Deep description of static and dynamic network ties in Honduran villages

TL;DR

The paper investigates how static and dynamic multiplex social networks are formed in rural Honduran villages across four relationship layers: health, friendship, financial, and adversarial. Using near-census data from 176 villages (n=20,232 individuals) across two waves (2016 and 2019), the authors apply multivariate mixed-effects zero-inflated negative binomial models, dyadic assortativity tests, and meso-scale community analyses (Louvain modularity, NMI) to link individual attributes and village context to network dynamics. Key findings show strong co-movement among cooperative layers but only modest dyadic overlap, with pronounced gender and religion-based assortativity; education and financial factors gain influence over time, while communities align most with friendship. The study provides actionable insights for designing multiplex-aware interventions in rural development, highlighting the importance of leveraging cross-layer ties and cultural factors to improve diffusion, resilience, and resource sharing.

Abstract

We examine static and dynamic social network structure in 176 villages within the Copan Department of Honduras across two data waves (2016, 2019), using detailed data on multiplex networks for 20,232 individuals enrolled in a longitudinal survey. These networks capture friendship, health advice, financial help, and adversarial relationships, allowing us to show how cooperation and conflict jointly shape social structure. Using node-level network measures derived from near-census sociocentric village networks, we leverage mixed-effects zero-inflated negative binomial models to assess the influence of individual attributes, such as gender, marital status, education, religion, and indigenous status, and of village characteristics, on the dynamics of social networks over time. We complement these node-level models with dyadic assortativity (odds-ratio-based homophily) and community-level measures to describe how sorting by key attributes differs across network types and between waves. Our results demonstrate significant assortativity based on gender and religion, particularly within health and financial networks. Across networks, gender and religion exhibit the most consistent assortative mixing. Additionally, community-level assortativity metrics indicate that educational and financial factors increasingly influence social ties over time. Our findings provide insights into how personal attributes and community dynamics interact to shape network formation and socio-economic relationships in rural settings over time.

Figures (10)

• Figure 1: Village‑level coupling and dyad‑level specialization of cooperative layers.Left: Correlations among village random effects from the multivariate degree model (Friendship, Health, Financial), faceted by wave (W1, W3). When random effects are unavailable, the panel uses per‑village mean degree as a proxy (results are nearly identical). Right: Distribution of dyad‑level overlap (Jaccard index) between pairs of cooperative layers across villages, by wave. Violin shapes show the across‑village distribution; points mark medians.
• Figure 2: Upper Left: Number of communities per village (aggregated, min size $=4$). Upper Right: Alignment (NMI) between communities and single attributes. Lower Left: Mean $|z|$ (95% CI) from within‑village multinomial logit. Lower Right: Share of villages with mean $|z|{>}1$ by attribute.
• Figure 3: Dyad-level assortativity in health ties (Wave 1). For each village, we fit dyad-level logistic regressions predicting the presence of a health tie from attribute similarity. Panels report village-specific odds ratios (points) with 95% confidence intervals for each attribute; villages are ordered by the odds ratio within each panel. Red points indicate villages where the similarity coefficient is statistically significant ($p<0.05$). The dashed horizontal line marks OR$=1$ (no association). See Table 18 for corresponding numerical details.
• Figure 4: Dyad-level assortativity in health ties (Wave 3). For each village, we fit dyad-level logistic regressions predicting the presence of a health tie from attribute similarity. Panels report village-specific odds ratios (points) with 95% confidence intervals for each attribute; villages are ordered by the odds ratio within each panel. Red points indicate villages where the similarity coefficient is statistically significant ($p<0.05$). The dashed horizontal line marks OR$=1$ (no association). See Table 19 for corresponding numerical details.
• Figure 5: Dyad-level assortativity in friendship ties (Wave 1). For each village, we fit dyad-level logistic regressions predicting the presence of a friendship tie from attribute similarity. Panels report village-specific odds ratios (points) with 95% confidence intervals for each attribute; villages are ordered by the odds ratio within each panel. Red points indicate villages where the similarity coefficient is statistically significant ($p<0.05$). The dashed horizontal line marks OR$=1$ (no association). See Table 20 for corresponding numerical details.