Access to credit is vital for economic well-being, yet millions of rural households in India remain financially vulnerable despite the expansion of formal institutions (Rajeev & Nagendran, 2024). This issue is particularly acute in the semi-arid tropics, where agrarian distress, farmer suicides, and caste-driven social stratification persist alongside weak financial infrastructure (Bhende, 1983). In this context, social networks serve as crucial conduits for informal credit (Wydick et al., 2011), enabling households to invest in agriculture, adopt new technologies, and smooth consumption during economic shocks. These networks, deeply embedded in kinship, caste, and trust, shape financial access. However, the formation and evolution of these networks—and their influence on credit access—remain insufficiently understood (Banerjee et al., 2024).

This study examines informal borrowing-lending networks in Kanzara village, addressing three key questions. First, it investigates how endogenous network structures shape borrowing relations. Second, it explores how household characteristics and tie-specific factors influence these ties. Third, it assesses the role of caste in credit access. Using longitudinal data from the ICRISAT VDSA survey (2007, 2013) and a Stochastic Actor-Oriented Model (SAOM) (Snijders, 2017), we find borrowing relations are highly reciprocal and caste-homophilous, with dominant castes lending within their groups while lower castes depend on them. Spatial proximity, patron-client ties, and external financial links further shape borrowing dynamics. These findings underscore the need for financial inclusion policies to address caste-based inequalities in credit access.

References

Banerjee, A., Breza, E., Chandrasekhar, A. G., Duflo, E., Jackson, M. O., & Kinnan, C. (2024). Changes in social network structure in response to exposure to formal credit markets. Review of Economic Studies, 91(3), 1331-1372.

Bhende, M. J. (1983). Credit markets in the semi-arid tropics of rural south India.

Rajeev, M., & Nagendran, P. (2024). Has Farmers' Access to Credit Improved Over Time? An Analysis of NSSO Data. In Institutions and Public Policy for India’s Sustainable Development (pp. 117-136). Routledge India.

Snijders, T. A. (2017). Stochastic actor-oriented models for network dynamics. Annual review of statistics and its application, 4(1), 343-363.

Wydick, B., Hayes, H. K., & Kempf, S. H. (2011). Social networks, neighborhood effects, and credit access: evidence from rural Guatemala. World Development, 39(6), 974-982.

Economic inequality remains a persistent issue, shaping societal structures and influencing public attitudes toward redistribution. While socioeconomic status and ideology are well-established predictors of redistributive preferences, the role of social networks has gained increasing attention. This study examines how homophily -the tendency of individuals to associate with others who share similar characteristics- affects beliefs about inequality and support for redistribution.

Using Chile as a case study, a country with high levels of inequality and pronounced social stratification, we analyze how educational and ideological homophily shape perceptions of meritocracy, tolerance for inequality, and redistributive policies. Based on data from the Longitudinal Social Survey of Chile (ELSOC, 2016) and the United Nations Development Programme (UNDP) survey for Chile (2016), we apply a case-control model to estimate homophily patterns, followed by multiple linear regressions to assess their impact on redistributive preferences.

Our findings indicate that educational homophily reinforces meritocratic beliefs and weakens support for systemic redistribution. In contrast, ideological homophily produces divergent effects: in left-leaning contexts, it tends to challenge inequality-justifying narratives, while in right-leaning contexts, it reinforces them. These results highlight the role of social networks in shaping economic attitudes, demonstrating the potential of a social network approach to understand better how beliefs, attitudes, and preferences about inequality are formed.

The principal eigenvector of an adjacency matrix is the basis of several important metrics in social network analysis. Bienenstock and Bonacich (2021) recently proposed using the Gini-coefficient of the principal eigenvector of the communication matrix to measure the extent to which the centralization of communication in an organization prevents that organization from incorporating outside information. This problem can be exacerbated when informational flows are one way. On directed graphs, nodes that are informed may not be influential, and influential nodes may not be informed. This subtle distinction is highlighted by the relationship between the forward and reverse versions of the DeGroot influence model with the corresponding reverse and forward versions of PageRank. We thus introduce two separate Gini-coefficient measures of influence inequality and information inequality for directed graphs and explore their implications for understanding influence and access in organizational networks.

Cryptocurrencies have recently been in the spotlight of public debate due to their embrace by the new US President, with crypto fans expecting a 'bull run'. The global cryptocurrency market capitalisation is more than $3 trillion, with 1 Bitcoin exchanging for more than $109,000 at the end of January 2025. Monitoring the evolution of these systems is key to understanding whether the popular perception of cryptocurrencies as a new, sustainable economic infrastructure is well-founded. In this paper, we have reconstructed the network structures and dynamics of Bitcoin from its launch in January 2009 to December 2023 and identified its key evolutionary phases. Our results show that network centralisation and wealth concentration increased from the very early years, following a richer-get-richer mechanism. This trend was endogenous to the system, beyond any subsequent institutional or exogenous influence. The evolution of Bitcoin is characterised by three periods, Exploration, Adaptation and Maturity, with substantial coherent network patterns. Our findings suggest that Bitcoin is a highly centralised structure, with high levels of wealth inequality and internally crystallised power dynamics, which may have negative implications for its long-term sustainability.

Through a survey, carried out for my thesis, based on the study of 464 YouTube channels and their videos, I am interested by the collective dynamics that structure the space of scientific popularization on the Internet. I study the nature of the social relations created by popularizers to structure their sphere and develop their activity. I observe how these relationships evolve and multiply over time and the individual's place in the network. I look at which actors are mobilized and the nature of these relationships. They collaborate not only with each other, but also with scientific institutions, with media and with commercial partners. Video collaborations are dated, so they can be studied over time. In addition, YouTube channels are studied in terms of their scientific discipline, seniority, audience and gender, enabling us to compare them in terms of the differences that can be observed about their trajectories and networks.

In many low-fertility countries, individuals with higher socioeconomic status (SES) have recently started to have more children than those with lower SES. However, the reasons for this shift are still not fully understood. Recent research increasingly highlights the role of social networks in fertility decisions, yet their role in explaining SES differences in fertility remains largely unexplored. To address this gap, this study uses population-level, multiplex social networks derived from Swedish register data to examine network mechanisms that could explain SES differences in fertility. Focusing on a random sample of women employed in small workplaces (≤ 50 employees) between 1990 and 2022, we analyse how exposure to workmates’ childbirth events affects the focal individual’s probability of having a child and how the effect differs across SES groups. Additionally, we employ instrumental variables to disentangle social influence from other contextual factors and use agent-based simulations to assess their macro-level consequences. In this study, we find evidence for two well-established inequality-producing mechanisms. First, women are primarily exposed to childbirth events among same-SES workmates, reinforcing baseline fertility differences. Second, workmates who share the same SES category influence each other’s fertility to a greater extent than those who do not, further deepening baseline fertility differences. These preliminary findings underscore how social networks shape fertility differentials, offering new insights into the mechanisms underlying SES differentials in childbearing behaviour in contemporary low-fertility societies.

Adolescent friendship networks are characterized by low interaction across both socioeconomic and racial lines. Using data from the National Study of Adolescent Health and a new exponential random graph modeling approach, this study examines the degree, pattern, and determinants of socioeconomic segregation and its relationship to racial segregation in friendship networks in high school. The results show that friendship networks are overall less socioeconomically segregated than they are racially segregated. However, the exclusion of low-SES students from high-SES cliques is pronounced and, unlike racial segregation, unilateral rather than mutual: many friendship ties from low-SES students to high-SES peers are unreciprocated. The decomposition of determinants indicates that about half of the socioeconomic segregation in friendship networks can be attributed to differences in socioeconomic composition between schools. The other half is attributable to students’ friendship choices within schools and driven by stratified courses (about 13 percent) as well as racial and socioeconomic preferences (about 37 percent). In contrast, relational mechanisms like triadic closure – long assumed to amplify network segregation – have only minor effects on socioeconomic segregation. These results highlight that SES-integrated friendship networks in educational settings are difficult to achieve without also addressing racial segregation. Implications for policymakers and educators are discussed.

Homophily is ubiquitous—people tend to associate with similar others in different settings of social life, from education to relationships to employment. This preference for in-group ties is a trade-off: it strengthens social groups through segregation. While homophily nurtures a feeling of belongingness, it can also limit the access to out-group opportunities and exacerbate inequalities. Although this trade-off is a core building block of social networks, it remains poorly understood and analytically unexplored.

The homophily trade-off is particularly critical in networks with minorities. When minority members favor in-group ties, they inherit not only the opportunities but also the limitations of their group. For instance, at social gatherings, homophily within small social groups limits individuals’ contact pool, resulting in fewer connections on average for these numerical minorities [1]. These inherited limitations may also stem from the social capital of a group. For example, immigrants relying on intra-ethnic contacts to find jobs might end up in low-wage positions, potentially leading to an ethnic mobility trap that hinders upward social and economic mobility [2]. Nevertheless, despite the negative effects on minority groups, this trade-off remains poorly understood, lacking an analytical approach to identify its underlying mechanisms and understand how intrinsic structural limits in networks sustain it, producing inequalities.

In this work, we explore homophily in networks analytically to disentangle its inherent trade-off. We investigate when homophilic ties are detrimental to minority groups, introducing the concept of the homophily trap—scenarios where increased homophilic interaction among minorities negatively affects their structural opportunities within a network. To study these scenarios, we use a generative network model to construct networks of different group mixing and minority sizes. We show that homophily traps arise when the minority group size falls below 25% of a network. Below this threshold, higher homophily within the minority group leads to fewer structural opportunities for the group: in-group ties come at the expense of lower structural visibility. This trade-off makes it difficult for numerical minorities to both maintain a high number of connections and belong to homophilic social groups. By disentangling the trade-off of homophily analytically and systematically, we build a foundation for understanding how homophily shapes structural opportunities in networks.

[1] Oliveira, M. et al. Group mixing drives inequality in face-to-face gatherings. Communications Physics 5, 1–9 (2022).

[2] Wiley, N. F. The ethnic mobility trap and stratification theory. Social Problems 15, 147–159 (1967).

Interpersonal status orders are a ubiquitous feature of human societies. Social scientists usually explain the emergence of prestigious, highly popular elites as a functional adaptation of social systems, as a result of cognitive biases and network mechanisms, or as a consequence of actors hoarding resources and power.

We add to previous work by arguing that status orders can emerge as an unintended by-product of communication. To explore this theoretical argument, we build an agent-based model that simulates face-to-face encounters in which agents talk about absent others and form status evaluations afterward. Our model demonstrates that the simple assumption that actors discuss others can produce highly skewed distributions of status and a systematic decoupling of status from quality. Moreover, our model yields that actors tend to misperceive the quality of others more strongly if they are further away in the network of face-to-face interactions. Model explorations also show that inequality and decoupling are amplified by large and highly connected networks and cognitive biases occurring in communication and status evaluations.

In sum, our study adds a new facet to the longstanding debate on the emergence of status orders by exploring the interplay between communication, networks, and cognitive biases with agent-based simulations for the first time. Thereby, we arrive at the surprising conclusion that the seemingly trivial act of talking about others in their absence could be an important driver for the emergence of status orders in human groups.

The pervasiveness of socioeconomic inequality could extend into social media platforms like Twitter. However, relevant empirical evidence remains rare and fragmented. This study leverages a recently developed method for estimating Twitter users’ individual socioeconomic status (SES) based on the brands they follow to examine socioeconomic inequality in social capital and communication behaviours on Twitter. First, this paper establishes that higher SES Twitter users exhibit higher social capital across multiple measures, including degree, reciprocity, topological diversity, local clustering coefficient, and effective size. As a result, the paper advances efforts to quantify the relationship between socioeconomic outcomes and social capital in large-scale digital networks. Second, compared with the existing scattered evidence, this paper provides a more comprehensive picture of the relationship between SES and communication behaviours on Twitter. The analysis demonstrates that higher SES users use more complex and future-oriented language in their tweets. Also, while high and low SES users mostly talk about similar topics, they diverge in hashtag usage and attitudes toward immigration. These findings suggest that socioeconomic inequalities are not only reflected but also potentially reinforced on social media, underscoring the critical roles of social capital and communication behaviours in perpetuating inequality. The study highlights the need for further research to explore the underlying mechanisms and integrate SES as a critical factor in social media research.

People seek wealthy partners for upward mobility, while the wealthy benefit from the cooperation of those around them, reinforcing resource advantages. This research examines how exchange networks and cooperation co-evolve, focusing on when network partners influence cooperation and whether people choose partners based on wealth or cooperation homophily.

We collected data from 1,080 participants via Prolific, assigning them to 40 randomly generated networks of ~28 players. Participants engaged in repeated interactions where they decided how much to cooperate with network alters. Networks varied by (1) whether players started with equal endowments and (2) whether cooperation from wealthier partners had a productive effect. A stochastic actor-based model estimated both network formation and changes in cooperation levels.

Cooperation increases faster in networks with resource inequality and wealth productivity, suggesting inequality promotes cooperation. Regarding network formation, we find a wealth heterophily tendency—poorer players prefer ties with wealthier players over other poorer individuals, regardless of endowment equality. Further, in unequal endowment conditions, players move in the opposite direction of their alters’ cooperation levels. They are less likely to cooperate when many alters have high cooperation rates, suggesting resistance against defection.

Networks amplify inequality, as individual disadvantages and network effects accumulate. While cooperation is generally valued, this study highlights its potential downsides in stratified resource environments, where prosocial behaviors benefit actors unequally.