(Job Market Paper)
Recent efforts by economists to exploit peer effects by creative peer assignment have come up short due in part to endogenous peer selection. That is, even conditional on random assignment, agents choose their peers, and failure to account for this selection may crucially bias predictions of the effects of alternative policies. To address this shortcoming of the literature, I build a two-part model in which (1) agents form a network; (2) conditional on the realized network, outcomes are determined by a process that allows for non-linear peer effects. To overcome difficulties in identification and estimation of network-formation games, agents in my model make continuous linking decisions subject to a budget constraint. I show that, under certain conditions, this model has a unique strictly positive equilibrium, which can then be used for identification and estimation. In modeling peer effects, I explicitly model network endogeneity as an omitted variable problem, and further propose a method to recover these omitted variables in estimating the network-formation game. I estimate the parameters of the two-part model using innovative data on networks and outcomes from a randomized study in Rajasthan, India, then show that the model performs well in matching predictions to realized out-of-sample outcomes. This paper makes important contributions to the methodology of peer effects estimation as well as the theory and econometrics of network formation, while providing an important link between structural and experimental approaches to policy evaluation.
Network Partitioning and Social Exclusion under Different Selection RegimesAvailable here.
(with Clara Delavallade and Rebecca Thornton)
Participation, Learning, and Equity in Education -- Can We Have It All?
While most social programs are based on some form of exclusion of sub-populations, we know little about how being excluded, and the selection process, affect social inclusion. This paper compares peer effects of an after-school program, under three different (randomly assigned) network-formation regimes: endogenously formed, popularity vote, and randomly assigned. We find substantial evidence of homophily within endogenously-formed and elected networks. When participation was randomly assigned, we find segregation of friendships due to the program. We do not find this among elected networks, mainly because they were already highly partitioned. Lastly, we find that social exclusion – not being elected in a school with popular voting – reduced education aspirations and self-confidence.Available here.
(with Clara Delavallade, Gaurav Shukla, and Rebecca Thornton)
The Sustainable Development Goals have set a triple educational objective: improving access to, quality of, and gender equity in education. Despite growing evidence regarding the effectiveness of policies targeting one of these three objectives, however, this study is the first to document the effectiveness of policies targeting each of these objectives simultaneously. Using random variation in exposure across 230 primary schools in rural India, we examine the impact of a multi-faceted educational program on students’ participation and performance, as well as the impact’s heterogeneity across gender and initial performance and the sustainability of the impact over two years. While the program specifically aimed to promote girls’ school enrollment and communities’ sensitization to girls’ education, the learning component of the program targeted and involved boys and girls equally. We find that the multi-faceted program reduced gender gaps in school retention and improved learning during the first year. However, the approach of targeting different educational goals (access, quality, and equity) does not yield sustained effects on either school attendance or learning, nor does it prove to be an effective strategy for bridging gender inequalities in schooling performance over the two year period.
In collecting data on network connections, a common practice is to prompt respondents to name up to a certain number of network links, potentially leading to censoring. This censored data is then used to estimate parameters of peer effects models in a wide variety of economic applications. In this paper, I first provide an analytic form of the bias induced by this practice, showing that this bias decreases as the number of observed links increases. I then conduct a series of Monte Carlo experiments to demonstrate the magnitude of the bias, providing suggestive evidence that it may be substantively meaningful. Using network data from Add Health, I show that different censoring rules induce substantially different estimates of peer-effects parameters. After documenting the possible bias, I propose a number of strategies for researchers working with censored network data. These findings and proposed solutions have potentially wide-ranging applications to research on peer effects through networks as well as the practice of collecting network data.
Extended Abstract available here.
When Interventions Affect the Network: A Decomposition of Treatment Effects in a Partial Treatment Setting
(with Clara Delavallade and Rebecca Thornton)
Identification and Estimation of Peer Effects in the Presence of Endogenous Network Formation: A Control-Function Approach