Association for Public Policy Analysis & Management

Aspiration is a fundamental determinant of one’s effort and investments (Sewell and Hauser, 1975; Kao and Tienda, 1998; Genicot and Ray, 2017; Dalton et al., 2016). Due to its consequences for individuals’ future outcomes, understanding the process of aspirations formation helps to inform public policies. The contributions from the theoretical literature argue that individual aspirations emerge in social contexts, through individuals’ comparisons with similar others (Appadurai, 2004; Ray, 1998, 2006; Genicot and Ray, 2017). This important social element of aspirations construction calls for investigations on how exactly peers influence an individual’s level of aspiration.

The present work uses a unique social networks data collected from middle school students in Brazil and investigates peer effects on students’ college aspirations - that is, students’ willingness to pursue a college degree. In line with previous literature, I first document that college aspiration is positively associated with students’ effort in school - measured by their report on how many hours they usually study, and on whether they take part in study groups - and with their probability of graduating or passing grades in high school. Next, I rely on network structures to address the main challenges that emerge in the identification of peer effects (Manski, 1993). Differently from standard linear-in-mean models, this work does not assume that all individuals in a student’s reference group are equally connected or have the same influence on each other. Instead, it acknowledges that individuals in social networks are idiosyncratically connected, taking this into consideration in the identification strategy (as in König et al. (2018) and Santavirta and Sarzosa (2019)).

An advantage of my dataset is the possibility to link it with administrative records. I explore this to model friendship formation based on homophily in predetermined characteristics and on random allocations of students into classes when enrolling in the first year of middle school. Based on these predicted friendship ties, the identification strategy uses friends of friends’ characteristics as instrumental variables for friends’ aspiration (as in Bramoullé et al. (2009) and De Giorgi et al. (2010)). It also uses network fixed effects to eliminate other possible correlated effects.

I find evidence of positive, significant, and quite large peer effects on aspiration: if a student passes from having no friends aspiring to a college degree to having all her nominated friends aspiring to it, her probability of also aspiring to a college degree increases around 20 percentage points. Heterogeneous exercises show that non-whites and students from less educated mothers are those more influenced by their friends, which shows the importance of peers in influencing the aspiration levels of minority students. Information diffusion and compliance with social norms seem to explain at least part of these results. I finally explore more tangible impacts of peers’ aspiration, showing that, even after controlling for students’ own aspiration, peers’ aspiration increases the likelihood of finishing high school.