6-8. WTFNS

• Workshop on the Frontiers of Network Science is an opportunity for our group to bond over research - some that is more early stage, and some that is better developed. It also allows us to bring leading scholars from the field of Network Science who bring their work to NYUAD. The keynote speakers' talks are open to the NYUAD community.

3. From defection to in-group favoritism to cooperation: Simulation analysis of the social dilemma in dynamic networks (Hirofumi Takesue) - 9:00AM-10:15AM, A5-101

• Abstract: The emergence of cooperation in social dilemmas is a core question in the social sciences. The existence of a conditional strategy, wherein individuals only cooperate with the members of their own group (in-group favoritism) was proposed as a solution. However, the empirical literature has suggested that in-group favoritism prevents the realization of potentially profitable interactions with outgroup members. This observation calls for a theoretical analysis that would help understand at what point in-group favoritism and unconditional cooperation become prevalent in societies. Here, we conducted computational simulation in which agents located in social networks imitate traits of successful neighbors or sever social ties with defective neighbors and connect them to other agents. The results of our simulation showed transitions from unconditional defection to in-group favoritism to unconditional cooperation as link-rewiring occurred more frequently. This indicates the usefulness of the dynamic networks framework for understanding the emergence of cooperation and in-group favoritism.

17. Only the Best People: Loyalty, Competence, and Connections in Russian Elite Networks (Noah Buckley) - 9:00AM-10:15AM, A5-101

• Abstract: Autocratic leaders face a well-known tradeoff between the loyalty and the competence of their subordinates (Egorov and Sonin 2011; Zakharov 2016). However, personal connections between those subordinates can also comprise a threat to loyalty to the dictator. In this paper, I develop a theory that incorporates personal connections into the loyalty-competence dilemma. I show that regime members who have high levels of competence and good connections are very threatening to loyalty, as they can more easily coordinate against the dictator, are well-informed, and may serve gatekeeping roles. Regime members who are connected but have lower levels of competence are, in contrast, easily co-opted into the regime and are unable to select winners in a rebellion. I test this theory using comprehensive social network data on all Russian federal-level political elites from 2000-2012. Incorporating measures of competence and network centrality into panel regressions and temporal ERGM social network analysis, I demonstrate the risks and rewards that personal connections can generate in autocracies. These findings contribute to literatures on coup-proofing and the roles of institutions in autocratic survival and authoritarian power-sharing (Svolik 2012).

31. Opening the black box of weak ties: Information asymmetries and network-based job search (Thibaud Deguilhem (University of Bordeaux), Santiago Gomez (University of Heidelberg), Javier Mejia, Eric Quintane (Los Andes University)) - 9:00AM-10:15AM, A5-101

• Abstract: This paper explores the role of tie strength in the job-search process. We argue that the existing work on the field lacks clarity regarding the fact that different aspects of tie strength contribute to solving different information problems during the job-search process. We offer a theoretical framework that allows disentangling this problem, focusing on the differential effects of frequency and trust. High-frequency ties are better solving incomplete information problems, while high-trust ties are better solving information asymmetry problems. Depending on which type of problem dominates, some type of ties would be more or less useful. To test our theory, we collected a new dataset of individuals that found their current job using their social network. As our theory predicts, we find that, on average, high-frequency ties are associated with better outcomes in the job search, while high-trust ties are associated with worse outcomes. However, the signs of this correlations switch when considering individuals changing occupations, where information asymmetries dominate incomplete information problems.

17. It Takes a Village: Peer Effects and Externalities in Technology Adoption (Romain Ferrali, Guy Grossman (University of Pennsylvania), Melina Platas, Jonathan Rodden (Stanford University)) - 9:00AM-10:15AM, A5-101

• Abstract: Do social networks matter for the adoption of new political communication technologies? We collect complete social network data for sixteen Ugandan villages where an innovative reporting mobile platform was introduced, and show robust evidence of peer effects on technology adoption. However, peer effects were not observed in all networks. We develop a formal model showing that while peer effects facilitate adoption of technologies with minimal externalities (like agricultural practices), it can be more difficult for innovations with significant positive externalities to spread through a network. Early adopters might exaggerate benefits, leading others to discount information about the technology's value. Thus, peer effects are likely to emerge only where informal institutions support truthful communication. We show that the observable implications of our model are borne out in the data. These impediments to social diffusion might help explain the slow and varied uptake of new political communication technologies around the world.

3. Diversity, Homophily, and Scientific Impact - the Preeminence of Ethnicity over Age, Gender, Discipline, and Affiliation (Bedoor AlShebli, Talal Rahwan, Wei Woon - Masdar Institute/Khalifa University) - 9:00AM-10:15:AM, A5-101, co-sponsored with the Identity Politics working group

• Abstract: Inspired by the social and economic benefits of diversity, we analyze over 9 million papers and 6 million scientists to study the relationship between research impact and five classes of diversity: (i) ethnicity, (ii) discipline, (iii) gender, (iv) affiliation, and (v) academic age. Using randomized baseline models, we establish the presence of homophily in ethnicity, gender and affiliation. We then study the effect of diversity on scientific impact, as reflected in citations. Remarkably, of the classes considered, ethnic diversity had the strongest correlation with scientific impact. To further isolate the effects of ethnic diversity, we used randomized baseline models and again found a clear link between diversity and impact. To further support these findings, we use coarsened exact matching to compare the scientific impact of ethnically diverse papers and scientists with closely-matched control groups. Here, we find that ethnic diversity resulted in an impact gain of 10.63% for papers, and 47.67% for scientists.

9. Web of lies - Lying in social networks, an experimental approach (Kinga Makovi & Manu Munoz) - 9:00AM-10:15AM, A5-101

4. The co-evolution of professional and market networks in global art worlds (Kangsan Lee) - 9:00AM-10:15AM, A5-101

18. Web of lies - Lying in social networks, an experimental approach (Kinga Makovi & Manu Munoz) - 9:00AM-10:15AM, A5-101

25. Bad apples in a barrel: Field evidence about corruption in organizations (Romain Ferrali) - 8:30AM-9:30AM, A5-101

• Abstract: Organizations matter for corruption because they structure social relationships that may either deter corruption through increased monitoring or foster it by giving access to additional accomplices. I introduce a model that shows that which of these two functions prevails has important implications for organizational design, for it determines whether maximally or minimally connected organizations are optimal. Answering this question requires field measurements of social relationships and fraudulent behavior of individual members. I use a fine-grained, original dataset on the daily operations of a large company that allows measuring, over time, dishonest behavior at the individual level and interactions among employees. I identify agents that are likely to be dishonest. A series of natural experiments show that honest agents reinforce their behavior after being exposed to corrupt individuals, who in turn tend to become more honest. This suggests it would be sensible to redesign government agencies to reduce isolation among bureaucrats.

7. Diversity, coordination, and spillovers - Presentation of experimental design (Yan Long & Manu Munoz) - 9:00AM-10:15AM, A5-101

• Abstract: Work group diversity has great influences on group processes and performance. However, much is still unknown about the effects of diversity. In our experiments, we study diversity in preferences over group outcomes, which reflects different values and different goals among group members. We hypothesize its effects on successful coordination and the spillovers it may have on subsequent group tasks: specifically, the higher the degree of diversity in the group, the more difficult it is to successfully coordinate; however, if successful coordination is achieved, the more severe coordination conflicts are overcome, the more likely people are to trust and appreciate each other. This results in larger positive spillovers in subsequent interactions. We design a two-part experiment to test our hypotheses.

14. Do men and women reap different benefits from networking on the labor market? (Malte Reichelt & Kinga Makovi) - 9:00AM-10:15AM, A5-101

24. The network formation origin of tribal societies (Javier Mejía) - 9:00AM-10:15AM, A5-101

• Abstract: This paper proposes a network formation model for explaining the stability of tribal societies. The model is supported by the idea that every two members of a tribe should have benefited from being connected to each other in order for the whole tribe to be stable. It also considers the constraints that the ecosystem brought to social interaction in pre-modern contexts. The model has three predictions. First, both homogeneous and heterogeneous tribes could have been stable regardless of technological development. Second, the social complexity of tribes was a function of technological development (having access to agriculture should have enabled the emergence of larger and more complex societies), interaction costs (if they were too low or too high, no complex society should have emerged), and environmental conditions (poor ecosystems should not have allowed the formation of complex societies). Finally, the model predicts that collapses of agricultural societies could not come from environmental pressures, but from high interaction costs. The predictions are consistent with some of the most relevant human history patterns.