DGS Sektion Modellbildung und Simulation Sprecher: Andreas Diekmann 31. Kongress der DGS in Leipzig Oktober 2002

Transcription

1 DGS Sektion Modellbildung und Simulation Sprecher: Andreas Diekmann 31. Kongress der DGS in Leipzig Oktober 2002 Sektionsveranstaltung: Solidarität und soziale Normen Modelle und Mechanismen (Solidarity and Social Norms Models and Mechanisms) Organizer: Thomas Voss Tuesday October 08, 14:15-17:00 hours Michael Hechter* (University of Washington, Seattle): From Class to Culture Guillermina Jasso* (New York University): Justice and Status Mechanisms for Solidarity and Social Norms Michael Macy* / Robb Willer (Cornell University): The Emperor s Dilemma Popular Enforcement of Unpopular Norms Karl-Dieter Opp* (Universität Leipzig): What is is always becoming what ought to be How Political Action Generates a Participation Norm Tuesday October 08, 17:15-18:00 hours Mitgliederversammlung und Vorstandswahlen Wednesday, October 09, 14:15-17:00 hours Andreas Flache* (Universiteit Groningen): When Social Control May Fail to Solve Cooperation Problems. Theoretical Mechanisms and Empirical Evidence Andreas Diekmann (Universität Bern) / Thomas Voss (Universität Leipzig): Social Norms and Reciprocity Uwe Matzat (Universität Düsseldorf): Social Networks and the Emergence of Norms in Internet Groups: A Test of the Coleman-Model Uwe Blien (IAB, Nürnberg): Towards an explanation of social norms in a conceptional multilevel approach Nicole J. Saam (Universität München): Subsidiarity, Decentrality and Innovation Harald Wiese (Universität Leipzig): Globalization from the Point of View of Cooperative Game Theory * Invited speaker Time per presentation will be 25 minutes and 40 minutes for invited speakers (starred*). Presentation time includes time for discussion. 1

2 Uwe Blien Towards an explanation of social norms in a conceptional multilevel approach Subjects do not only act in accordance with standards of expedient rationality but also in accordance with normative criteria. In the present context it is argued that a multilevel structure is necessary to explain norms, as on the one hand these develop spontaneously in social interactions and on the other hand the adoption of standards of superordinate social units is decisive for them. Two key roots are important for an analysis of the development of social norms, historically in the process of human evolution and today in continuous interactions. One of these roots consists in the tendency to behave in social relationships in accordance with the demands of reciprocity, the other consists in the tendency to identify with changing social units in which the subjects participate. If it is tried to map the normative standards of the behaviour on a standard rational choice calculation, then this is possible in formal terms. However, in this case the separation of the objective function and the restrictions is no longer clear. In the paper competing approaches of sociological theory are first dealt with briefly and then an elaboration is given based on experiments used in social psychology, sociology and economics. A classification into the processes of human evolution is done in as far as the development of norms can be understood as an evolutionary stable strategy and can not be equated for instance with the development of altruism. An exposition of an extended formal rationality calculation and an outlook for research conclude the paper. PD Dr. Uwe Blien Institut für Arbeitsmarkt- und Berufsforschung der Bundesanstalt für Arbeit (IAB) D Nürnberg Germany Tel.: Blien Uwe <Uwe.Blien@iab.de> 2

3 Andreas Diekmann/Thomas Voss Social Norms and Reciprocity (Abstract) The paper discusses the argument that reciprocity may be a commitment device that helps to enforce social norms with sanctions in social dilemma situations. Reciprocity is understood in this context as an internalized motive or norm, it does not mean conditional cooperation. Negative reciprocity is the motive to retaliate if a partner has cheated or deviated from a social norm. Rational choice explanations of informal norms have to address the so-called second order problem: Norms can be enforced by (threats of) sanctions, but these sanctions may be costly. Therefore, rational actors may have no incentives to apply such sanctions. Using elementary game theoretic tools, conditions can be derived such that the second order problem may be resolved. However, empirical evidence partially refutes some of the implications of this theoretical analysis. The paper deals with the argument that non-standard assumptions about human motivations (negative reciprocity) may explain that rational actors who are involved in one-shot situations apply sanctions even if there are material costs associated with (actively) punishing norm deviations. In particular, the paper examines whether motivation functions that represent reciprocity may help to justify subgame perfect Nash equilibria that require to conform to a (first order) norm and to a (second order) metanorm even in one-shot situations. Implications for the sociological research agenda on social norms are discussed informally. Prof. Dr. Andreas Diekmann Universität Bern Institut für Soziologie Lerchenweg 36 CH-3000 Bern 9 Switzerland Tel /12 andreas.diekmann@soz.unibe.ch Prof. Dr. Thomas Voss Universität Leipzig Institut für Soziologie Burgstraße 21 D Leipzig Germany Tel.: voss@sozio.uni-leipzig.de 3

4 Andreas Flache When social control may fail to solve cooperation problems. Theoretical mechanisms and empirical evidence. Recent theoretical research suggests that bilateral exchanges of social rewards between group members, like in friendships, may sometimes undermine group solidarity rather than sustain it. Theoretical models identified two different mechanisms that may bring about this counterintuitive effect. A backward-looking social learning model (Flache and Macy 1996) emphasized that bilateral exchanges emerge faster than more complex multilateral co-operation, favoring the emergence of cohesive bilateral ties at the expense of actors willingness to impose peer pressure. A forward-looking rational choice model (Flache, forthcoming (JMS)) highlights the social costs of peer pressure in situations of imperfect information. Here, rational actors may abstain from using peer pressure, because particulalry under imperfect information conditional cooperation may need to inflict costly sanctions that may disrupt privately beneficial social exchanges. This paper summarizes both theoretical arguments and the resulting predicted conditions that explain variation in the strength and effectiveness of social control in task groups. These hypotheses are then submitted to an empirical test in two different semiexperimental studies. A vignette study was conducted under about 200 employees of a dutch university. Furthermore, about 100 business students participated in a questionaire study that accompanied a three weeks period of intensive project work. In these projects, students formed artificial management teams and where they were graded on basis of group performance. The analysis of both data sets indicates that the predicted failure of social control does occur, but only in exceptional cases. Particularly, it can be found that in certain teams of the management game actors who are more popular in the team are at the same time less willing to invest in the group effort. Moreover, the data show that the conditions that explain variation in the effectiveness social control are consistent with both theoretical models, notably salience of group output, shadow of the future, and embeddedness in social network with deviants. Dr. Andreas Flache Rijksuniversiteit Groningen, ICS Grote Rozenstraat 31 NL-9712 TG Groningen Niederlande Tel: a.flache@ppsw.rug.nl 4

5 Michael Hechter From Class to Culture This paper seeks to explain why class politics has receded in advanced industrial societies during the last century, while cultural politics has increased. The novelty of my approach lies in its application of a general theory, the theory of group solidarity, to a bewilderingly complex historical problem. The leading explanations in the literature attribute the cause of this shift to structural changes -- in particular, to the growth of the middle class at the expense of the working class. In contrast, I contend that the principal cause is the shift from class-based organizations to organizations whose members principally share cultural affinities. This shift in patterns of organizational affiliation is strongly affected by the replacement of direct for indirect rule in multicultural societies. Whereas indirect rule tends to promote class-based organization, direct rule favors culturally-based organization. Whereas the expansion of direct rule since the 1950s - in the form of the welfare state -- resulted in the muting of class politics, at the same time it has increased the salience of cultural politics. Prof. Dr. Michael Hechter University of Washington Department of Sociology Box Seattle, Washington USA Tel: hechter@u.washington.edu 5

6 Guillermina Jasso Justice and Status Mechanisms for Solidarity and Social Norms Two basic processes observed in human behavior and social phenomena are the sense of justice and the sense of status. Singly or in combination they have a long reach, showing their effects across vast and disparate areas of social life, from family phenomena to criminal behavior, from religious institutions to international migration, from health to norms. The two processes have a similar structure. In both, the process begins with individuals reflecting on their own or others personal quantitative characteristics (such as attractiveness, bravery, income, or wealth), thus generating a justice evaluation, in the case of justice, and a status score, in the case of status, and thence a myriad of behaviors, often involving the subgroups formed by the categories of qualitative characteristics (such as race, religion, or language). However, the two processes operate differently according to current mathematical descriptions, the mathematical function that generates the justice evaluation from the quantitative characteristics differs in important ways from the mathematical function that generates the status score from the quantitative characteristics. Hence, new factors arise which can be used to explain observed phenomena, in particular: (1) whether individuals and societies are concerned with justice or with status or with a weighted combination of the two, (2) whether the valued quantitative characteristics are of the cardinal kind (such as wealth and other material possessions) or the ordinal kind (such as beauty or intelligence), and (3) the number of available qualitative characteristics and their number of categories (such as gender, nativity, and citizenship, each with two categories, and race, religion, and language, each with multiple categories) which can be used to form subgroups. Different combinations of these ingredients generate different outcomes. In this paper we use the justice and status framework to derive predictions for social solidarity and social norms. The predictions are ceteris paribus and fully testable. As predictions, they signal the fundamental importance of the Weltanschauung for example, a society that values cardinal goods and is concerned with justice will differ markedly from a society that values ordinal goods and is concerned with status as well as of the numerical proportions in the categories of qualitative characteristics and the inequality in the distributions of cardinal goods (the twin Blau factors of inequality and heterogeneity). More importantly, they raise the challenge of rigorous testing, so that the strength of the operation of justice and status mechanisms and their scope can be empirically assessed. Such testing will not only improve our understanding of (1) justice and status mechanisms and (2) solidarity and norms, but will also pave the way for discovery of other more fundamental principles generating observed phenomena. Prof. Dr. Guillermina Jasso Department of Sociology New York University 269 Mercer Street, 4th Floor New York, NY Tel: gj1@nyu.edu 6

7 Michael Macy/Robb Willer The Emperor s Dilemma: Popular Enforcement of Unpopular Norms In functionalist accounts, norms are useful to society at large, while choice theory posits utility to those who enforce them. The Emperor s Dilemma uses game theory and agent-based modeling to explore an alternative possibility that members of a group may enforce obligations to act in ways that no one actually wants or needs. In this game, based on the Andersen fable, everyone privately wants to laugh at the Emperor but fears being criticized by the others as unfit for office. The explanatory problem is not why people conform to an unpopular norm that is easy, they fear social criticism. The question is why anyone who cannot see the Emperor s clothes would enforce this norm in the first place? Familiar examples of self-enforcing norms range from the comic to the tragic: Fawning admiration for a highly prestigious but incomprehensible scholar (whose brilliant new ideas cannot be seen by intellectual lightweights). Teenagers who seek peer acceptance by celebrating self-destructive behaviors and ridiculing posers (whose conformity is insincere). The politically incorrect whose exposure affirms the moral standing of those who are sufficiently indignant. Game theory proves that enforcement of an unpopular norm can be a subgame perfect equilibrium (hence the threat of sanctions is credible) but cannot explain how a population falls into the trap or how the house of cards collapses. A stochastic threshold model shows how crusades can be triggered by a random walk, even when no one actually believes in the cause. Once established, the crusade eventually collapses by the same mechanism. Professor Michael Macy Ph.D. Cornell University Department of Sociology 358 Uris Hall Ithaca, NY U.S.A. mwm14@cornell.edu 7

8 Uwe Matzat Social Networks and the Emergence of Norms in Internet Groups: A Test of the Coleman-Model While early research on Internet groups suggested that norms would have only a restricted impact on the behavior of members in computer mediated communication (e.g., Kiesler, Siegel & McGuire 1984), newer research corrects this impression (e.g., Postmes, Spears & Lea 1998). This paper applies the Coleman model (Coleman 1990: Chapter 10 & 11) to explain the emergence of a norm in academic discussion groups of the Internet. According to Coleman (1990) under the following conditions the evolution of a group norm that proscribes a certain form of behavior is expected: a) the behavior has positive effects on other group members b) it cannot be enforced by bilateral social exchange c) a high degree of closure of the members social networks facilitates social control. The paper uses the ideas to explain under which conditions a group norm emerges in academic discussion groups, so-called mailing lists. It analyses under which conditions a group norm develops that proscribes the provision of help and information in group discussions. More precisely, under which conditions are academic members of Internet Discussion Groups expected to answer questions of other group members? It tests whether the embeddedness of academic Internet Discussion Groups in offline social networks facilitates the norm emergence in online discussions. Questionnaire data of active and passive members of about 50 international academic mailing lists combined with archived data of their online communication behavior are used to test this idea. Two results of the multivariate data analysis are most remarkable. Firstly, the higher the degree of embeddedness the stronger is the perceived norm in the discussion group. Secondly, the stronger the norm in the discussion group the more often the members answer questions during the group discussions. Uwe Matzat University of Düsseldorf Institute of Social Sciences Universitätsstraße Düsseldorf Germany Tel.: matzat@uni-duesseldorf.de 8

9 Karl-Dieter Opp What is is always becoming what ought to be. How Political Action Generates a Participation Norm 1 There are different processes that lead to the emergence of norms. One process begins with a behavioral regularity. It is often claimed that if such a behavioral regularity has been established the respective behavior becomes a norm. In other words, what is becomes what ought to be. This paper addresses the question whether this proposition is a general law or whether it holds only under certain conditions; if so, what are those conditions? To answer these questions the paper focuses on a particular kind of behavior, namely political action. The question is under what conditions political action generates a norm to participate. In a first part, some propositions are suggested. In a second part, these propositions are tested with a two-wave panel study in which the same individuals were interviewed 1993 and Prof. Dr. Karl-Dieter Opp Universität Leipzig Institut für Soziologie Burgstraße 21 D Leipzig Germany Tel.: opp@sozio.uni-leipzig.de 1 The quotation is from Homans 1974, p

10 Nicole J. Saam Subsidiarity, Decentrality and Innovation The Edinburgh European Council of December 1992 defined the basic principles of the concept of subsidiarity and laid down the guidelines for interpreting Article 3b, which entrenches subsidiarity in the Treaty on European Union. Its conclusions were set out in a declaration which still serves as the cornerstone of the subsidiarity principle. The norm of subsidiarity also holds in German federalism, e.g. in the field of social politics. The subsidiarity principle is intended to ensure that decisions are taken as closely as possible to the citizen and that constant checks are made as to whether action at Community level is justified in the light of the possibilities available at national, regional or local level. Specifically, it is the principle whereby the Union does not take action (except in the areas which fall within its exclusive competence) unless it is more effective than action taken at national, regional or local level. It is closely bound up with the principles of proportionality and necessity, which require that any action by the Union should not go beyond what is necessary to achieve the objectives of the Treaty. As a consequence the subsidiarity principle fosters decentralization. This lecture will introduce a multi-level model which simulates the relation between different degrees of decentrality in political systems and the capacity of innovation of these systems. It is based on evolutionary market process and competition theory (Schumpeter, Hayek) and on the New Economics of Jurisdictional Competition (Oates/Schwab 1988, Frey/Eichenberger 1995, Bratton/McCahery 1997). The latter approach transfers the market paradigm to public services. Competition between territorial units is interpreted as competition between jurisdictions. The central hypotheses which the simulation investigates is that the average growth rate of the knowledge accumulation process in judicial competition increases with the number of independently innovating jurisdictions. PD Dr. Nicole J. Saam Ludwig-Maximilians-Universität München Institut für Soziologie Konradstraße 6 D München Germany Tel: nicole.saam@soziologie.uni-muenchen.de 10

11 Harald Wiese Globalization from the point of view of cooperative game theory Building on the Shapley value, several values for coalition structures have been presented in the literature, most notably by Aumann and Dreze (1974) and by Owen (1977). They capture ''social structure'' in some sense or other. We find that they can be amended so as to model different aspects of globalization. Thus, the paper has two aims: First, defining globalization with the help of cooperative game theory. Second, generating hypotheses about winners and loosers of globalization. Long abstract: Globalization is an often-used buzzword in public discourse. It has a negative connotation for many people. The attac movement, for example, organizes protests against the negative impacts, as they see it, of globalization. However, the meaning of globalization is far from clear. In this paper, we try to use cooperative game theory in order to elucidate the meaning of globalization and in order to develop some theory about who stands to win or to gain from globalization. We find that partitional values that build on the Shapley value can be of use for that purpose. The Shapley value itself is not a partitional value. It can be used to assess the payoffs for players in a globalized world. For that purpose, one needs a coalition function v that describes the economic possibilities open to various groups of people, called coalitions. Formally, the coalition function assigns a ''worth'' v(k) to any subset K of the set of individuals (players) N. The Shapley value, then, is a function that maps coalition functions into payoffs for all players. The general idea is to devide the worth v(n) according to some average of the ''marginal contributions''. However, in a world of trade barriers, illiberal migration policies, and high transaction costs, some gains from trade cannot be realized. Then the sum of the players' payoffs is less than v(n). One simple way to model the world's segmentation is by way of partitions. The sets making up the partition are called components. For example, Aumann and Dreze (1974) propose a partitional value that assumes component efficiency (the players in a component share the worth of this component). One might then interpret these components as countries or blocks of countries that do not (or hardly) interact with outsiders. In this framework, the EU enlargement with respect to countries such as the Czech Republic or Poland can be viewed and modeled as an enlargement of the component hosting the individuals living in the former EU member states. Of course, the existence and the effects of components need not be binary: either the components exist or do not exist. For example, trade may largely be confined to a component but some trade will nevertheless occur beyond component boundaries. Western Germany had a low volume of trade with communist Albania but it was not zero. Another related, but different aspect concerns outside options. For example, the people in the former Republic of Yugoslavia were free to travel and work abroad while the communist GDR did not allow people to seek work in other countries. The outside-options for GDR inhabitants were very small and therefore, medicins could be paid a salary minimally above a worker's salary. For the AD-value, the payoffs for a component's players are dictated by the marginal contributions of players within this component. This implies that differing outside options of players do not bear on the pay-off. In this paper, we propose a value that introduces outside options into the AD-value. According to the above examples, we define globalization as follows: First, globalization is the process of enlarging components (EU enlargement). Second, gloablization occurs when the volume of trade decreases within small components and increases within larger components (trade liberalisation). Third, globalization can be said to happen whenever the outside options of individuals within a component become more important (GDR versus Yugoslavia). 11

12 From the point of view of the attac movement, globalization has to be countered by government interference and globalized unions. In this paper, we will model the effect of unions that spread across national borders. Unionization can be modeled with the help of the Owen value (see Owen (1977) and Hart and Kurz (1983)). We construct a concoction of these values and make use of two partitions: First, the AD-partition capturing trading blocks. Second, the Owen partition capturing unionization. This value is, of course, rather complicated. An axiomatic treatment is not the main focus of the paper and is given for the simpler values, only. Instead, the paper concentrates on the application to the problem of globalization. Prof. Dr. Harald Wiese Universität Leipzig Wirtschaftswissenschaftliche Fakultät Jahnallee 59 D Leipzig Germany Tel: Harald Wiese wiese@wifa.uni-leipzig.de 12