Bayesian learning in social networks

We extend the standard model of social learning in two ways. First, we introduce a social network and assume that agents can only observe the actions of agents to whom they are connected by this network. Secondly, we allow agents to choose a different action at each date. If the network satisfies a connectedness assumption, the initial diversity resulting from diverse private information is eventually replaced by uniformity of actions, though not necessarily of beliefs, in finite time with probability one. We look at particular networks to illustrate the impact of network architecture on speed of convergence and the optimality of absorbing states. Convergence is remarkably rapid, so that asymptotic results are a good approximation even in the medium run.     

Signaling in social network and social capital formation

This paper presents a model of social capital and social network formation. The key interaction within the model is that whom an individual chooses to become friends with affects the value (social capital) of the friendship. In the model, how a player searches for and then forms friendships reveals how willing she is to engage in cooperation with a potential friend. Individuals observe their local network structure (friends and cliques) and the actions of players within these. Willingness to cooperate is private information and is captured by the discount factor of an individual. Cooperative types have high discount factors and can signal their type by forming a clique through befriending a friend of a friend. Uncooperative types do not form these kinds of friendships because of the local observability of their actions to all members of a clique. Thus, when a player meets someone with whom she shares a friend, her belief that the individual has a high discount factor is greater than the population average. In this sense, people “trust” each other more when they share a friend in common. Finally, I relate the primitives of the model to characteristics of the implied social network by nesting the sequential equilibrium in an algorithm of network formation proposed by Jackson and Rogers (Am Econ Rev 97(3):890–915, 2007). The model highlights a trade-off between maximizing the total amount of social capital in a society and distributing it equitably across individuals.     

Job contact networks

Many workers hear about or obtain their jobs through friends and relatives. The aim of this paper is twofold. First, we relate both individual and aggregate labor market outcomes to the network structure of personal contacts. Second, we study strategic network formation. To this purpose, we develop a model specifying at the individual level both the decision to form contacts with other agents, and the process by which information about jobs is obtained and transmitted. We show that equilibrium networks always exist and that only moderate levels of network asymmetry can be sustained at equilibrium. Also, we establish a general non-monotonicity result on information flow and unemployment with respect to network size in symmetric networks.     

Endogenous formation of multiple social groups

This paper explores a model of group membership formation in which agents decide to join or not multiple social groups. The membership formation process induces a bipartite graph structure with social groups listed on one side and agents listed on the other side. Among members of multiple social groups, we consider two decisive types of agents: the grand star and the mini star. The former type is the unique agent in a society who participates in all social groups. The latter type includes agents who participate in more than one, but not all, social groups such that every social group pair has one and only one common member. We analyze the efficiency and stability conditions of group membership formation, and we establish sufficient conditions under which a connected graph that contains either a grand star or a set of mini stars becomes the unique strongly efficient and stable graph.     

Influential listeners: An experiment on persuasion bias in social networks

This paper presents an experimental investigation of persuasion bias, a form of bounded rationality whereby agents communicating through a social network are unable to account for repetitions in the information they receive. We find that, after repeated communication within a social network, social influence depends not only on being listened to by many others, but also on listening to many others. We show that persuasion bias can be viewed as an extreme case of a generalized boundedly rational updating rule in which agents receive more or less attention depending on how many other agents they listen to. The results indicate that behavior in the experiment is consistent with an updating rule according to which agents' social influence is proportional to their indegree.     

Anti-coordination and social interactions

This paper studies the first model of social interactions with anti-coordination. Agents have fixed partners with whom they play a common bilateral game of anti-coordination, like the chicken game. Partnerships are represented as links of a network. The paper asks: How do social interactions interplay with the incentives to anti-coordinate? How does the social network affect individual choices in equilibrium? The analysis shows that network effects are much stronger than when partners play a coordination game. The network notably affects how agents respond when a strategy becomes more advantageous. In the standard benchmark of complete interactions, more agents play the strategy. This result does not generally hold when the network is incomplete. On bipartite networks, agents' choices are unaffected, while on core-periphery networks, less agents may play the strategy.     

Social media networks,fake news,and polarization

We study how the structure of social media networks and the presence of fake news affects the degree of misinformation and polarization in a society. For that, we analyze a dynamic model of opinion exchange in which individuals have imperfect information about the true state of the world and exhibit bounded rationality. Key to the analysis is the presence of internet bots: agents in the network that spread fake news (e.g., a constant flow of biased information). We characterize how agents’ opinions evolve over time and evaluate the determinants of long-run misinformation and polarization in the network. To that end, we construct a synthetic network calibrated to Twitter and simulate the information exchange process over a long horizon to quantify the bots’ ability to spread fake news. A key insight is that significant misinformation and polarization arise in networks in which only 15% of agents believe fake news to be true, indicating that network externality effects are quantitatively important. Higher bot centrality typically increases polarization and lowers misinformation. When one bot is more influential than the other (asymmetric centrality), polarization is reduced but misinformation grows, as opinions become closer the more influential bot’s preferred point. Finally, we show that threshold rules tend to reduce polarization and misinformation. This is because, as long as agents also have access to unbiased sources of information, threshold rules actually limit the influence of bots.     

Social network and private provision of public goods

Using a simple model with interdependent utilities, we study how social networks influence individual voluntary contributions to the provision of a public good. Departing from the standard model of public good provision, we assume that an agent’s utility has two terms: (a) ‘ego’-utility derived from the agent’s consumption of public and private goods, and (b) a social utility which is the sum of utility spillovers from other agents with whom the agent has social relationships. We establish conditions for the existence of a unique interior Nash equilibrium and describe the equilibrium in terms of network characteristics. We show that social network always has a positive effect on the provision of the public good. We also find that, in networks with “small world”-like modular structures, ‘bridging’ ties connecting distant parts of social network play an important role inducing an agent’s contribution to public good. Assumptions and results of the model are discussed in relation to the role of social capital in community-level development projects and to the effect of innovation networks on firms’ R&D investments.     

Adolescent alcohol use,sociability and income as a young adult

We use data from the National Longitudinal Survey of Adolescent Health (Add Health) to study how sociability and adolescent alcohol use impact personal income as a young adult. We find that factors which enhance not only individual sociability but also social interaction at the community level are positively linked to future earnings of adolescents. Adolescents whose friends and friends of friends have greater sociability reap long-term labour market rewards into adulthood. After adjusting for individual and community sociability, the effect of teenage alcohol consumption on labour market earnings as young adults is reduced. Our results suggest that earnings premiums associated with adolescent alcohol consumption may be partially explained by social network dynamics.     

An experimental study of persuasion bias and social influence in networks

In many areas of social life, individuals receive information about a particular issue of interest from multiple sources. When these sources are connected through a network, then proper aggregation of this information by an individual involves taking into account the structure of this network. The inability to aggregate properly may lead to various types of distortions. In our experiment, four agents all want to find out the value of a particular parameter unknown to all. Agents receive private signals about the parameter and can communicate their estimates of the parameter repeatedly through a network, the structure of which is known by all players. We present results from experiments with three different networks. We find that the information of agents who have more outgoing links in a network gets more weight in the information aggregation of the other agents than under optimal updating. Our results are consistent with the model of “persuasion bias” of DeMarzo et al. (2013. Q. J. Econ., 909) and at odds with an alternative heuristic according to which the most influential agents are those with more incoming links.     

社会网络、非正规金融与创业

社会网络可以缓解信息不对称问题,从而促进民间借贷。本文使用中国农村的调查数据,发现拥有更多社会网络的农民,会有更多的民间借贷渠道,从而更有可能创办自营工商业。农村个体工商业的初始投资和后续发展所需的资金很大程度上来自于亲友的借款。本文还发现,在正规金融越不发达的地方,民间借贷对农民创办自营工商业所发挥的作用越大。这表明,依托亲友关系的非正规金融弥补了农村正规金融发展滞后的缺陷。     

(Non-)existence and Scope of Nash Networks

For the connections model of strategic network formation, with two-way flow of information and without information decay, specific parameter configurations are given for which Nash networks do not exist. Moreover, existence and the scope of Nash network architectures are briefly discussed. We would like to thank two referees for helpful comments.     

Transmission of information within transnational social networks: a field experiment

Several studies have shown a relationship between the stocks of migrants and country-level investment in the home country; however the mechanism through which this relationship operates is still unexplored. We use a field experiment in which participants who are recent immigrants send information about risky decisions to others in their social network in their home country. The results demonstrate how this information influences decisions in the home country. We find that the advice given by family members and decisions made by friends significantly affects an individual’s risky decision-making.     

Rewards and the private provision of public goods on dynamic networks

We study an evolutionary model of a public good game with rewards played on a network. Giving rewards to contributors transforms the game but gives rise to a second-order dilemma. By allowing for coevolution of strategies and network structure, the evolutionary dynamics operate on both structure and strategy. Players learn with whom to interact and how to act and can overcome the second-order dilemma. More specifically, the network represents social distance which changes as players interact. Through the change in social distance, players learn with whom to interact, which we model using reinforcement dynamics. We find that, for certain parameter constellations, a social institution, prescribing prosocial behavior and thus solving the second-order dilemma, can emerge from a population of selfish players. Due to the dynamic structure of the network, the institution has an endogenous punishment mechanism ensuring that defectors will be excluded from the benefits of the institution and the public good will be supplied efficiently.     

Social network structure,segregation, and equality in a labor market with referral hiring

We examine the effects of social network structure on inequality in a model of referral hiring that focuses on groups rather than individuals. More random social networks yield higher employment rates than less random ones if the population is integrated or job vacancy information flows are random. However less random social networks allow for better containment of job information inside a group in a segregated population with non-random job information flows, resulting in higher employment rates. We report on the robustness of these findings with respect to the size of minority and majority groups and the amount of social segregation.     

How helpful are social networks in finding a job along the economic cycle? Evidence from immigrants in France

Around 50% of individuals obtain or hear about jobs through social networks. This hiring trend may become problematic when the labor market is tight and people need less social contacts to find a job. Using a one-period static model where network members may receive job offers directly from the firm or indirectly through employed members in the network we show that the share of new hires finding a job through social connections (ie network matching rate) decreases with the job finding rate. Using French data for the period 2003–2012, we test this prediction with immigrants, a population subgroup for whom networks play a major role in occupational decisions. We propose two network matching rate indicators, one based on direct recommendations and another one internalizing the positive externality on the employment probability induced by peers. We find a decreasing relationship between the network matching rate and the job finding rate. Social connections are less helpful for finding jobs during economic expansions.     

Resource availability in the social cloud: An economics perspective

Social cloud has emerged as a case of sharing economy, where socially connected agents share their computing resources within the community. This paper considers the social cloud as an endogenous resource-sharing network, where agents are involved in closeness-based conditional resource sharing. This study focuses on (1) the impact of agents' decisions of link addition and deletion on their own local and global resource availability as well as on others' global resource availability (as spillover effects), (2) the role of agents' closeness in determining spillover effects, (3) agents' link addition behavior, and (4) stability and efficiency of the social cloud. The findings include the following: (i) Agents' decision of link addition (deletion) increases (decreases) their local resource availability. However, these observations do not hold in the case of global resource availability. (ii) In a connected network, agents experience either a positive or a negative spillover effect and there is no case with no spillover effects. Agents observe no spillover effects if and only if the network is disconnected with three or more components. Furthermore, an agent experiences negative spillover if there is no change in its closeness. Although an increase in the closeness of agents is necessary to experience positive spillover effects, the condition is not sufficient. (iii) We study the relation between agents' distance from each other, and their local as well as global resource availabilities. We prove that the local resource availability of an agent from another agent increases with decrease in the distance between them and that maximum local resource availability is obtained from the agent with the least closeness. Using these results, we discuss which agent to add a link to, so as to maximize the local resource availability. We discuss why such results are difficult to establish for global resource availability. However, in a two-diameter network, we show that for an agent, link formation always increases the global resource availability. (iv) We also study resource-sharing network formation and its efficiency in a strategic setting. We prove the existence of a pairwise stable network. Furthermore, we provide a set of conditions for a few prominent network structures (star, complete, wheel, and bipartite networks) to be pairwise stable. We show that the “connected in pairs, otherwise disconnected” network is better than a connected network, in terms of social welfare.     

Peer Effects and Social Networks in Education

We develop a model that shows that, at the Nash equilibrium, the outcome of each individual embedded in a network is proportional to his/her Katz–Bonacich centrality measure. This measure takes into account both direct and indirect friends of each individual, but puts less weight to his/her distant friends. We then bring the model to the data using a very detailed dataset of adolescent friendship networks. We show that, after controlling for observable individual characteristics and unobservable network specific factors, a standard deviation increase in the Katz–Bonacich centrality increases the pupil school performance by more than 7% of one standard deviation.     

To what extent do network effects moderate the relationship between social media propagated news and investors’ perceptions?

As a prominent social media tool, Twitter enables prompt dissemination of financial news and information, which can have a substantial impact on investors’ perceptions and decision-making processes. The propagation of financial news and information through Twitter can either positively or negatively affect investors’ perceptions. As per network theory, the impact of information on one's perception and behavior is known as the network effect. Since Twitter is also a network, we tried to contribute more to this theory in this study by considering other factors that can have an impact on the perceptions of investors. We argue that the impact of financial information and news on investors’ perceptions is moderated by other factors such as connectivity, social ties, and network size of the network. To establish the links between them, we considered three key factors in investors’ networks: (1) network connectivity (network structure); (2) social ties circle (friends, family, colleagues); and (3) size of the network (number of contacts). The results of this study indicate that highly connected investors receive more information and hence, the impact of news is derived from the connectivity of investors within the network. The findings of the study also show that the social ties circle plays a crucial role in determining the impact of the news. The findings further indicate that the impact of news on investors’ perceptions also depends on the theme of the news.     

On the formation of interaction networks in social coordination games

There are many situations where two interacting individuals can benefit from coordinating their actions. We examine the endogenous choice of partners in such social coordination games and the implications for resulting play. We model the interaction pattern as a network where individuals periodically have the discretion to add or sever links to other players. With such endogenous interaction patterns we see multiple stochastically stable states of play, including some that involve play of equilibria in the coordination game that are neither efficient nor risk-dominant. Thus the endogenous network structure not only has implications for the interaction pattern that emerges, but it also has a significant impact on the play in the coordination game relative to what would arise if the same interaction network were exogenous.