Last week, Christina Hoff Sommers published an op-ed in The New York Times that called attention to the growing gender gap in educational achievement, a difference that emerges as early as elementary school. While Sommers mentions a number of factors that could explain this trend, she emphasizes "boy friendly" policies as her preferred solution:

What might we do to help boys improve? For one thing, we can follow the example of the British, the Canadians and the Australians. They have openly addressed the problem of male underachievement. They are not indulging boys’ tendency to be inattentive. Instead, they are experimenting with programs to help them become more organized, focused and engaged. These include more boy-friendly reading assignments (science fiction, fantasy, sports, espionage, battles); more recess (where boys can engage in rough-and-tumble as a respite from classroom routine); campaigns to encourage male literacy; more single-sex classes; and more male teachers (and female teachers interested in the pedagogical challenges boys pose).

Claudia Buchmann, co-author our new book, The Rise of Women, published a letter to the editor in the Times responding to Sommers:

The way to deal with boys’ underachievement in school is not through "boy friendly" policies like more recess, single-sex classrooms and male teachers but through strong academic climates and clear, consistent information about occupations and the educational pathways that lead to them. After years of research for our forthcoming book, "The Rise of Women: The Growing Gender Gap in Education and What It Means for American Schools,” Thomas A. DiPrete and I found that in schools where academic effort is expected and valued, boys compete for high grades and more often achieve them. Such schools reduce gender gaps and promote healthy, multifaceted gender identities for both boys and girls. Boys also perform better when teachers and parents help them understand how their future success is linked to their efforts in middle and high school. Rather than remaking schools in ways that reinforce gender stereotypes, we need schools that set high expectations for student achievement and treat students as individuals.

In their book, Buchmann and DiPrete review the latest research on some of the policy ideas Sommers endorses, and find good reason to be skeptical. On same-sex schools, for example, they write:

Caroline Hoxby (2000) used exogenous variation in classroom composition from the Tennessee Project STAR data to obtain a plausible estimate of the impact of gender classroom composition on student achievement. She found that both males and females have small but systematic advantages in math and reading in classrooms that are more heavily female. Furthermore, the gains appear to hold net of students’ ability. Diane Whitmore (2005) obtained similar findings with the same data. In a study of Israeli schools, Victor Lavy and Analía Schlosser (2007) used random variation in the fraction of classrooms that are female in order to estimate the impact of gender composition on school outcomes. Like previous studies, they found higher cognitive outcomes in classes with a higher fraction of girls. They concluded that higher fractions of girls raise the intermediate factors that, as shown in chapter 6, tend to be associated with girls—namely, lower levels of classroom disruption, better relationships between students and teachers and among students, a higher overall level of satisfaction with school, and less teacher fatigue. Given that students have differing needs, some boys would probably be better served by single-sex schools. In general, however, existing evidence suggests that boys would not be helped by sex-segregated classrooms.

Similarly, on the need for more male teachers to teach boys:

Many scholars have speculated that boys might benefit by having more male teachers, especially in elementary and middle school, where teachers are predominantly female. To date, however, there is little hard evidence supporting this conjecture. [...] Helena Holmlund and Krister Sund (2008), who examined upper secondary students’ gains in natural science and social science in Sweden over three years, were able to control for unmeasured student and school characteristics using fixed effects. They found no effect of gender matching on performance as measured by grades. Marcel Helbig (2010) similarly found no relationship between the percentage of teachers who are female and sixth-grade boys’ reading or mathematics competence in Germany.2 This study used data from the Progress in International Reading Literacy Study (PIRLS), which allows a stronger test of the same-sex teacher effect because it contains data on teacher gender and individual achievement data for all students in the class. Moreover, in Germany the process of student assignment to classrooms largely rules out the possibility of nonrandom selection of teachers. Using these data and a variety of modeling strategies, Martin Neugebauer, Marcel Helbig, and Andreas Landmann (2010) found no evidence that student achievement is affected by whether the teacher is the same gender or a different gender. Bruce Carrington and his colleagues (2007) took a different approach by interviewing three hundred seven- and eight-year-old students in England; they found no evidence that gender matches between teacher and student have an effect on academic engagement.

In the coming months, we will discuss Buchmann and DiPrete's policy suggestions, as well as their broader assessment of the growing gender gap in education. For now, you can read the introduction to their book for free.