ResearchEducation & Skills

All-boys schools: What do we know?

Jul 24, 2025
Ben Smith
A group of boys at an all-boys school

Why revisit single-sex schooling?

Across the United States boys now trail girls on almost every rung of the education ladder—from kindergarten readiness through college completion. The gap is widest for Black boys and Latino boys and for those growing up in low-income districts. In the mid-2000s, the U.S. Department of Education loosened Title IX regulations to permit voluntary, gender-separate school options.

Two decades later it is time to ask: Has the experiment paid off?

AIBM commissioned a report from Nina Hankins, a recent graduate of the Goldman School of Public Policy, on the state of single-sex schooling that combined a systematic review of largely post-2014 research on single-sex schooling with fresh analysis of enrollment trends; semi-structured interviews with fourteen practitioners, researchers, and organizational leaders focused on single-sex schooling; brief case studies of high-performing single-sex schools; and a scan of the current Title IX policy landscape. Here we present our summary of its findings. See the full report here.

Key takeaways

  • After a 2006 Title IX rule change allowed single-sex public schools, their number rose from 2 in 1995 to 121 in 2016 but have declined since, with the number of all‑boys campuses falling 22%, from 58 to 45.
  • Practitioners and researchers both noted that students thrive in schools with strong adult‑student relationships, shared values, active learning, and teaching that meets boys where they are—though practitioners saw all-boys schools as uniquely equipped to deliver these consistently and with fidelity.
  • On average, rigorous international studies show only a small academic edge for boys in gender-separate settings, but there are no rigorous studies of all-boy public schools in the United States.
  • Boys’ behavior and engagement do seem to improve in well-run single-sex schools. However, evidence is thin and sensitive to context, age, and peer quality.
  • Single-sex schools may offer important lessons for co-ed schools and may be succeeding in ways that we cannot assess because of the lack of evidence.

Key facts on single-sex schools

Until a few decades ago, single-sex schooling was restricted to the private sector. Only two public single-sex schools operated in 1995, but following a change in the law in 2006, the number rose. While data on intervening years is spotty, by 2016 the count had increased to 121 (63 for girls and 58 for boys). Since that date, the number of all-boys campuses has fallen slightly, from 58 to 45. Single-sex schools still account for only a tiny fraction of the more than 99,000 public schools nationwide.

 

Figure 1

Data note

The estimate of fully single‑sex public schools in 2022 was based on male and female enrollment data from the Civil Rights Data Collection (CRDC) after removing juvenile‑justice, alternative, and virtual campuses. The list of remaining schools was checked against NCES data to confirm it was (a) still operating, (b) genuinely single‑sex, and (c) not a special‑purpose or very small rural school with coincidental one‑gender enrollment.

2016-2021 data was calculated based on the Feminist Majority Foundation’s criteria which also reproduced the 2022 count exactly.

1995 data was based on media reports cited in Vanze (2010), though other sources claim three single-sex schools at this time.

 

All-girls schools now represent a majority (56%) of public single-sex schools with the change driven more by the closure of all-boys institutions rather than growth in girls’ schools.

These single-sex public schools are much more likely to be educated students of color: eight out of ten students are Black or Latino. They are also quite likely to be charters: around two-fifths of the campuses operate under charter or magnet rules that grant unusual freedom over staffing, scheduling, and culture.

Despite the growth in public single-sex schools, the vast majority of single-sex schools are still private. As of 2020, approximately 793 (2.6%) of the estimated 30,492 private schools in the United States are all-boys schools and 640 (2.1%) are all-girls schools, according to the NCES.

What’s good about boys’ schools?

Several common themes emerged from conversations with both practitioners and academics. Both groups highlighted several school-wide factors as important ingredients for student success, though interviewees diverged on how best to provide them. Some thought that co-ed settings could accomplish this, while others, including many of those working inside all-boys schools, saw single-sex schooling as a uniquely effective way to deliver them.

Key takeaways included:

  • Relationships are critical. Many identified close, high-quality relationships with boys as keys to their success. Boys benefit when staff express genuine interest in their lives and go out of their way to understand their experiences.
  • Effective schools help boys understand who they are. Relationships improve when teachers and staff help boys self-reflect and articulate their thoughts and feelings. Practitioners mentioned activities like talking circles or senior speeches as particularly important for their student bodies.
  • Shared values, traditions, and practices turn schools into communities. Expert interviewees frequently emphasized the importance of ritual and ceremony as an important mechanism for developing character and inculcating shared values.
  • Teachers must understand boys’ and girls’ unique developmental needs. One size does not necessarily fit all, and many interviewees felt strongly about giving staff and teachers the professional development they need to address boys’ educational needs.
  • Experiential, active learning is important. Classroom structures should channel boys’ need for physical movement, challenge, and social connection. Interview participants stressed the value of project-based learning, exposing students to new environments, and getting them outside as opportunities for more transformative experiences.

These ideas are obviously not the exclusive preserve of all-boys schools. Co-ed schools could potentially capture the gains seen in strong boys’ programs by embedding structured relationship time (advising, mentoring), identity-reflection rituals, and active learning. Doing so with the same fidelity as a dedicated all-boys school may be a challenge, however. In particular, deliberately allocating blocks of time to mentoring or sustained staff training on male developmental needs requires administrative buy-in and can be easily squeezed out by conventional academic priorities.

What’s the empirical evidence?

On academic outcomes

While the research base has matured somewhat since the early-2000s, there is a stark absence of methodologically strong research on single-sex schools in the United States since 2006. (There is some on single-sex classrooms). The only studies that either randomize admission or apply rigorous controls for self-selection are from other countries.

They tend to find only small differences in boys’ mathematics or literacy performance between single-sex and coeducational settings. To highlight a few of the more rigorous and more recent studies:

  • Pahlke et al. (2014), in one of the most comprehensive meta-analyses of controlled studies, found a weighted average effect size for boys of 0.06 in math and 0.11 in verbal performance when attending a single-sex school—statistically significant but very small improvements. Among the subset of U.S.-focused studies, the weighted effect sizes were close to zero and insignificant. These effect sizes suggest minimal academic impact of single-sex education on boys’ performance when proper controls are added.
  • Lee and Park (2024) found no academic benefit for boys in mathematics or literacy in South Korea. They did find some academic gains for girls (though also a decline in STEM participation for girls in single-sex schools).
  • Clavel and Flannery (2023), using Programme for International Student Assessment data from Ireland, where about one-third of secondary students attend single-sex schools, similarly found no significant difference in mathematics or reading performance between single-sex and co-ed schools after accounting for individual, parental, and school-level factors.

While the individual studies are often highly heterogeneous (pointing to the importance of some contextual factors, like age), the aggregate positive effects, where they exist, are relatively small compared to other interventions and compared to the size of some current academic gaps between boys and girls.

 

Figure 2

Data note

How to read the benchmark points in the figure

An effect size reports how far an intervention shifts outcomes in fractions of a standard-deviation—so each 0.20 SD roughly raises a student from the 50th to the 58th percentile, while a 0.50 SD improvement would move that same student to about the 69th percentile. Each estimate comes from a published meta-analysis or large‐scale dataset.

  • Phonics vs. whole-language (d = 0.26)Hansford et al. (2024). Mean difference in effect size between structured (phonics-based) literacy program and balanced (whole-language) literacy programs. See table 3.
  • Typical yearly growth, grades 5-8 (d ≈ 0.25)Scammacca, Fall, and Roberts (2015). Average of math and reading gains for 50th percentile learners. See “mean of tests” column in Tables 2 and 3.
  • Current male-female reading gap, Grade 8 (d = 0.23)National Center for Education Statistics (2023). Based on NAEP 2022 national results. See narrative under figure 6. Female = 265 vs. male = 256; 9-point gap/38-point SD ≈ 0.24.
  • Large-scale, high-dose tutoring (d = 0.21)Kraft, Schueler & Falken (2024). Based on the pooled-effect size across the sample of large-scale programs only. Note that the effects are largely driven by elementary literacy tutoring programs. See pg. 5.
  • Single-sex schools – boys’ math (g = 0.06)Pahlke, Hyde & Allison (2014). Meta-analysis, controlled-study subset, math outcome (Table 2).
  • Single-sex schools – boys’ verbal (g = 0.11)Pahlke, Hyde & Allison (2014). Reading/verbal outcomes in the controlled-study subset (table 2).

On behavior and engagement

However, academic scores are only part of the story. Behavioral outcomes—attendance, effort, disciplinary records—also matter for long-run success, especially for disadvantaged boys. Here the data is cautiously encouraging:

  • Delinquency: When Trinidad & Tobago schools converted from co-ed to single-sex, juvenile-delinquency rates among boys fell (Jackson 2021).
  • Attendance and engagement: In Seoul’s lottery-based high-school system, single-sex campuses cut absences and raised homework effort for both sexes, with larger gains for girls but meaningful ones for boys (Lee & Park 2024).
  • Teacher expectations: Boys in boys’ schools are roughly 0.3 SD less likely to agree that “teachers have low expectations of me” (Lee & Park 2017).

The evidence here is limited, but it does suggest that single-sex schools with structured environments and positive student-teacher dynamics can mitigate some of the behavioral risks that derail boys.

Big knowledge gaps persist

Taken as a whole, even with many decades worth of studies, large blind-spots remain:

  • We need more longitudinal studies examining outcomes across students from various educational settings through college and into their careers.
  • Most single-sex public schools in the United States are urban, majority-minority, and in low-income neighborhoods. We need methodologically rigorous research specifically examining outcomes for this subset of boys.
  • There are a range of non-academic outcomes like social-emotional skills, identity development, and motivation that need further exploration and understanding.
  • Many public charter schools use lottery systems and it would be useful to see strong quasi-experimental research on public single-sex schooling in a U.S. context that exploits this.
  • More investigation of the developmental timing question is needed, specifically when and for whom single-sex education might be most beneficial.
  • More resources should be directed toward understanding which specific elements of school environments most benefit boys.

Single-sex education is neither a panacea nor a relic. At its best it bundles proven ingredients like strong mentoring, purposeful rituals, and experiential learning into a format that helps boys succeed and which could potentially be implemented in co-ed settings. However, because the most credible studies are international, we still lack rigorous evidence on U.S. public boys’ schools that might help us understand the unique context in which they operate. We need rigorous (experimental, quasi-experimental, lottery-based) U.S.-centered studies to learn which students truly benefit from single-sex models, and we need to better understand the potential benefits beyond academics.