High school size

High school size

High School Size and the Education of All Students in 9-12: What the Research Suggests

by Sandra Stotsky

For reasons that go beyond rational thinking, the size of American high schools has suddenly become a major educational issue. On the basis of size alone, it seems, American high schools have been declared obsolete and dysfunctional for all students. What is strikingly absent from these declarations, often by people who have never taught at the high school level, is evidence. There is no evidence that size is a systemic problem independent of the student body in a high school—or that the difficulty many students have in doing high school level work is a function of the high school curriculum.

Many large urban high schools with a generally low achieving student body and a high drop-out rate are dysfunctional. But some large urban high schools have a high-achieving student body and almost no drop-outs. In 2004-2005, examination schools in New York City, for example, ranged from Bronx High School of Science with 2617 students and Stuyvesant High School with 3059 students to Brooklyn Technical High School with 4062 students, with similar numbers at other very high performing (but not examination) high schools, such as Benjamin Cardozo High School with 3972 students and James Madison High School with 3978 students. New York City parents clearly do not think these large high schools are dysfunctional; this past spring almost 30, 000 students took the entrance test for the fewer than 8,000 available seats in the examination schools. Moreover, according to a New York Times article on November 18, 2005, Mayor Michael Bloomberg is now proposing to build more examination high schools in New York City, among other kinds of schools; at present the mayor's plan includes seven new selective high schools, including one to be called Brooklyn Latin and another to be a math and science school affiliated with Columbia University. So far the mayor has not specified that they must be tiny. Clearly, large high schools may or may not be dysfunctional.

Our high schools enroll students with different interests, abilities, and learning paces. Public priorities should ensure that all high schools, large or small, offer all students coursework based on demanding academic standards and taught by academically qualified teachers. Except for students who need small, therapeutic environments for learning, size should matter only when a high school is too small to offer a full curriculum for all its students, as are many rural high schools, or too large for community feeling and effective administration, as are many of the urban high schools that enroll large numbers of students with poor reading, writing, and mathematics skills and little interest in academic coursework. Although we are regularly told by a group I call Educators Without Evidence, an informal organization with virtually monopolistic access to the media, that high schools should enroll no more than 300 to 400 students, the empirical evidence so far suggests otherwise.

I. “Ideal” Size in Research Studies
Is there an “ideal size for a high school? We find an interesting convergence of evidence from several independent sources and periods of time. One source, a study reported in 1997 by Valerie Lee of the University of Michigan and Julia Smith of Western Michigan University, was highlighted by Diane Ravitch in an op-ed in the Washington Post on November 6, 2005. After analyzing progress in mathematics and reading from 8th to 12th grade for 10,000 students in a federal data base from almost 789 public and private schools of varying size, the two researchers concluded that “the ideal high school” enrolls between 600 and 900 students. Size matters, they believe, because it affects social relations within the school and the school’s ability to provide a strong curriculum for all students. Very large schools lack a sense of community and cannot shape student behavior, while very small schools cannot offer a full academic curriculum. In the Lee and Smith study, low-income students made the greatest academic gains in schools of 600-900 students. Academic gains for both low-income and high-income students declined in schools enrolling fewer than 600 students, and declined again in schools enrolling fewer than 300.

Similar numbers for the “ideal” size of a high school—and similar trends—turned up in a 2002 study conducted by a group of researchers at the National Foundation for Educational Research in England. Controlling for background variables and using a secondary national value-added data set from almost 3000 high schools throughout England, the researchers analyzed the relationship between student achievement and single-sex education as well as size. They found that student outcomes improved with size up to a certain point and then declined. Best results were obtained in medium-sized schools with a cohort of about 180 to 200 students per grade, and the worst in the very small or very large schools. Boys and girls also did better in single-sex schools, especially girls in single-sex comprehensive schools.

Two studies that are still useful for details on what is desirable in a high school curriculum were published almost a half century ago but, interestingly, come up with a similar number. For two landmark reports, The American High School Today, published in 1957, and The Comprehensive High School, published in 1967, James B. Conant used a long list of specific criteria for judging the adequacy of a high school’s curriculum at a time when test data on student outcomes were unavailable. Criteria ranged from instruction in calculus, the offering of four years of one modern foreign language, and instruction in music and art to courses for slow-learners. Based on visits to or surveys of 2000 schools, Conant concluded that “an excellent comprehensive high school can be developed in any school district provided the high school enrolls at least 750 students and sufficient funds are available” (1967, p. 2). At the time, Conant was Chairman of a Committee of the National Association of Secondary School Principals on a Study of the American Secondary School, and he and his committee wanted to address the full range of student abilities and interests in our public high schools, not just those at risk or those who lacked the opportunity to take Advanced Placement coursework.

Another very recent empirical study deserves mention because of its comprehensiveness, even though it didn’t come up with an “ideal” size. Michael Hicks and Viktoriya Rusalkina of Marshall University conducted a study of the relationship between school consolidation and student achievement for the West Virginia School Building Authority in 2004. Their analysis of all West Virginia high schools, consolidated and single, found a tendency for higher achievement in larger, not smaller, schools. In fact, they found that both teacher education and larger schools “correlated with higher test scores among certain groups” (p. 30).

Finally, the results of a survey on school size reported by Public Agenda in March 2002 should be noted. According to Public Agenda, the survey results suggest that, for now, “neither teachers nor parents see reducing school size as a priority.”

II. Benefits of a Larger Size for Different Groups of Students
A larger size is better for all groups of students but for different reasons. As Conant suggested in 1957 and 1967, academically motivated students benefit from high schools that can offer honors and Advanced Placement courses and a four-year sequence in at least two foreign languages. Hicks’s comments, quoted in the Charleston Gazette in 2004, echo that suggestion: “Students in larger high schools are more likely to take Advanced Placement courses and college entrance exams.”

Larger high schools can also benefit students with learning disabilities or underachieving students who are potential drop-outs by providing the intensive help with reading and mathematics these students need, as well as a choice of career-oriented curricula and academic coursework at their level of skill. This is what an award-winning vocational/technical high school in Massachusetts does. It enrolls over 900 students, a large number of whom are special education students or are well below grade level in reading and mathematics for other reasons; it has almost a 100% pass rate on the state’s grade 10 high stakes tests and an attrition rate of less than 1%; and it sends about half of its graduates on to some form of post-secondary education (see the range of offerings of the Blackstone Valley Regional Vocational Technical High School at www.valleytech.k12.ma.us).

Finally, larger high schools may provide more challenging academic coursework than tiny high schools for the broad middle range of students who now graduate from high school but learn much less than they can. About 50% of these students now take remedial courses in reading and mathematics at the post-secondary level. Small high schools tend to offer a one-size-fits-all curriculum, which can make sense only if students have chosen to attend these high schools.

III. Features of useful research on high school size
Educators and policy makers who believe that wise policy making can be informed by what the evidence suggests on high school size should look to see whether a study is designed to provide the following information:

1. True comparison groups. The district average is not a true comparison score for small high schools broken out of a big one if assignment is not random. Charter or pilot schools are also not true comparison groups for a small high school broken out of a big one because they have self-selected populations. If selection is random from an application pool, a true comparison group is the group of students who also applied to the charter or pilot school but didn’t get in.

2. Breakdown of results by type of student. Average scores should be provided for those below grade level in reading and math and apt to drop out, those on or about grade level, and those well above grade level.

3. Detailed information on teachers’ academic qualifications for the subjects they teach.

4. Detailed information on the content and level of difficulty of the courses.

5. Detailed information on whether students who take Advanced Placement courses must take the Advanced Placement test and pass with a 3, 4, or 5.

6. Pre-post data on reading gains using district or state tests.

7. Use of measures of effectiveness other than retention and graduation rates to gauge the effectiveness of the small high school for the grade 9 students who are unlikely to drop out and who are likely to graduate. Overall, about 70% of our students do graduate.

IV. Ideas for Structural/Curricular Reforms
Perhaps the best way to harness the energies of our educators in a positive direction is to provide monetary incentives to a school for regularly increasing the percentage of students passing a rigorous end-of-course test for algebra I in grade 8 (designed by mathematicians). Why Algebra I in grade 8? Because it is the gateway course for high school mathematics and science. We need to revise the negative system of incentives the federal government and state governments have inadvertently created by massively funding remediation efforts to help schools get their low-performing students up to a minimal level of competency, but giving nothing much except paper awards to schools for increasing the numbers of those doing well. Schools with a large number of low-achieving students need extra funds, but we may accomplish much more over the long-term by rewarding schools that show a regular increase in the percentage of high performers with even more funding, putting positive incentives in place.

To enable schools to take advantage of positive incentives, they should be encouraged to be flexible in ways they group students and in the materials they use and to give teachers a greater voice in these matters. Concern about the lack of significant growth in the percent of grade 8 students performing at the two highest levels on the state’s grade 8 mathematics tests led the Massachusetts Department of Education to fund a study that gathered information from a random sample of administrators and teachers throughout the state in 2003. A significantly higher number of teachers in schools that both increased the percent of grade 8 students performing at the two highest performance levels and simultaneously decreased the percent of grade 8 students performing at the lowest performance level reported having a voice in choosing their instructional materials and using accelerated and leveled algebra I classes to address the needs of above grade students (http://nonpartisaneducation.org/Review/Articles/v1n1.htm).

Second, we need a strong discipline-based academic curriculum for grades 5-8, with each subject taught by a content specialist. The KIPP (Knowledge Is Power Program) schools are outstanding models of what is needed at the middle school level. According to a recent evaluation, these schools produce a higher level of achievement than demographically similar schools.

Structurally, the middle grades might be attached to a high school but in a separate wing, or as part of small neighborhood K-8 elementary schools, some of which might be single sex. But however the physical structures are configured, the curriculum in grades 5-8 should be clearly connected to a discipline-based 9-12 curriculum, not to the self-contained inter-disciplinary curriculum in K-4 classrooms. A middle school curriculum that is little more than a glorified elementary curriculum is what should be declared obsolete.

Finally, grade 9 should provide an intensive, transition-year program to students with serious limitations in reading, writing, and mathematics. In addition, all graduating grade 8 students should be able to choose the academic or technical/vocational school program they wish to pursue in high school, as well as to attend either a co-ed high school or a single-sex high school. Boys’ schools should be staffed chiefly by male teachers. No matter what curriculum a student pursues in grades 9 to 12, however, all students should be required to study the four core subjects of English, history/geography/U.S. government, science, and mathematics and take end-of-course tests in these courses. All high schools should also provide leveled courses in each core subject in one time block to permit acceleration or remediation (as does Blackstone Valley Regional Vocational Technical High School), with summer school courses available to facilitate acceleration or remediation. Such requirements will allow students to transfer from one high school to another, if they so choose, and to qualify for post-secondary study no matter what high school they have attended.


Conant, James. B. The American High School Today (a first report to interested citizens). New York: McGraw-Hill, 1957.

Conant, James B. The Comprehensive High School (a second report to interested citizens). New York: McGraw-Hill, 1967.

Educational Policy Institute. Focus on results: an academic impact analysis of the Knowledge Is Power Program (KIPP). A research paper prepared for the KIPP Foundation. August 2005.

Hicks, Michael J. & Rusalkina, Viktoriya. School Consolidation and Educational Performance. A monograph prepared for the West Virginia School Building Authority. Huntington, West Virginia: Center for Business and Economic Research at Marshall University, May 2004.

Lee, Valerie E. & Smith, J.B. “High School Size: Which Works Best and for Whom?” Educational Evaluation and Policy Analysis, 1997, Fall, 19 (3), 205-227.

Lee. Valerie E. “Effects of High-School Size on Student Outcomes: Response to Howley and Howley.” Education Policy Analysis Archives, Vol. 12, No. 53, September 24, 2004. “…equity without excellence is not something we should encourage in schools” (p. 4).

Ravitch, D. “Downsize High Schools? Not Too Far.” Washington Post, Sunday, November 6, 2005, page B07.

Spielhofer, T., O’Donnell, L., Benton, T., Schagen, S. & Schagen, I. The Impact of School Size and Single-Sex Education on Performance. Local Government Association Research Report 33. Slough, England: National Foundation for Educational Research, July 2002.

Stotsky, S., Bradley, R, & Warren, E. “School-Related Influences on Grade 8 Mathematics Performance in Massachusetts.” Nonpartisan Education Review, March 24, 2005, 1(1). Retrieved 10/5/05 from http://nonpartisaneducation.org/Review/Articles/v1n1.pdf

Public Agenda. Sizing Things Up: what Parents, Teachers, and Students think about Large and Small High Schools. March 2002.

Nonpartisan Education Review HOME