Project Summary Interventions that reduce childhood malnutrition in resource-limited settings around the world are critical, and recent work suggests that social relationships can affect health-promoting behaviors using social accountability (SA) interventions. Our project, Social Networks and Health in Uttar Pradesh, India, will study social networks in 80 villages where the State Government of Uttar Pradesh (UP), India, introduced accountability interventions to improve delivery of key health and nutrition services. Our general model posits greater information intervention effectiveness in cohesive, well-connected networks. We propose to examine how (1) network structure is related to health at baseline, (2) the social network moderates intervention effectiveness at the village level, and (3) position within the network is associated with household-level health after the intervention. Starting in 2016, the government implemented SA interventions that included information about rights and entitlements to publicly provided services and healthcare opportunities. A recent cluster-randomized evaluation of these efforts found large improvements in average childhood health outcomes, including reductions in indicators of malnutrition and increases in immunization rates, although it also noted effect heterogeneity across treated villages and likely unequal effectiveness within villages. Our study will leverage unique social network data collected in 2016, prior to the SA interventions, along with two waves of household survey data from before and after the cluster-randomized trial of the intervention. In Aim 1, we will estimate the relationship between village-level network structure—measured using complete network data and a novel complete- network clustering technique—and pre-intervention community-level health outcomes. This aim sets the foundation for explaining between-village effect heterogeneity. In Aim 2, we will determine whether network characteristics account for observed health outcome heterogeneity and healthcare service use net of baseline, two years after implementation of SA interventions. This aim tests whether core information and social pressure related network features can account for the effectiveness of the intervention. For Aim 3, we will use a hierarchical linear modeling (HLM) framework to evaluate the relationship between individual households’ positions within the village social network and household-level health. This aim helps us identify within-village heterogeneity to test expectations that those with greater exposure to informed others will be healthier. This novel project will explore the potential influence of network structure on information diffusion, social position, and health outcomes within real-world, village-level social networks and build a foundation for future research on dynamic networks and community-based health interventions that can leverage the growing availability of large-scale, dynamic, community-bas...