Girls Next Door

The study, conducted in Greece by Sofoklis Goulas, Rigissa Megalokonomou, and Yi Zhang, set out to elucidate whether social networks outside school matter for students’ learning and future career. Specifically, whether the gender of members of that network plays any role on academic performance, post-secondary admission, degree choice, and expected wage. The researchers define a relevant peer group for high school students as their same-cohort peers who reside near them but attend different schools in the local area. The reasoning is that, unlike university students, high-school ones tend to interact with neighbors in their same age group, participating in after-school activities, sports and the like.

The data the authors used allowed them to identify all same-cohort students who live nearby (e.g., a block away) and attend another neighboring school, no more than one mile away. Helpful features for this were that in Greece students are assigned to schools based on the proximity of their residential address, and that schools are built very close to each other.

They then compared the outcomes and choices of students from consecutive cohorts that had similar characteristics and faced the same environment, except for the fact that one cohort had a higher share of female neighbors than the other.

They found that having a higher share of female neighbors increases both genders’ national exam scores, matriculation rates and matriculation scores, and it also makes both genders more likely to enroll in an academic university (versus a technical school) and a more selective/prestigious university department.

Neighbors’ gender is also found to impact enrollment in STEM degrees for females but, curiously, not for males. The impact of this kind of student peer networks can be long-lasting, affecting even the expected wage of females. The effects are larger for higher proportions of females in the neighborhood—i.e., above 56%. This result underscores the value of peer networks outside school in high-density environments, such as urban areas.

Why this should be so is suggested by findings of a study on in-class effects by a different researcher: individuals tend to be less disruptive, more focused, and better behaved in environments in which females outnumber males. This may also apply out-of-class: if a larger proportion of females leads to lower levels of disruption or violence in a neighborhood, this will in turn lead to increased trust, more discipline and a greater feeling of safety among individuals in the local community. It would also improve student behavior and engagement in learning activities.

These positive externalities of a higher share of females may be more relevant in neighborhoods in which young individuals have more interaction channels.

The trick, in any case, appears to be the intensity of direct social interaction. So how do you foster such direct interaction, in an age when those young cohorts tend to spend their free time glued to their smartphone screens?

Access to communal spaces, that’s the thing. Such spaces may foster social interaction and social integration, and encourage learning engagement. And, bingo, the authors found that gender neighbor effects are amplified for females in local communities that have a higher density, or “intensity”, of public facilities. The intensity of facilities in the local community, strangely enough, does not much seem to affect males’ outcomes. Why that is so should be the subject of further research.