachers—has yet to be meaningfully implemented anywhere in the country.
How, then, do we get an autonomous school system that is accountable to the public? Some economists believe the most promising strategy is to reorganize the public education system according to the principles of competition and choice. The basic idea is to permit parents and students to choose their schools. Good schools, ones that provide the kinds of academic results that most parents want, would be rewarded and would flourish. Bad schools, schools that few want to attend, would not be propped up by interest groups that are able to influence the political process. They would be denied resources, go out of business, and be replaced. Opponents of this view argue, among other things, that it would further disadvantage poor students, whose parents are not as well equipped as wealthy parents to make informed choices.
For me and many others, however, the logic of educational choice holds great appeal. But there is more to recommend educational choice than its "survival-of-the-fittest" rationale. In particular, educational choice would create powerful disincentives for bureaucratic growth and powerful incentives for empowerment of professionals at the school level. In a marketplace where parent and student satisfaction are crucial to a school's longevity, schools must respond to diverse family wants and needs. This can be accomplished only if the people closest to parents and students—namely, teachers and principals—have ample authority.
This is not just a matter of theoretical speculation. The New York City public schools, for example, employ more than 6,000 central office personnel—an administrator/student ratio of 1 to 150. The Catholic schools of the Archdiocese of New York, a smaller system than its public counterpart but nevertheless the twelfth largest school system in the country, employs only 30 central office personnel—a ratio of 1 to 4,000. Political control of schools encourages bureaucratization; market control dramatically discourages it. If reform aspires to create a school system that is based not only on accountability for results but on school autonomy and the professionalism of educators, a system organized around the principles of competition and choice is doubly desirable.
Of course, no market is perfect. For a system of educational choice to work efficiently and equitably, the government would need to play a significant role. Experts differ over precisely how this role would be played. But most agree that the government would need to fund the system and provide students from poor families the financial wherewithal to compete effectively with students from middle-class families for admission to schools of their choice. The government would need to ensure that all parents have ample and accurate information about how schools are performing so that students are not excluded from schools because of parental ignorance. The government might do this by operating parent information centers, mandating achievement tests, or like the Federal Trade Commission, regulating "truth in advertising" by schools. The government might also find private firms like Consumers Union offering all sorts of comparative assessments of schools, thereby reducing the need for government information.
The government would also need to design an admissions process that guarantees all students a school. Schools and students should be permitted to match up voluntarily as much as possible, but there must also be a safety net. Schools might be required, for example, to accept some mandatory placements from a lottery of students who do not find schools voluntarily. The government might also want to establish basic chartering criteria to guarantee that schools that participate in the system are nondiscriminatory, healthy, safe, and bona fide institutions of learning—not degree mills. If a system of educational choice is to make a real difference in the supply of schools, however, the government must also recognize the limits of its role and the need for market forces to operate. The government must trust and respect the professional judgment of teachers and principals, and the values, concerns, and intelligence of parents.
Increasingly, governments are doing just that. Most cities have introduced specialized "magnet" schools to motivate children and to bring together children of different races voluntarily. Several cities have fully "magnetized" their school systems, with encouraging results. Eight states now provide freedom of choice among most of their public schools.
A number of cities and towns also provide vouchers for students to attend private schools or colleges. Since 1986 Minnesota has had more than ten thousand juniors and seniors using vouchers to attend college instead of high school—a program that not only is popular with participants but that also has stimulated large increases in advanced placement classes in high schools threatened with college competition. Since 1990 Wisconsin has permitted low-income parents in Milwaukee to choose nonsectarian private schools. Despite court battles, opposition from the state and local education agencies, and great uncertainty about the program's future, roughly seven hundred families are happily (according to independent surveys) participating, and parents are much more involved in their children's education than they were before.
Of course, not all experiences with choice—public or private—are positive. Because public schools are often forced by government to be the same, parents have sometimes met public school "choice" with indifference. Because markets take time to weed out inferior products, the Milwaukee voucher program initially included one educationally dubious school that served its students poorly—before it closed and returned its students to the public schools. Nevertheless, choice was the keystone of the education strategy offered by former President Bush, and is the focus of school reform debates nationwide.
Public opinion polls indicate that the concept of educational choice is supported by a majority of Americans, especially poor Americans and racial minorities who are often trapped—without choice—in collapsing urban school systems. The business community is panicked about the quality of the work force and has grown impatient with traditional school reforms. As educators come to see that there is little hope for acquiring autonomy without also providing the accountability that choice allows, educators, too, may become supporters of the idea. In any event, choice will be the focus of educational debate over the next decade.
John E. Chubb is a distinguished visiting fellow at the Hoover Institution and executive vice president of Edison Schools.
Boaz, David, ed. Liberating Schools: Education in the Inner City. 1991.
Chubb, John E., and Terry M. Moe. Politics, Markets, and America's Schools. 1990.
Coleman, James S., and Thomas Hoffer. Public and Private High Schools: The Impact of Communities. 1987.
Finn, Chester E., Jr. We Must Take Charge: Our Schools and Our Future. 1991.
Sizer, Theodore R. Horace's Compromise: The Dilemma of the American High School. 1985.
Related Material on Econlib:
EconTalk podcasts: Eric Hanushek on Educational Quality and Economic Growth and Making Schools Better: A Conversation with Rick Hanushek. Hosted by Russ Roberts.