Running Head: LEADER ENDORSEMENT IN SOCIAL DILEMMAS Leader Endorsement in Social Dilemmas: Comparing the Instrumental and Relational Perspectives

This chapter examines the role of leadership in overcoming social dilemmas within groups. First, based on prior theorising and research we present two alternative perspectives on leader endorsement in social dilemmas, an instrumental and relational perspective. Next, we systematically compare these perspectives in a series of experiments investigating leadership in social dilemmas created within small groups in the laboratory. The results of our studies suggest that when their personal identity is salient group members more strongly endorse leaders who are perceived to be instrumental in solving the free-rider problem. In contrast, when a social identity is salient members more strongly endorse leaders who fulfil their relational needs. Based on these findings we propose a differential needs model of leader endorsement in social dilemmas. 2 "Leadership is getting someone to do what they don't want to do, to achieve something that they want to achieve" --Tom Landry (legendary American football coach) Introduction Leadership is commonly defined as a process of influence to attain important group, organisational and societal goals (Bass, 1990; Chemers, 2001; Haslam, 2001, Hollander, 1985, Yukl, 1989). In order to achieve these goals leaders must ensure that conflicts, which frequently emerge between the self-interest of individual members and the collective interest of the group, are resolved. Such conflicts are better known as social dilemmas (Dawes, 1980; Komorita & Parks, 1994; Wilke, 1991). Good leadership is necessary in order to manage these dilemmas, which occur at every level of society. In work teams, for example, an important role of the manager is to ensure that all team members contribute towards the completion of a group task. In sports, a team manager must be able to motivate players to put in their best performance to beat other teams. On a larger scale, political leaders and authorities must secure the cooperation of citizens for the maintenance of important public goods, such as schools and hospitals, and natural resources, such as land and water (Van Vugt, Snyder, Tyler, & Biel, 2000). Without some form of leadership many groups and organisations would not be able to deal adequately with social dilemmas, because in the absence of leaders and authorities freeriding in groups would be too widespread (Olson, 1965; Yamagishi, 1986). Consistent with this view, B. M. Bass, a leadership theorist, recently stated that successful leadership involves “the moving of followers beyond their selfinterests for the good of the group, organisation, or society” (Bass, 1997; p. 130). Appointing a leader seems a sensible solution to deal with social dilemmas in groups, but it is not a straightforward solution. Leadership is a complex political process involving continuous negotiations between group members about what kind 3 of leaders they desire. For example, groups must decide where the leader should come from (e.g., from inside or outside the group), how they should be assigned to the group (e.g., by election or appointment), what power base they should have (e.g., reward, coercive or legitimate power), what should be their leadership style (e.g., task or relation-oriented) and personal attributes (e.g., a highly skilled versus highly committed leader), and, finally, how they can be replaced if necessary (Bass, 1990, French & Raven, 1959; Hollander, 1985; Levine & Moreland, 1998; Yukl, 1989). Furthermore, once leaders are in place, group members must decide whether to cooperate with their directives (Lippitt, & White, 1968; Tyler & Degoey, 1995; Van Vugt & De Cremer, 1999). In this dynamic process, group members will presumably not be focused solely on the perceived instrumentality of leadership in resolving the social dilemma at hand. They might also consider the consequences of having a leader for the social climate in the group and the quality of interpersonal relationships between group members (cf. goal achievement vs. group maintenance; Cartwright & Zander, 1968). It is important to distinguish between two kinds of solutions to social dilemmas, individual and structural solutions (Messick & Brewer, 1983; Van Vugt et al., 2000). Individual solutions involve group members’ voluntary efforts to produce valuable goods for their group. The vast majority of social dilemma research has been devoted to studying social-psychological determinants of voluntary cooperation, such as communication, trust, and the development of prosocial norms (for overviews, see Komorita & Parks, 1994; Liebrand, Messick, & Wilke, 1992; Schroeder, 1995). Over the past two decades, however, researchers have become interested in the emergence of structural solutions to social dilemmas, the main question being when groups opt for a change in the structural features of groups in order to resolve the social dilemma. Structural solutions that have been empirically investigated include the introduction of contribution and distribution rules (Sato, 1987; Van de Kragt, Orbell & Dawes, 1983), reward and punishment systems (Komorita, Parks, & Hulbert, 1992; Yamagishi, 1986), systems of exclusion (Kerr, 1999), formal authorities (Tyler & DeGoey, 1995), 4 and leadership (Foddy & Crettenden, 1994; Foddy & Hogg, 1999; Messick et al., 1983; Rutte & Wilke, 1984; Samuelson et al., 1984; Van Vugt & De Cremer, 1999; Wilke, 1991; Wit, Wilke, & Van Dijk, 1989). In this chapter, we concentrate on the role of leadership because it is potentially the most viable solution to social dilemmas within small groups (Levine & Moreland, 1998; Messick & Brewer, 1983; Van Vugt & De Cremer, 1999). We investigate two interrelated aspects of leadership in social dilemmas, leadership emergence and leadership influence. First, we examine why group members want to voluntarily assign a leader to the group and what type of leader they choose. Second, we study the influence of different leader types on the voluntary cooperation of members. Together we refer to the voluntary acceptance and cooperation with leaders as leader endorsement. Our main research aim is to demonstrate that the endorsement of leadership in social dilemmas is a function of the prevailing needs within the group. If members are concerned primarily about their short-term personal welfare, they will endorse leaders who are believed to be instrumental in solving the social dilemma task. In contrast, if members assign greater priority to the long-term group welfare, they will endorse leaders whose primary goal is to strengthen the relationships between group members. In this regard, our research is inspired by the traditional leadership literature, which o draws a distinction between the instrumental and relational roles of leadership in groups (Bass, 1990; Cartwright & Zander, 1968; Hemphil, 1961; Yukl, 1989). Extending this work we argue that the importance of these two leadership functions in social dilemmas is contingent upon the prevailing needs of the group. Leadership as a Solution to Social Dilemmas Social dilemma is the generic term for two special classes of social conflicts, the resource dilemma and the public good dilemma (Komorita & Parks, 1994; Messick & Brewer, 1983). A resource dilemma involves a potential conflict between a group of people over the distribution of a finite resource. Many natural resources have this property (Van Vugt et al., 2000). A public good dilemma entails a potential 5 conflict between group members over the contributions necessary to create a commonly shared good, for example, a group facility or a successful team performance (Stroebe & Frey, 1982). Because public good dilemmas are relatively more common in small groups, the present research focuses on these types of conflicts (although our conclusions may speak to both dilemma types). At the heart of the public good dilemma lies the freerider problem (Kerr, 1983; Olson, 1965). In creating public goods, group members must decide whether to cooperate by making a contribution or to free-ride on the contributions of others. Free-riding is personally more attractive, but if it is too widespread the group may fail to secure the good, which leaves every group member worse off. Cooperation is thus the most sensible strategy from a collective viewpoint. Yet members may be reluctant to cooperate because (a) it is tempting to free-ride, and (b) even if they cooperate there is a risk that they are being exploited by other group members (“the sucker’s pay-off;” Komorita & Parks, 1994). Hence, both motives of greed and fear can explain the emergence of free-riding in social dilemmas (Kerr, 1983). One of the main tasks of a group leader in a social dilemma is to prevent freeriding. In a fully cooperative group task each member will be inclined to cooperate spontaneously, and therefore the demands on leadership are relatively straightforward. The main leadership function is to coordinate the, often diverse, efforts of group members and bring them together. In a fully competitive group task (“zero-sum game”) the role of leaders is also fairly clear. They must primarily serve as an arbitrator to mediate between the individuals (or groups) with opposing interests. Social dilemmas, however, are mixed-motive conflicts in the sense that for individuals there are incentives both to free-ride -to enhance their personal welfare -as well as to cooperate for the group. Here the role of leadership becomes more complicated. In order to achieve the group goals, leaders must deter and punish freeriding. At the same time, however, they should contribute to a positive group climate to ensure that members, particularly those with a cooperative inclination, enjoy being in this group 6 and are tempted to stay and contribute to the group’s welfare (Van Vugt, Jepson, & De Cremer, 2001). Social dilemmas thus provide an ideal laboratory to test hypotheses about the different roles and functions of leadership in groups. Instrumental Perspective on Leadership in Social Dilemmas Rational decision-making theories, such as game theory, rational choice theory and social exchange theory (Hardin, 1968; Luce & Raiffa, 1957; Olson, 1965; Thibaut & Kelley, 1959), postulate that individuals in social dilemmas are primarily concerned about their short-term self-interest. According to the notion of self-interest, group members endorse leaders when they are instrumental in providing favourable outcomes. Hence, they will cooperate more with leaders who provide material rewards for cooperation and punishments for freeriding. Note that a leader who achieves group success will also be endorsed, according to this perspective, because all members benefit materially if the leader successfully resolves a social dilemma. Group members will therefore primarily look for information about leader characteristics which suggests that they are capable of solving the free-rider problem by modifying members’ selfish behaviours. This instrumental perspective on leadership in social dilemmas has received considerable empirical support (Messick et al., 1983; Rutte & Wilke, 1984; Samuelson & Messick, 1986; Samuelson et al., 1984). For example, Messick et al. (1983) showed that group members were more likely to choose a leader if their group had previously failed to collectively sustain common resource pool. Furthermore, in applied research on a water shortage in California (Tyler & Degoey, 1995) it was found that residents who perceived the shortage as more threatening were more likely to defer control over their personal water use to the water authorities. Finally, in a public good experiment, Yamagishi (1986) found that the introduction of a sanctioning regime to punish free-riders increased the average contribution level in groups. These findings, however, do not tell the whole story about the role of leadership in social dilemmas. First, if group members are solely concerned about 7 material outcomes, they would not hesitate to give up complete decisional freedom to a leader so as to eliminate the free-rider problem. In fact, even in a collective crisis individuals are quite reluctant to vote for an autocratic leadership regime (Rutte & Wilke, 1984, 1985; Samuelson, 1993; Tyler & Degoey, 1995). When given the choice between various structural solutions group members tend to prefer democratic solutions, such as majority and unanimity rules, above autocratic leadership (Rutte & Wilke, 1985) -perhaps due to a feared loss of personal control or concerns about the possibility of corruption and exploitation by the leader. Second, there are no straightforward effects of the use of reward and punishment schemes by leaders in social dilemmas. For example, in one study it was found that a weak sanctioning regime resulted in a lower level of cooperation than when there was no sanctioning at all (Tenbrunsel & Messick, 1999). Furthermore, Yamagishi (1986) showed that a sanctioning regime had more effect on members with a low trust in other people’s voluntary cooperation than on members with high trust in others. Finally, in a recent field study on water conservation we found that when water authorities sanctioned excessive water use – by installing water meters in properties – this intervention had more effect on residents who identified weakly with their residential community than residents with a strong community identification (Van Vugt, 2001). These results suggest a more group-based view on the role of leadership in social dilemmas. The emergence of leadership and the subsequent influence of leaders may be influenced by the prevailing motives and needs within a particular group. Sometimes group members are more focused on the material outcomes that leaders could potentially provide, judging their ability to solve the problem of freeriding. At other times, however, they may look for their leaders to satisfy other important, non-instrumental needs. A Relational Perspective on Leadership in Social Dilemmas An alternative motive for leader endorsement in social dilemmas is the extent to which the leader is capable of fulfilling the relational needs of group members. 8 People join groups for a multitude of different reasons, only some of which are instrumental (e.g., the achievement of some specific group goal). An important alternative motive for group membership is that it allows individuals to fulfil a desire to establish positive social relationships with other people. Group membership gives people a sense of identity and belonging, which is regarded by many theorists as essential for the survival and psychological well-being of an individual (Baumeister & Leary, 1995; Brewer, 1979; Deci & Ryan, 2000; Tajfel & Turner, 1979; Turner, Hogg, Oakes, Reicher, & Wetherell, 1987). For example, social identity theory (Tajfel & Turner, 1979) assumes that people's sense of who they are, their identity, is shaped in part by the social groups with which they are associated. This aspect of people’s self-concept is described as one’s social identity (Tajfel, 1972): “that part of an individual’s self-concept which derives from his [or her] knowledge of his [or her] membership of a social group (or groups) together with the value and emotional significance attached to that membership (1972, p. 273). Social identity theory argues that people seek a positive social identity, which is achieved through a positive distinctiveness from relevant other groups. Whereas social identity theory was originally formulated as a theory to describe the intergroup dynamics underlying social identity processes, more recent adaptations of this theory have focused more on the intragroup dynamics of social identity (Hogg & Abrams, 1988; Turner et al., 1987). Recently, for example, social identity researchers have started to investigate the emergence of leadership in groups (Hains, Hogg, & Duck, 1997; Hogg, 2001; Turner & Haslam, 2000). These insights are useful for understanding the different roles of leadership in social dilemmas. Within social dilemmas, the emergence and effectiveness of leadership may be dictated, in addition to their perceived instrumentality, by the perceived influence on the relational needs of members (i.e., identity and belongingness needs). Leaders can fulfil these needs by developing pleasant social relationships with group members and by creating an encouraging social climate so that members enjoy their group membership and wish to stay. Leaders who facilitate 9 these goals will be perceived as legitimate (French & Raven, 1959). Hence, members will cooperate voluntarily with them in solving social dilemmas without the promise of specific material rewards or the threat of punishments (Lind & Tyler, 1988; Tyler, 2000; Tyler & Dawes, 1993; Tyler & Lind, 1992). A Differential Needs Model of Leadership What specific factors determine whether instrumental or relational needs are more salient in the endorsement of leaders in social dilemmas? Following an instrumental perspective on leadership in social dilemmas, leaders' primary role is outcome-directed. Leaders must help to increase the material pay-offs for the group and its members and, to ensure this, they must demonstrate the ability to detect and punish free-riding. In contrast, a relational perspective asserts that to secure acceptance and cooperation from group members leaders must show a concern about the social relationships within the group. This differential needs-model thus suggests that impressions about the role of leaders may differ between members, as well as between groups, depending on their dominant needs and motives. A key difference between group members presumably lies in the extent to which they consider themselves to be part of the group. Either they perceive themselves essentially as a unique individual, in which case their personal identity is more salient, or as member of a group, in which case their social identity is more salient (Turner et al., 1987). When their social identity is salient group members tend to believe that they have very much in common with other group members, both in terms of opinions, values, shared goals and interests (Spears, Oakes, Ellemers, & Haslam, 2000). The salience of a social identity thus blurs the distinction between an individual’s selfinterest and the group’s interest – they are perceived as overlapping – which effectively solves the social dilemma conflict (Brewer, 1979; De Cremer & Van Vugt, 1999). Furthermore, a salient social identity enhances depersonalised trust in other group members’ cooperative intentions (Kramer & Brewer, 1984). Both mechanisms, a cooperative goal and depersonalised trust, may explain why group members exhibit 10 greater voluntary cooperation in social dilemmas when a social identity is activated (Brewer, 1979; Brewer & Kramer, 1986; De Cremer & Van Vugt, 1999; Kramer & Brewer, 1984). Conversely, when a personal identity is salient, individuals perceive little communality between their attitudes and interests and those of the other group members. Hence, they are less likely to cooperate spontaneously and they expect no cooperation from others either. The Moderating Role of Identity The attitudinal and behavioural differences between social and personal identifiers yield important implications for the perceived role of leadership in social dilemmas. First, lacking in trust, personal identifiers probably see a greater urgency to voluntarily accept a leader as a solution to the social dilemma. Moreover, they will be looking for evidence in leaders which shows that they are capable of overcoming the free-rider problem in their group. Finally, the primacy of self-interest over the group interest dictates that the behaviour of personal identifiers is shaped by the expected material rewards and punishments received from the leader. Accordingly, personal identifiers will be more accepting of and influenced by instrumental leadership in overcoming a social dilemma. What about social identifiers? Because the conflict between their self-interest and the collective interest is absent, or at least less intense, they are more optimistic that the social dilemma can be resolved through the voluntary contributions of themselves and others in the group. Hence, they presumably perceive less need to move from an unstructured group setting to a situation with a group leader. Yet, when a leader is already there they will assign less weight to the specific instrumental qualities that the leader brings to the group. After all, they have an intrinsic motivation to cooperate with the leader in securing the collective good. What social identifiers presumably care about more is the possible impact of leadership on the psychological experience of their group membership, in other words, their social