Authors: Shivkumar Vishnempet Shridhar†, Selena T. Lee†, Yanick Charette, George Iosifidis, and Nicholas A. Christakis* By: Human Nature Lab (Yale), United States of America
Network-cycle motif participation is associated with individual and collective wealth in Honduran villages
This github repo describes how to caclulate network cycle membership, and investigates how individual participation in cycles is related to wealth
Geodesic cycles, or loops of nodes connected in a sequence, are an important network motif, and their prominence or deficiency is associated with both beneficial and detrimental properties in diverse kinds of networks. Here, we examine network cycles formed by informal borrowing/lending and friendship ties among 24,702 rural Hondurans (in 176 isolated villages), and we explore their association with personal and community wealth across two time points. We find that cycles of different lengths (i.e., 3 or 4 ties) constitute an over-represented motif, and their quantity is strongly associated with individual wealth, i.e., richer individuals are involved in more cycles. Furthermore, we introduce a new metric of cycle composition, defined as the average of some measures (such as wealth) of a node’s alters in its cycles, and find that this metric outperforms cycle quantity as an indicator of both current and future wealth. A longitudinal analysis also reflects a higher participation rate in future cycles among wealthier individuals. Finally, cycle composition is a strong indicator of overall village wealth. In some, the potential for the flow of money in a village through structural social network cycles is associated with both individual-level and village-level wealth.
This folder describes the estimation of cycles in real (obeserved) networks against Null (synthetic) networks.
This folder describes the relationship between individual participation in cycles/paths and cycle composition with various measures of wealth.