Seminars: 2013 Winter / Spring

Seminar series of the MTA TK „LENDÜLET” Research Center for Educational and Network Studies (RECENS)

Winter/Spring 2013

 

January 8, 2013

Nikita Krugljakov (Maastricht University): Social Closure and Social Norms

Coleman (1988) proposes that social structure is important to effectively internalize external effects. In a lab experiment we study whether social closure is sufficient to discipline subjects not to take money from others. Shareholder A chooses how much money to take from stakeholders B1 and B2. The stakeholders observe the amount taken. In the closure treatment stakeholders communicate with each other. In the noclosure treatment each stakeholder communicates with a bystander who is not affected by A's choice. After communication each stakeholders individually invests in punishing the shareholder. Punishment is costly. Our main result suggests that shareholders take more money from stakeholders in the closure treatment than in the noclosure treatment. This observation contradicts Coleman’s hypothesis. To test the robustness of our results we conduct a follow up experiment. We allow stakeholders to observe each other’s punishment choices. Furthermore, we look at the effect of decreased punishment costs.

 

 

January 15, 2013

Simone Righi (MTA TK „Lendület” Research Center for Educational and Network Studies (RECENS)): Essays in Information Aggregation

I will discuss my work in the field of network theory and its applications to Economics and Social Sciences. I will show how several properties of empirical social networks and financial markets can be reproduced as emerging properties of relatively simple theoretical models. The main focus will be on socio-economic systems characterized by the presence of many heterogeneous agents, each endowed with a piece of information (an opinion, a belief or a preference), which can be shared with some of the other members of a population (in a distributed fashion). To obtain a global understanding of the problem studied, different aggregation techniques will be used.

 

 

January 22, 2013

Zoltán Kmetty (Eötvös Lóránd University, Faculty of Social Sciences, Research Center for Social Methodology): Structural Rifts – A New, Network Based Method for the Exploration of Social Cleavages

Co-author(s): Júlia Koltai

From a lots of points of views, the research of networks opened new perspectives on the field of social structures. The homophiles that are observed in social networks and the segregations, which are presented in social structures are equivalent processes on the micro- and macro level. The name-generator technique, which is a current method in network research, is primarily able to detect the close system of relationships, so hereby it enables the exploration of the narrower, micro-environmental mechanisms of effects. The position-generator or rather the resource-generator methods primarily assign the people’s instrumental action-zones and only restrictedly appropriate for the definition of social cleavages – although they examine the networks of the individuals in a wider context. (Nevertheless it should be noted that by particular aspects, some of the extension of these methods enable the determination of the social cleavages.) In 2006 a resource-generator-like method was applied in the General Social Survey (GSS) for the measurement of the size of the personal network and social embeddedness (DiPrete et al. 2011). The funds of the method were laid down by McCarty et al. (McCarty 2001) at the beginning of the 2000s. In the GSS questions were asked about how many persons does the respondent know who’s name is XY (for example „How many of the people you are acquainted with are named Michael?”). Based on the different populational frequency of male and female names, it is possible to estimate the network-size of the respondent. However, the size of the network does not necessarily mean useful information per se, as it does not show indication for the expedience of the network. The curiosity of this method is that it was also asked that how many persons do the respondents know and trust in different social groups (for example African Americans, white people, homosexuals, who attend religious services on a regular basis, etc.). With the help of these questions it is possible to examine how strong are the segregational relations between the different social groups and how do they strengthen or weaken in the case of trusted relationships. Therefore, the method does not concentrate on the instrumental and expressive help-network – which is useful for the ego –, but the ties and segregational borders between social stratums, classes and milieus. In 2012 a research was made in Budapest, where questions, which were based on the American GSS network block, but implemented to the Hungarian situation, were asked. We will present the possibilities provided by the method and also the limits of these; furthermore we try to draw those cleavages that divide the people of Budapest the most.

 

 

February 12, 2013

Ádám Szeidl (Central European University, Department of Economics): Treasure Hunt: Social Learning in the Field

Co-author(s): Markus Mobius, Tuan Phan

We seed individuals in a real-world social network with information about quiz questions to experimentally measure the effectiveness of social learning. Using data on both the pre-existing social network and the actual conversation network, we find strong evidence that people learn from direct and indirect friends, but also that—unlike in standard models—information transmission is imperfect. We then compare two theories of social learning: a DeGroot-style model in which people double-count signals that reach them through multiple paths, and a “Streams” model in which people tag the source of information and hence do not double-count. In the conversation network, the weight a decision maker attaches to the signal of a person increases in the number of paths between them when the person is an indirect friend, but not when she is a direct friend. This fact is consistent with the Streams model in which multiple paths only increase the weight by increasing the transmission probability when—like with an indirect informer— transmission is imperfect. Structural and reduced-form estimates which exploit this and other variation generally support the Streams model combined with probabilistic transmission in our setting.

 

 

February 19, 2013

Szabolcs Számadó (Eötvös Loránd University, Department of Plant Systematics, Ecology and Theoretical Biology): Pre-Hunt Communication Provides Context for the Evolution of Early Human Language

The origin of human language is one of the most fascinating and most difficult problems of evolution. Here I argue that pre-hunt communication was the starting context of the evolution of human language. Hunting of big game created a shared interest; animals and hunting actions are easy to imitate; the need to plan created a pressure for increasing complexity; and finally, cultural inheritance of hunting tools and know-how made the transition unique. I further argue that this “first step” was actually a two-stage process where first indexical and iconic signs evolved to coordinate recruitment for the hunt; then later, in the second stage, the complexity of this communication system increased as a response to the increased demand to coordinate group-hunting effort (including division of labor). I provide a review of the fossil record and show that the available evidence is fully compatible with the theory.

 

 

February 26, 2013

André Grow (University of Groningen/ICS): Regional Variation in Status Values-An Explanation on Status Construction Theory

Co-author(s): Andreas Flache, Rafael P. M. Wittek

Why do the status values of social distinctions often vary widely across geographic regions? For example, why are the elderly more respected than the young in some parts of the world, but are less respected in others? Social scientists have hitherto mostly focused on regional variation in sociocultural factors that create power and resource differences between social groups, which ultimately lead to status differences between them. Combining status construction theory with insights from social network research, in this paper we explore a new mechanism that can account for regional variation in status values, even in the absence of variation in sociocultural factors that much earlier research has focused on. Status construction theory, on the one hand, holds that goal directed interactions among members of different categories (e.g. between men and women) can induce status beliefs among them that imbue the social distinction with status value. Social network research, on the other hand, suggests that goal directed interactions are often spatially clustered, so that they occur more often within geographic regions than between geographic regions. We argue that the constraints that spatial clustering of interactions imposes on the behavioral processes that status construction theory describes can lead to regional variation in status values.

The mechanism that we describe is based on a complex interplay of interaction and influence processes among a large number of autonomous individuals. The outcomes of such complex interaction systems are difficult to assess by mere verbal reasoning. We therefore develop an agent-based computational model that enables us to rigorously assess the logical consistency of our theoretical argument. Systematic computational experimentation with our model suggests that spatial clustering of goal directed interactions can indeed account for regional variation in status values.

 

 

March 5, 2013

Tamás Keller (TÁRKI, Social Research Institute)

 

 

March 12, 2013

Elias Carroni (University of Namur): Eliciting Local Network Externalities through Price Discrimination

Co-author(s): Simone Righi

Nowadays, consumers' preferences are more and more sophisticated and it is difficult for producers of new goods to efficiently capture their consumers' idiosyncratic willingness to pay. However, it is recognized that consumers are embedded in structured social interactions. Exploiting the networked structure of society, producers may find it profitable to offer discounted prices to those consumers who convince other people to buy. In presence of search costs and heterogeneous willingness to pay, these network-activation based reward programs may serve to expand the client base, through the local flow of information between informed and uninformed consumers. In this paper, we propose a simple model to analyze the incentives for a monopolist to propose such program. We observe the emergence of a trade-off between the size of the customer base reached and the price charged to uninformed consumers. The balance in this trade-off is driven by the degree distribution in the consumers' social network.Fixing a network density, the lower the average degree (i.e. the more positively skewed the degree distribution), the more profitable is to run the program.

We then endogenize the market structure in the context of an incumbent-entrant model. We find that offering a network-activation based discount program can be a tool for the incumbent to prevent entry.

 

 

March 19, 2013

Jan Kratzer (TU Berlin): The Social Network Positions of Lead Users and Opinion Leaders

 

 

April 9, 2013

Vera Messing (Institute for Sociology, Centre for Social Sciences, Hungarian Academy of Sciences): The Consequence of Ethnic Segregation: Peer Group and Student – Adult Relationships in Multiethnic Schools

The EDUMIGROM research project investigated future prospects of 14-17 years old youth attending ethnically diverse schools in 9 countries of the European Union by using both quantitative and in-depth qualitative methods. In the old member-states, second generation migrant students were focused on, while in countries of Central East Europe Roma /Gypsy youth participated in the research. In the course of the research there was a special attention attributed to ethno-social segregation. The proposed presentation first demonstrates mechanisms leading to various forms of ethno-social segregation in the 9 countries and then describes factors which have an important influence on students’ everyday life beyond academic performance. The second part of the presentation will focus on how the various forms of segregation have an influence on students’ ties to their peers, on their relationship to adults in and outside the school. It will describe what role an ethnically diverse teacher body may play as well as how various teaching methods influence the general atmosphere and performance in ethnically and socially diverse schools.

 

 

April 30, 2013

Marcel A.L.M Van Assen (Tilburg University)

co-author: Jacob Dijkstra (ICS / University of Groningen)

We propose a game theoretical model of one-shot network public goods formalizing the ‘closure argument’ that cooperation is more frequent in denser groups or networks. Equilibrium analyses show that (i) an ‘inefficiency problem’ exists: players all preferring mutual cooperation need not all cooperate; (ii) in dyads, groups and networks with degree independence, first order stochastic dominance shifts of the distribution of cooperation preferences or the degree distribution (weakly) increases cooperation, and (iii) the latter result does not hold for networks with degree dependence. Hence the closure argument always holds in networks satisfying degree independence but not in other networks. 

 

 

May 2, 2013

Flaminio Squazzoni (University of Brescia): When Competition is Pushed too Hard

Peer review is a cornerstone of science, whose quality and efficiency depends on a complex, large-scale collaboration process, which is sensitive to motivations of scientists. Recent findings on typical failures of peer review, due to judgment bias, parochialism, barriers against multi-disciplinarity and innovation, and cases of misconduct have contributed to calls for a reconsideration of the rigour and reliability of the process. This has become even more problematic recently as increasing pressures towards competition implied that scientists harshly compete for funds and reputation at a national and international level.In this talk, I shall examine the impact of strategic behaviour of referees on the quality and efficiency of peer review by presenting an agent-based model that simulates the peer review process. We modelled a population of N scientists (N = 200) randomly selected each to play one of two roles: author or referee. The task of authors was to submit an article with the goal of having it accepted to be published. The task of referees was to evaluate the quality of author submissions. We assumed the presence of knowledge asymmetry so that referee opinion was potentially subject to evaluation bias. We built two simulation scenarios to investigate large-scale implications of certain referee behaviours. The first one was inspired by “the luck of the reviewer draw” idea. In this case, we assumed that referees randomly fell into Type I and Type II errors, i.e., recommending submissions of low quality to be published or recommending against the publishing of submissions which should have been published. In the second scenario, we assumed that certain referees tried intentionally to outperform potential competitors by underrating the value of their submissions. Results showed that when publication selection increased, the presence of a minority of cheaters may dramatically undermine the quality and efficiency of peer review even compared with a scenario purely dominated by “the luck of the reviewer draw”. We also found that peer review outcomes are significantly influenced by differences in the way scientists identify potential competitors in the system. Certain mechanisms, such as the stratification of scientists in local competing groups and the presence of niches of competition, might reduce the negative effect of cheating on the quality of the peer review process as scientists can develop more competent judgment. 

 

 

June 18, 2013

Christina Prell (University of Maryland)