Quantitative Research Process Matrix

See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/334179940

What do the followers want?: The core functions of leadership

Article  in  The Leadership Quarterly · July 2019

DOI: 10.1016/j.leaqua.2019.04.003

CITATIONS

0 READS

586

2 authors:

Some of the authors of this publication are also working on these related projects:

Formal and Informal Help in and across Projects View project

Frontline Employee Innovation View project

Ghufran Ahmad

Lahore University of Management Sciences

12 PUBLICATIONS   38 CITATIONS   

SEE PROFILE

Christoph H. Loch

University of Cambridge

176 PUBLICATIONS   7,403 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Ghufran Ahmad on 13 November 2019.

The user has requested enhancement of the downloaded file.

Contents lists available at ScienceDirect

The Leadership Quarterly

journal homepage: www.elsevier.com/locate/leaqua

Full Length Article

What do the followers want?: The core functions of leadership M. Ghufran Ahmada,⁎, Christoph Lochb a Suleman Dawood School of Business, Lahore University of Management Sciences, Lahore, Pakistan b Cambridge Judge Business School, University of Cambridge, United Kingdom of Great Britain and Northern Ireland

Introduction

There are many leadership models, which partially contradict one another and may fail. For example, charismatic leadership, a widely usedmodel (Yukl, 2013), has been found tobecome routinized (Trice& Beyer, 1986) and fail on many occasions (Conger, 1990; O'Connor, Mumford, Clifton, Gessner, & Connelly, 1995). One likely reason of the failure is mismatch between the needs of the followers (or an organi- zation or a society) and the design features of a particular leadership model. The mismatch is likely to arise when one fails to ask a funda- mental question: what was the model designed to deliver? Is charisma required for its own sake, or are leaders performing the functions that are needed by the followers? Different leadership models prescribe different functions, some of which do not overlap (Burke et al., 2006; Lord, 1977; Yukl, 2013; Zaleznik, 1977). Thus, the question arises whether there are any essential functions of leadership which explain its universal presence (Brown, 1991) and necessity (Bennis, 2007)? In this experimental study,1 we first identify two important social

needs of the followers that require leadership: (a) when team members need to coordinate their actions to reacha collective goal, and (b)when they need to cooperate to achieve a goal, requiring that they resolve conflicts of interests. Second, we investigate the effectiveness of lea- dership in such situations. In examining leadership fromtheperspective of followers, we respond to the calls for study of leadership through the reversed lens of followers (Bligh, 2011;Carsten,Uhl-Bien,West, Patera, &McGregor, 2010;Uhl-Bien,Riggio, Lowe,&Carsten, 2014).Our study contributes to the leadership literature in three ways: a) we study the demand for leadership from an evolutionary perspective, in contrast with most leadership studies which focus on traits, behaviors, effects, and styles of leaders, b) we develop a unified framework and research design that integrates different strategic games, and c) we compare leadership with other alternatives as responses to strategic situations across a variety of incentive structures. Leaders in modern complex organizations are expected to possess a

suiteofpersonality traits (Zaccaro,Kemp,&Bader, 2004), change styles to suit situations (Vroom & Jago, 2007), and perform a myriad of roles (Kets de Vries, 2007). We argue that from the point of the origin of

leadership in the history of our species, achieving coordination and cooperation among individuals and organizational units are the two most formidable challenges that made leadership necessary from fol- lowers' point of view. Lawrence and Lorsch (1967, p. 142) call these activities integration, defined as “the achievement of unity of effort among the major functional specialists in a business”. Inspiring visions, shared values, charisma, reward systems, and other influence processes described in the major leadership models are the tools designed to achieve functional integration and may be used when necessary. Coordination involves concerted (mutually consistent) action by

two or more individuals to achieve a particular outcome. Coordination problems are prevalent in social, economic, and organizational activ- ities (Friedman, 1992; Schelling, 1960). Cooperation issues arise when there are conflicts of interest, when individuals can benefit from the actions of others without carrying their share of the cost; social di- lemmas such as the public goods dilemma and the common resource dilemma, represent cooperation issues (Komorita & Parks, 1996). Col- lective existence becomes difficult unless these problems are solved, and scholars from many disciplines have argued that leadership plays an important role in the resolution of these problems for the followers. We now review extant research on coordination and cooperation issues and the roleof leadership, beforedescribing theexperimental studyand its results.

Leadership theory and hypotheses

Leadership and coordination

Coordination problems are challenging to resolve (Camerer, 2003; Schelling, 1960) and are a major topic of research in economic and organizational decision making (Friedman, 1992; Shapira, 1997). Calvert (1992) and Foss (2001) have proposed leadership as a solution to the problems of coordination, and the effectiveness of leadership in resolution of coordination problems has been studied by Weber, Camerer, Rottenstreich, and Knez (2001). Leadership, defined this way, implies coordination of group activity; yet coordination problems have mainlybeen investigatedbyeconomists rather than social psychologists

https://doi.org/10.1016/j.leaqua.2019.04.003 Received 31 May 2018; Received in revised form 10 April 2019; Accepted 27 April 2019

⁎ Corresponding author. E-mail addresses: [email protected] (M.G. Ahmad), [email protected] (C. Loch).

1Wedonotpresent a general theoryof leadership. Instead,we focuson some important evolutionary functions of leadershipwhichmeet theneedsof the followers and how different environments may result in the exercise of different forms of leadership.

The Leadership Quarterly xxx (xxxx) xxx–xxx

1048-9843/ © 2019 Elsevier Inc. All rights reserved.

Please cite this article as: M. Ghufran Ahmad and Christoph Loch, The Leadership Quarterly, https://doi.org/10.1016/j.leaqua.2019.04.003

(Abele&Stasser, 2008).Mainstream leadership theories generally focus on traits, styles, charisma, relational skills, and inspirational abilities of leaders (Antonakis & Day, 2017), and some scholars exclude co- ordination from the portfolio of leadership activities by declaring it a “managerial” activity (Kotter, 1996; Zaleznik, 1977). However, recent theories of team and shared leadership emphasize coordination as one key leadership process (Yukl, 2013), particularly in virtual teams (Yoo & Alavi, 2004). Research on coordination problems identifies a number of solutions

with limited efficacy, such as leadership, communication, and focal points (Crawford, Gneezy, & Rottenstreich, 2008; Devetag & Ortmann, 2007; Weber et al., 2001). Leadership may work as a solution when it combines communication and focal power with the capacity to exclude non-cooperative group members. Whereas we study the effects of lea- dership in solving coordination challenges, our main focus is on studying the leadership from followers' perspective and evaluate the demand for leadership when followers are presented with coordination challenges. We specifically employ an evolutionary perspective which we explain below. Furthermore, in contrast with the typical research design for studyof leadership inwhich a subject acts as a leader (Sahin, Eckel, & Komai, 2015), we focus on followers and study the conditions under which followers demand leadership.

Leadership and cooperation

The idea that leadership may help resolve cooperation problems such as social dilemmas goes back to Hobbes (1651), Homans (1974), and Hardin (1968). In the early 1980s, Messick and others pioneered experimental research on leadership as a solution to the problem of social dilemmas. The results of these initial studies indicated that sub- jects preferred a leader when a common resource depleted quickly due to group members' mismanagement. Subjects reduced their personal harvests as leaders as compared to their harvests as individual group members, but they still harvested more for personal use as compared to their allocations to other group members (Messick et al., 1983; Rutte & Wilke, 1984, 1985; Samuelson & Messick, 1986; Samuelson, Messick, Rutte, & Wilke, 1984). People find it difficult to install leadership endogenously (Güth,

Levati, Sutter,&VanDerHeijden, 2007), and larger groups are likely to fail despite having leadership (Weber et al., 2001). People expect lea- ders to distribute resources equitably (De Cremer & van Knippenberg, 2002), and followers start distrusting leaders and reduce cooperation when the payoff differences between leaders and followers increase (Wilson & Rhodes, 1997; Wit & Wilke, 1988). Setting an example through initiating actiononexclusive information (Komai, Stegeman,& Hermalin, 2007) and through self-sacrifice (De Cremer & Van Knippenberg, 2005) makes leaders credible enough to be followed (Potters, Sefton, & Vesterlund, 2005, 2007). Sanctioning systems (Yamagishi, 1986) and autocratic leaders (Güth et al., 2007; Van Vugt, Jepson, Hart, & De Cremer, 2004) are able to force more cooperation, but this reduces trust among group members (Mulder, Van Dijk, De Cremer,&Wilke,2006)andcauses somepeople to leave thegroup (Van Vugt et al., 2004). Leaders are expected to make the group successful irrespective of circumstances and they are held responsible for group failure (Weber et al., 2001; Wit, Wilke, & Van Dijk, 1989). Some recent experiments and reviews have also studied and ana-

lyzed the role of leadership in solving cooperation challenges (Chaudhuri, 2011; Frackenpohl, Hillenbrand, & Kube, 2016; Sahin et al., 2015; Van Lange, Balliet, Parks, & Van Vugt, 2014). The current findings suggest that leadership may play limited role in resolving conflict of interest and facilitating cooperation with the help of com- munication and enforcement of exclusionary punishment on non-co- operative group members. In addition to leadership, communication and institutional frames may also offer limited solutions. In current research, we focus more on the demand for leadership by the followers insteadof theusual researchdesign inwhicha subject is expected toact

as a leader. From an evolutionary perspective, leadership is hypothe- sized to have evolved as a psychological adaptation or preference, is demanded by the followers under specific situations, and played an important role in the resolution of coordination and cooperation chal- lenges, as explained in the following.

Leadership and evolution

The role of biological factors in human behavior is receiving in- creased attention of researchers in the field of management and orga- nizational behavior (Markóczy & Goldberg, 1998; Nicholson & White, 2006; Shane, 2009). One area in which the influence of genetic factors has been identified is leadership (Arvey, Rotundo, Johnson, Zhang, & McGue, 2006). Identifying the genetic factors responsible for the emergence of leadership (Arvey et al., 2006; Johnson et al., 1998; Johnson, Vernon, Harris, & Jang, 2004) and exploring how environ- mental interactions influence expression of these genetic propensities (Zhang, Ilies, & Arvey, 2009) are not the focus of the current paper. This paper builds on theory of how the process of biological evo-

lution selects genes for their fitness advantages conferred to the or- ganisms bearing such genes (Dawkins, 2006). So “[o]ne quickly won- ders:Whathas caused this leadership structure toemerge?Whyhasone animal (the alpha) emerge to lead the collective? And how does this leadership cause this collective to flourish – or founder?” (Judge, Piccolo, & Kosalka, 2009, p. 855). Answering these questions requires moving beyond a trait based perspective to consider how a behavior proves advantageous to organisms in their struggle for survival. Re- cently, researchers have proposed that leadership evolved because it resolved coordination problems and facilitated cooperation in group living animals (see e.g., Glowacki & von Rueden, 2015; McNamara & Trumbull, 2007; Smith et al., 2016; Spisak, O'Brien, Nicholson, & Van Vugt, 2015; Van Vugt, Hogan, & Kaiser, 2008; von Rueden & Van Vugt, 2015). We build upon these ideas to propose and test a model of the essential functions of human leadership for the followers. Group living in primates evolved because it provided better defense

against predators (Alexander, 1974) and allowed greater access to and control over resources under conditions of high resource competition (Wrangham, 1980). We illustrate the role of these factors with the help of a strategic game of coordination, the so-called “matching game”.2 A strategic game represents a situation in which an individual makes a decision which results in a payoff that depends not only on his or her own on decision but also on the decision(s) of the other individual(s). Suppose two individuals A and B can go either left or right and both have to decide in which direction to move. Moving in either direction can give some benefit (say, resource access or predator avoidance). But the benefit accrues only if both individuals stay together—they are interdependent. Fig. 1 presents the decision choices available to each individual and their associated benefits. This simple game formalizes the benefit for the individuals both

going together either left or right; if they move in different directions, neither gets any benefit. If a group of individuals could not maintain cohesion by moving and staying together, group living would not arise (Conradt & Roper, 2005; King, Johnson, & Van Vugt, 2009). However, coordinated action is not the only problem. Sociality also involves co- operative interactions among group members, and sometimes some group members may obtain a benefit without paying any cost or without returning the favor. We illustrate the role of mutual coopera- tion with the strategic game of cooperation, the “prisoner's dilemma game,” as depicted in Fig. 2. We explain the game of cooperation in the context the survival

2Camerer (2003, p. 343) describes this game as “A matching game,” also known as a “pure coordination game.” This game should be distinguished from two other common games, namely matching pennies game and matching pro- blem (e.g., allocating students to schools).

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

2

challenge of maintaining resource access. Suppose the same two earlier presented individuals A and B were able to move together and conse- quently monopolized a resource patch. Now they face challenge of being displaced from their resource patch by another group of in- dividuals. If both individuals cooperate to collectively fight they can maintain

their hold. Or, both can decide to not cooperate (and avoid the fight) andmayeither runaway to someother less attractive resourcepatchor share the resource patch with additional individuals. Finally, one of them could be more cooperative and fight harder and sustain more injuries while the other individual fights less vigorously or hardly at all andconserves energy. In this case, thevigorousfighterbenefits the least due to injuries and energy consumed in fighting, although the pair continues its hold on the resource. The less active individual “free rides” and benefits from the altruistic action of another individual. This game highlights that the conflict of interest, in the form of the temp- tation to not cooperate or contribute one's share yet still reap the benefit, is wide spread (Komorita & Parks, 1996). Strategic games developed by evolutionary biologists capture the

essential elements of the adaptive challenges of coordination and co- operation and have been used extensively in studying the evolution of various traits (Silk, 2007; Smith, 1982). We have presented above two strategic games, the Matching and Prisoner's Dilemma games. Other important two-person games, which we introduce later in the Methods section, include the Stag Hunt, the Chicken, and the Battle of the Sexes games. Scholars have proposed many mechanisms for the evolution of co-

operation including kin selection (Hamilton, 1964), reciprocity (Axelrod & Hamilton, 1981; Trivers, 1971), indirect reciprocity (Alexander, 1974; Nowak & Sigmund, 2005), punishment (Clutton- Brock & Parker, 1995; Silk, 2005), costly signaling (Gintis, Smith, & Bowles, 2001), and group selection (Hamilton, 1975; Sober & Wilson, 1998). Leadership has also been proposed as an evolved solution to the problems faced by the followers or group members (Glowacki & von Rueden, 2015; McNamara & Trumbull, 2007; Smith et al., 2016; Spisak et al., 2015; Van Vugt et al., 2008; von Rueden & Van Vugt, 2015). Observational studies of primates provide evidence that dominant

individuals of a group act as leaders and perform some essential social functions, including control anddirectionof groupmovement (Goodall, 1986), maintenance of group cohesion (Chance, 1967), conflict re- solution and control of free riding (Aureli & de Waal, 2000), protection from predators (Harcourt & Stewart, 2007; Wrangham & Peterson, 1996), rank regulation, and resources allocation (Meder, 2007). “From a human perspective, evolution reached a major milestone when ani- mals began to live in groups, and authority and its precursors, dom- inance and deference, made this possible” (Heifetz, 1994, p. 49). The

relationship between dominance and leadership is also documented in human societies (Fleischer & Chertkoff, 1986; Giessner & Schubert, 2007; Lord, De Vader, & Alliger, 1986; Megargee, 1969; Nyquist & Spence, 1986). Hence, hierarchy provides the first model of (author- itarian) leadership for the resolution of the problems of coordination and cooperation. The anthropological literature provides evidence that leaders co-

ordinate group affairs and mediate conflicts among group members even in egalitarianhunter gatherer societies,whichare characterizedas fluid societies with lower levels of resource conflicts and coordination issues. Although status differences between leaders and other group members are minimal and leaders enjoy no special power and privi- leges, the leaders still exist and perform the needed social functions (Boehm, 1999; Lewis, 1974; Woodburn, 1982). Our description of the evolutionary origins of leadership identifies

major recurrent adaptive problems for continued group living during the evolutionary history of humans: how to coordinate group activity, control free riding, and resolve conflicts of interest among group members. The evidence suggests that leadership evolved to help facil- itate mutual cooperation and coordination among group members in the context of the adaptive problems and helped the group members overcome these problems. Scholarshaveargued that long-termevolutionarypressures result in

adaptations that allow humans to respond to such challenges in char- acteristic ways. These adaptations are emotional and cognitive in nature and variously labelled as pre-adaptations (Witt & Schwesinger, 2013), cognitive adaptations (Cosmides & Tooby, 1992), or evolved decision rules (von Rueden & Van Vugt, 2015). These evolved decision rules become operative automatically and unconsciously under specific contexts (Bargh&Chartrand, 1999; vonRueden&VanVugt, 2015) and represent deep evolutionary rationale which may be different from the economic rationality (Kenrick & Griskevicius, 2013). As an evolu- tionary adaptation, leadership is likely to be a psychologically encoded mental algorithm that gets enacted under specific circumstances re- sembling the adaptive problems that ancestors to modern humans consistently faced (Van Vugt et al., 2008). Therefore, we expect lea- dership to consistently emerge in response to the coordination and cooperation challenges.

Hypothesis 1A. The presence of a coordination need and conflict of interest (which prevents cooperation) each separately will increase the demand for leadership intervention.

Humans spent almost 95% of their time as a species in small-scale hunter-gatherer societies (von Rueden & Van Vugt, 2015). The an- thropological studies indicate that these hunter-gatherer societies were small communities with minimalist leadership characterized as

Individual B Go left Go right

Individual A Go left 2, 2 0, 0

Go right 0, 0 2, 2

Notes: The numerical figures in each cell represent benefits for each combination of collective actions (left/right) by both players; left side figure for individual A and right side figure for individual B.

Fig. 1. The Matching Game: Coordinated decisions benefit both players.

Individual B Cooperate Do not cooperate

Individual A Cooperate 2, 2 0, 5

Do not cooperate 5, 0 1, 1

Notes: The numerical figures in each cell represent benefits for each combination of cooperative actions by both players; left side figure for individual A and right side figure for individual B.

Fig. 2. The Prisoner's Dilemma Game: Cooperative actions benefit both players.

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

3

egalitarian leadership which generally arose informally in response to the coordination or conflict resolution challenges such as deciding di- rection of group movement for searching food and shelter. These ega- litarian leaders did not have any special authority or power to enforce their opinion or will on others (Boehm, 1999; Glowacki & von Rueden, 2015; Smith et al., 2016). However, these groups also faced frequent inter-group competition for scarce resources, escalating the level of coordination and cooperation challenges (Bowles, 2009; Gat, 2006; Glowacki, Wilson, & Wrangham, 2017; Keeley, 1997). Under such si- tuations a second form of leadership, known as authoritarian leader- ship, emerged to help a group address its challenges (Chagnon, 1997; Earle, 1997; Lowie, 1948; Meggitt, 1967; Price, 1981). The group granted more authority to an authoritarian leader, a privilege not available to an egalitarian leader (Boone, 1992; Hooper, Kaplan, & Boone, 2010; von Rueden & Van Vugt, 2015). Recent evidence suggests that solitary punishers may facilitate cooperation (O'Gorman, Henrich, & Vugt, 2009) and authoritarian leaders are likely to exercise such as role given they are granted more authority to control the behavior of other group members and enforce social norms (Boehm, 1999). The evolutionaryandhistorical evidence suggests that there are two

basic forms of leadership: authoritarian and egalitarian; the type of leadership being expressed depends upon the level of resource conflict and coordination challenges being faced by the group members. Authoritarian leadership facilitates quicker decision making and action when the level of coordination and/or conflict resolution challenge rises, otherwisepeople prefer egalitarian leadership (Boehm,1999;Van Vugt et al., 2008). Recent theoretical work (Hooper et al., 2010) also suggests that coordination and cooperation challenges may result in authoritarian leadership. Therefore, we hypothesize that as the level of coordination and cooperation challenge goes up, people will increase their preference for authoritarian leadership.

Hypothesis 1B. Higher coordination need and conflict of interest will increase the demand for authoritarian leadership more than the demand for egalitarian leadership.

Scholars have proposed many evolutionary strategies besides lea- dership to resolve the problems of coordination and cooperation. The strategies include kin selection, direct reciprocity, punishment, indirect reciprocity, costly signaling, and group selection. Why leadership evolved (or persisted) when alternatives were available? It may be ar- gued that leadership may have provided more benefits as compared with the alternatives, at least under some specific situations. According to the evolutionary analysis, an adaptation is a response to a recurring problem. If a problem is recurring, it means that available mechanisms are not sufficiently effective, requiring development of new solutions. In other words, an adaptation is more likely to evolve because it is a solution better than other available solutions. Out of the proposed alternative solutions, kin selection, group se-

lection, and costly signaling do not represent an appropriate point of comparisonwith leadership, especiallywhen thedemand for leadership is under study. In other words, itmaybe more appropriate to inquire: if people can solve coordination and cooperation challenges themselves, then why leadership is needed? Here, reciprocity and punishment- based solutionsprovidea suitablepoint of comparisonasboth solutions can be implemented individually without a need for leadership. Most contemporary research implicitly utilizes an underlying re-

ciprocity framework (as represented in the most commonly used game of the Prisoner's Dilemma) wherein players are expected to achieve coordination and cooperation through some form of communication. Arguably, the second most empirically investigated strategy is punish- ment which may be considered a derivative of reciprocity and hence has also been labelled as strong reciprocity (Fehr, Fischbacher, & Gächter, 2002). Camerer (2003, p. 403), in a comprehensive survey of research, concluded that coordination failure is commonand the effects of communication are often weaker than expected (see also Devetag & Ortmann, 2007). There is some evidence of punishment producing

cooperation (O'Gormanet al., 2009) yet its effectivenessmaybe limited due to its anti-social bias (Herrmann, Thöni, & Gächter, 2008) and adverse effects (Van Lange et al., 2014, p. 59). As we propose leadership as a mechanism for the evolution of co-

operation in humans, we expect leadership strategy to achieve more cooperation, and more value for the group than the two benchmark strategies: individual decision making (that is, relying on one's own preferences for fairness and reciprocity), and punishment (pressuring the other side into cooperation).

Hypothesis 2. The presence of leadership will increase the level of cooperation and, thus, a group's total payoff, as compared to individual decision making and punishment.

Apart from theevolutionarymechanismsdiscussedabove, outcome- based monetary incentives are another way of focusing people on producing value (taking the form of bonuses, commissions, etc.). Individual level incentives can also produce coordination, as the re- solution of coordination problems carries no conflicting motives. But conflicts of interest present dilemmas about the decision to cooperate when payoffs are at stake, which cannot easily be overcome with in- dividual incentives. Cooperation and coordination failures are not infrequent (Camerer,

2003) and administration of monetary incentives (or their withdrawal in the form of punishment) has unintended consequences (Gneezy, Meier, & Rey-Biel, 2011). Furthermore, monetary incentives tend to change personal and interpersonal behavior in unexpected ways (Vohs, Mead, & Goode, 2008) and there is increasing evidence that the use of monetary incentives may produce unethical or selfish behavior, espe- cially under the situations involving conflict of interest (Gino & Mogilner, 2014; Kouchaki, Smith-Crowe, Brief, & Sousa, 2013; Wang, Zhong, & Murnighan, 2014). Monetary incentives may make co- operative choices salient (Crawford et al., 2008). Hence, we expect that monetary incentives may solve the coordination problem, but not the problem representing conflict of interest.

Hypothesis 3. Outcome-dependent monetary payoffs increase coordination and a group's total payoff, but only in absence of a conflict of interest.

Methods

We conducted eight experiments, each examining the demand for a leader and the effect of the leader in a different strategic game. Each experiment involved 25 pairs of subjects playing the game over 10 re- peated rounds.3 A subject participated only once, that is, in one out of eight experiments only. 400 subjects were recruited at a social sciences lab in Paris close to the Sorbonne universities campus; they were aged between 18 and 26, studying in various academic fields. 164 subjects were male and 236 female. Eachparticipantwas randomlyassigned to anotherperson to forma

pair. The pair did not see each other, communicate, or know each other's identity during the study. The two players in a pair were de- signated as player A and player B. They played in isolation via com- puters, coordinating implicitly without being able to discuss strategies or announce intentions. We played six games (described in the following section) in which

each player was paid a fixed fee of €5 for participation in the experi- ment. To evaluate Hypothesis 3, we played two games with outcome- dependent monetary payoffs as described in the following section.4 The study was conducted in French; instructions for one game are shown in

3For our purposes, measuring the demand for and the cooperation benefit obtained by a leader, a pair is an appropriate minimal group. 4At the time of signing up for the experiment, participants were informed

that they will receive compensation in the form of currency (not medals).

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

4

the Appendix 1. The instructions introduced the experiment and ex- plained what each player was required to do in each round. The in- structions further explained the pay-off and governance structures as elaborated in the following sections below, while noting that the computer was pre-programmed to respond as a leader.5

The timeline of decisions is shown in Fig. 3. The players chose a governance structure, saw the leader's recommendations (if a leader was chosen by either of them), and made their decisions (by choosing their respective actions). The player who had chosen a punishment option decided whether or not to actually execute punishment and as- sign a low value payoff to the other player. The enforcing leader (if chosen) then punished deviations, if any. The design of experiment allowed each player to select a governance structure and make deci- sions independent of the choices made by the other player. For ex- ample, a participant could carry out individual decision or punishment even if the other participant had selected recommending or enforcing leadership (see Fig. 3). In other words, the choice of one player did not override the choice of the other player. The computerwas programmed to award a maximum cumulative penalty of zero payoff in case both individual and leader penalties were implemented. Finally, the players saw the results and proceeded to the next round. All decisions and outcomes were visible to both parties and recorded. Only those in- dividuals participated in the study who demonstrated their under- standing of the instructions by correctly responding to a questionnaire.

Coordination and conflict of interest in strategic games

We manipulated the level of coordination and cooperation difficulties with the help of six strategic games. We presented the decisions payoffs in non-monetary terms to engage the underlying cognitive mechanisms in a simple and clear way. Research on the effects of money suggests that monetary rewards may not work as incentives (Gneezy et al., 2011), may crowd-out non-monetary incentives (Ariely, Bracha, & Meier, 2009), may change the personal and interpersonal behavior of individuals (Vohs et al., 2008), includingbutnot limited to increased incidenceofmakingunethical and selfishdecisions (Gino& Mogilner, 2014; Kouchaki et al., 2013;Wang et al., 2014). Because, some of our strategic games permitted selfish choices, we did not use monetary incentives to represent decision payoffs to avoid the effects of monetary incentives. Our research focused on the effects of leadership as an evolutionary adaptation, separate from the ef- fects of monetary incentives. Research on non-monetary incentives also provides evidence of their efficacy (Sorauren, 2000). Therefore, we used the concept of a medal to create the payoffs in an intuitively appealing manner to serve as sufficiently motivating non-monetary incentive.6 See

Fig. 4 for the games and decision payoffs. The top-left game is called trivial because there is no inter-

dependence: player A gets the gold for choice A1, and player B for the choice B2, independent of what the other person does. There is simply no need for interaction, neither a need to coordinate nor any conflict of interest. The Trivial game (TR) is the benchmark where players are not interdependent. The top-right game is the Matching game (MT) discussed earlier: a

player is rewarded not based on his/her choice, but based on thematch with the other player (either A1 matched by B1, or A2 matched by B2). If they fail to coordinate, neither gets a medal. MT is high on co- ordination difficulty, as only two of the four possible decision choices yield payoffs; but there is no conflict of interest as both players benefit equally if they achieve coordination. The Battle of the Sexes game (BS) represents the same level of co-

ordination difficulty as that of MT, but has an added dimension of in- terest conflict becauseanypayoff-yielding collectivedecision involves a payoff difference between the two players. This conflict cannot be re- solved in one round, but it can be resolved if the partners play repeated rounds and take turns. The Stag Hunt game (SH) incorporates lower level of coordination

requirement than the previous games, as there is a fail-safe strategy for each player (A2 and B2) requiring no coordination though yielding a lower payoff. SH has no conflict of interest; the players must trust each other and coordinate in order to play the risky strategy (A1 and B1) with the higher payoff. The Prisoner's Dilemma game (PD) represents conflict of interest, as

we have described earlier. The only equilibrium is one where both parties chose the “defection action” (A2 and B2) and end up with bronze medals. PD represents well the cooperation problem (conflict of interest) of social dilemmas, where one party free-rides on the other, while both could be better off contributing to the common good at a slight cost to their personal payoffs. This game is lower on coordination scale than BS, as one strategy (A2 and B2) is fail-safe: no matter what the other player does, the player playing this strategy has a guaranteed payoff. The fear of loss associated with the risk of cooperation along with the temptation of a higher payoff produce non-cooperation. Finally, the game of Chicken (CH) represents a different version of

conflict of interest. This game has a built-in temptation to try for a

Players choose: • Individual decision • Punishment • Recommending

leader • Enforcing leader

Players decide (A1 or A2, and B1 or B2)

Leader (if chosen) recommends value maximizing actions

Players who chose punishment: • see the decisions • decide whether or

not to punish the other player

Players see results of the round

Enforcing leader (if chosen) punishes deviations

Time

Fig. 3. Timeline of decisions and information for each round.

5A separate trial run with 23 different participants was done to ensure that players completely understood the instructions including calculating the pay- offs accurately, executing the governance structure, and understanding that the leader responses were pre-programmed. The trial run demonstrated that the players comprehended the instructions and played without ambiguity, while fully understanding the currency and mode of payment. 6A preliminary survey with 49 separate participants with characteristics si-

milar to those who participated in the main studies showed that 1) people consideredgettingamedal a formof recognitionequivalent invalue to that of a

(footnote continued) monetary reward, 2) gold, silver, and bronze medals indicated differing levels of value, and 3) people preferred medals and felt motivated to obtain medals of higher denomination. The survey also revealed that people considered a silver medal given to each member of a pair to be a value-maximizing offer for the group as compared to a gold medal given to one member and a bronze medal given to the other member of a pair; it is consistent with the evolutionary preference for egalitarianism as observed in the hunter-gatherer societies (see main text). Results of the preliminary survey allowed us to create payoff structures for the strategic games. Subsequently, a post-game survey showed that participants recognized payoff differences inherent in the medals, were motivated to obtain medals, and the higher-denomination medals were more valued. In a separate set of studies we also used monetary incentives, which we discuss later.

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

5

higher payoff, and correspondingly, contains an interest conflict, but the fear of loss interferes: a player could try playing aggressively (by selecting A2 or B2) to go for the gold medal but the cost of aggression couldbeveryhigh (yieldingnomedal at all) if theotherplayer behaves similarly. In contrast, the second strategy (A1 or B1) has a guaranteed payoff of at least a bronze medal anda higher payoff (of a silver medal) if the other player also behaves “sensibly”. Contrary to PD, the higher loss may motivate group cooperation in CH. But the safer option is not an equilibrium (that is, each party would do better by unilaterally de- viating from it). Table 1 classifies the games in terms of the difficulty of their co-

ordination and conflict resolution challenges. TR is the no-inter- dependence benchmark. With two dimensions, the comparison is lex- icographic: we cannot compare MT (high coordination, no conflict of interest) with CH and PD (medium coordination, conflict of interest). However, we can say that BS is higher (or at least as high) on both dimensions than CH and PD, and MT is higher than SH. In order to examine the effect of outcome-dependent monetary re-

wards (Hypothesis 3), appealing to economic rationality, we also played the Prisoner's Dilemma game with monetary incentives (PDI) and the Stag Hunt game with monetary incentives (SHI): gold, silver, and bronze medals were worth 0.50€, 0.40€, and 0.30€ respectively,7

and the accumulated money over ten rounds was paid to each player at the end of experiment. In the instructions for two games, the figure in Appendix 1 representing payoffs (in terms of medals) also included specification of the monetary value of each medal in PDI and SHI games.

Governance structure

At the beginning of each of the ten rounds, before decisions to co- operate were made, each player chose among the following options in terms of governance of the round. They correspond to the group co- operation mechanisms discussed under Fig. 2: reciprocity, punishment, and leadership (both egalitarian and authoritarian).

1. Individual decision making. Each player makes a decision based on his or her personal preferences (and may use information from previous rounds).

2. Punishment. The player reserves the right to punish the other player by allotting him/her a low-value payoff after decisions have been made for the round.

3. A recommending leader. The player requests a recommendation by a leader for the pair. A leader (simulated by the computer) then makes a decision recommendation to the pair before they make a decision. The computer is programmed to always recommend to both sides the cooperative choices that would maximize value for thegroup.8 This is apure recommendationwithnoenforcement and represents an egalitarian leader who can suggest solutions but has no power to enforce his or her opinions over other group members.

Trivial Game Matching Game

B1 B2 B1 B2

A1 Gold, Silver Gold, Gold * A1 Black, Black Gold, Gold *

A2 Silver, Silver Silver, Gold A2 Gold, Gold * Black, Black

Battle of the Sexes Game Stag Hunt Game

B1 B2 B1 B2

A1 Black, Black Silver, Gold * A1 Black, Silver Gold, Gold *

A2 Gold, Silver * Black, Black A2 Silver, Silver * Silver, Black

Prisoner’s Dilemma Game Chicken Game

B1 B2 B1 B2

A1 Silver, Silver Black, Gold A1 Silver, Silver Bronze, Gold *

A2 Gold, Black Bronze, Bronze* A2 Gold, Bronze * Black, Black

Fig. 4. The Six Strategic Games: A1 and A2 are player A's choices, and B1 and B2 player B's. In each cell, corresponding to one choice combinationof the two players, both players' payoffs are presented, player A to the left and player B to the right. The payoffs are represented as “medals”: gold is thebest, silver the second, bronze the third, and black cor- responds to “no medal”, the worst outcome. For example, in the Trivial game, player A choosing A1 and player B B1 (the upper left cell) results in player A receiving a gold and player B a silver medal. * represent equilibria, or stable decision combinations where no subject wants to deviate unilaterally. Cooperation means that each player decides for the goodof the group (thehighest sumofmedal values), possibly sacrificing personal interest if it conflicts with the interest of the group. In the Matching game and the Battle of the Sexes game, equilibria re- present cooperation. In the Stag Hunt, Prisoner's Dilemma, and Chicken games, decision combination A1 &B1 represents cooperation. In the Trivial game, decision combination A1 & B2 represents “pseudo” cooperation (as there are no trade-offs).

Table 1 Coordination difficulty and interest of interest in the strategic games.

Trivial Stag Hunt Chicken Prisoner's Dilemma Matching Battle of the Sexes

Coordination challengea No Medium Medium Medium High High Conflict of interestb No No Yes Yes No Yes

Notes a The coordination challenge is medium if there is a “safe” choice by both players that prevents the lowest (no medal) outcome irrespective of the choice of the

other player; the coordination challenge is high if there is a risk (depending on the choice of the other side) of “no medal” for every choice of each player. b A conflict of interest is present if one player's choice that increases his/her own payoff reduces the other player's payoff.

7 In a separate set of studies, we used monetary incentives only and did not use medals. The studies are reported later.

8 In case of the Battle of the Sexes game, the computer was programmed to offer value-maximizing choice on alternate basis. For example, if Player A first selected leadership option (either recommending or enforcing), then computer recommended option with Player A getting the gold medal (and silver for Player B), followed by alternative recommendation for Player B receiving the gold (and Player A silver), irrespective of player A or B selecting leader option (either recommending or enforcing) the next time. The programming was im- plemented for both recommending and enforcing leaders.

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

6

4. An enforcing leader. The player requests a recommendation by a leader who will enforce the decision by allotting an inferior payoff to the player not following the recommendation. As in option 3, the leader recommends a value maximizing choice to both sides. This option is designed to represent an authoritarian leader who has the power to enforce his or her will upon others.9

Results

The demand for leadership

We first consider the question whether subjects demand more lea- dership when they face difficult coordination and conflict of interest challenges (Hypothesis 1). Comparing the fraction of subjects who chose a leader option (coordinating or enforcing) offers a first level of aggregated evidence on this question (we turn to the richness of in- dividual behavior later); this is graphically summarized in Fig. 5. There are no significant trends in subjects' governance choices over

time; theydonot start outwith somerandomchoiceand thenconverge. Therefore, the initial periods arenot just transient “noise”, and itmakes sense to consider all the data points over all periods; in absence of trends over time, we can aggregate over periods in the remaining analysis. Now turning to our question, we report the summary statistics for

leadership in Fig. 5. Although there is noise (some subjects demand leadership even in TR, where it does not make sense to do so because individuals do not influence each other), the demand for leadership clearly goes up statistically with the coordination and conflict of in- terest challenge. As there are no trends, we can count how many times an individual

demands the leadership option over the ten rounds.10 A leadership gov- ernance is used differently across the games: F(5, 54)=6.47, p < .001. Theuse of leadership is larger collectively in thefive strategic games than in TR, t(54)=3.60, p < .01. InMT, subjects solicit leadership more than in SH, t(54)=2.93, p < .01; in BS, leadership is used more often than in PD, t(54)=1.96, p=.06, and in CH, t(54)=3.00, p < .01. This corre- sponds to the ranking of the games in Table 1. Table 2 compares the usage frequency of the two types of leader-

ship. The use of the enforcing leadership option increases in (a) MT as compared to SH, t(18)=2.40; p=.03, and (b) BS as compared to PD, t (18)=1.83; p=.08, and CH, t(18)=2.57, p=.02. None of these comparisons are statistically significant for recommending leadership; its usage increases only weakly with the conflict of interest. We also compare the overall demand for leadership for a) increased

level of coordination challenge at the same level of conflict of interest, and b) increased level of conflict of interest and the same level of co- ordination challenge. In the first case, the demand for leadership in- creases in (i) MT as compared to SH, t(18)=2.70; p=.02, (ii) BS as compared to PD, t(18)=2.42; p=.03, and (iii) BS as compared to CH,

0 5

10 15 20 25 30 35

1 2 3 4 5 6 7 8 9 10

Trivial Game Individual Decision Punishment Leadership

Leadership: M = 10.60; SD = 3.60 0 5

10 15 20 25 30 35

1 2 3 4 5 6 7 8 9 10

Matching Game Individual Decision Punishment Leadership

Leadership: M = 16.40; SD = 4.12

0 5

10 15 20 25 30 35

1 2 3 4 5 6 7 8 9 10

Stag Hunt Game Individual Decision Punishment Leadership

Leadership: M = 11.90; SD = 3.28 0 5

10 15 20 25 30 35

1 2 3 4 5 6 7 8 9 10

Ba!le of the Sexes Game Individual Decision Punishment Leadership

Leadership: M = 17.90; SD = 2.96

0 5

10 15 20 25 30 35

1 2 3 4 5 6 7 8 9 10

Prisoner’s Dilemma Game Individual Decision Punishment Leadership

Leadership: M = 14.90; SD = 2.56 0 5

10 15 20 25 30 35

1 2 3 4 5 6 7 8 9 10

Chicken Game Individual Decision Punishment Leadership

Leadership: M = 13.30; SD = 3.83

Fig. 5. Number of times a given governance was chosen per game and over 10 consecutive rounds (count of occurrences per round, out of 50 subjects' choices per round). “Leadership” combines the recommending and enforcing leadership structures, their separation is shown in Table 2.

9A post-game survey showed that 1) participants understood that the com- puter was programmed to respond as a leader, and 2) the programmed leader advice they received during the experiment was not different from the advice they were likely to get from a real-life leader.

10Statistical analysis (not reported here) confirmed that round-wise data met parametric conditions. Hence we apply ANOVA for the reported analyses. All reported p-values are two-tailed unless otherwise specified.

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

7

t(18)=3.00, p=.01. In the second case, the demand for leadership increases in (i) PD as compared to SH, t(18)=2.28; p=.04, but not in (ii) CH as compared to SH, t(18)=0.88; p=.39, and (iii) BS as com- pared to MT, t(18)=0.94, p=.36. The comparison shows that a co- ordination challenge increases demand for leadership more than a conflict of interest. Thus, we have support for Hypotheses 1A and 1B, that subjects

choose a leadership option in response to a coordination and conflict challenge, and resort to authoritarian leadership when the difficulty level increases.11 Still, there is a significant amount of “noise”: perhaps subjects are not able to estimate the nature and level of challenge well, or subjects interpret the game in a normative sense and want to show the other side how to behave, even if it does not have an influence on themselves. We explore this noise further in the following.

The benefit of leadership

We now examine the benefit that each strategy brings to the group through promoting cooperation between the players (Hypothesis 2). In order to assess the effect of each type of strategy on the decisions made by subjects, we estimated the following logistic regression for each game (excluding TR), with the dependent variable the decision to co- operate in a particular round.

= t

t Probability (cooperation in round (yes/no) )

function (governance structure in round , cooperation by other (yes/no) , punishment by other (yes/no) ).

t 1

t 1

As the dependent variable is dichotomous, we estimate the prob- abilityofmakingacooperativedecision.Thisprobability inaparticular round may also be influenced by the cooperative decision and the choice of punishment strategy by the other player in the previous round. Tables 3-A–3-E describes the odds of reaching a cooperative decision for each independent variable. The odds of taking a cooperative decision increase substantially with

the use of leadership in all strategic games, whereas the cooperative de- cision taken by the other player in the previous round influences the de- cision to cooperate in the current round only in three games. Interestingly, the use of the punishment strategy by a subject fails to yield cooperation andmay indeedhave theopposite result, as is suggestedby lower-than-one odds ratio in BS (we further investigate this aspect later). In order to compare the outcomes over various governance struc-

tures, imagine we value the medals in the same way as in the manip- ulations with incentivized payoffs: a bronze medal is worth €0.30, a silver €0.40 and a gold €0.50.12 We examine the value achieved by the group, as the medal obtained by an individual is determined by the choices of both subjects.

Table 2 Leadership usage.

M (SD)

Trivial Stag Hunt Matching Prisoner's Dilemma Chicken Battle of the Sexes

Recommending leadership 7.40 (2.91) 6.20 (1.75) 7.40 (2.50) 7.80 (2.15) 7.30 (2.06) 8.30 (1.89) Enforcing leadership 3.20 (2.10) 5.70 (3.06) 9.00 (3.09) 7.10 (2.18) 6.00 (2.40) 9.60 (3.72)

Notes: Count of occurrences for 50 subjects over ten rounds.

Table 3-A Logistic regression results: Stag Hunt.

Independent variables Cooperation

B Wald χ2 RSE p OR 95% CI OR ME

Enforcing leadership 2.07 13.43 0.57 <.001 7.96 [2.62, 24.12] 1.97 Recommending leadership 0.68 6.02 0.28 .01 1.98 [1.15, 3.42] 0.58 Punishment 0.27 0.92 0.28 .34 1.32 [0.75, 2.30] 0.17 Cooperation by other t–1 0.47 5.44 0.20 .02 1.60 [1.08, 2.37] 0.89 Punishment by other t–1 0.42 2.74 0.25 .10 1.52 [0.93, 2.48] 0.86

Table 3-B Logistic regression results: Matching.

Independent variables Cooperation

B Wald χ2 RSE p OR 95% CI OR ME

Enforcing leadership 1.04 11.32 0.31 <.01 2.81 [1.54, 5.14] 1.15 Recommending leadership 1.03 7.79 0.37 <.01 2.80 [1.36, 5.76] 1.14 Punishment 0.08 0.07 0.30 .79 1.08 [0.61, 1.93] 0.19 Cooperation by other t–1 1.53 64.28 0.19 <.001 4.60 [3.17, 6.68] 1.41 Punishment by other t–1 0.02 0.00 0.28 .94 1.02 [0.59, 1.76] 0.66

11 In response to a post-game question, subjects noted that they selected leadership: 1) to force the other player to make appropriate choice (in case of enforcing leadership) and 2) to recommend appropriate choice to the other player (in case of recommending leadership). The subjects did not say that they selected leadership to recommend or enforce a choice on themselves. Subjects also informed that they did not take turns to win medals in any game.

12The medals are ordinal and do not determine one cardinal evaluation. In this valuation, there is a jumpbetweennomedal and thebronzemedal.Wealso tested the valuation with the jump between bronze and silver (i.e. bronze=0€.10, silver=€0.40), and between silver and gold (i.e., bronze=€0.10, silver=€ 0.20, gold=€0.50). The results remained qualita- tively the same as the ones we report.

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

8

The trivial game and the retaliation effect of punishment Table 4 shows that in TR the average value achieved per round by a

group is close to one (i.e., both subjects obtain a gold medal), and the values achieved with individual decision making and the two types of leadership are statistically not distinguishable—this is what we expect given one player's outcome is independent of what the other player does. However, the value obtained goes down by almost one third (to €0.69) when one player or both choose the punishment option, and indeed, between 10 and 20% of subjects do choose this option even in TR (see Fig. 6). Across all six games, the punishment governance does not seem to

be successful in producing cooperation; it results in the lowest average value obtained in all games. The punishment governance is often seen as a potential alternative to leadership, an alternative emphasizing in- dividual bottom-up “correction” of behavior rather than centralized, top-down driven behavioral correction. However, punishment is used even when there is no need for any punishment at all (in TR and MT), and when it is used, it represents an aggressive action that triggers retaliation spirals. We see retaliation spirals byanalyzing thedynamicsof thedecisions

over time in a regression: whether an individual wants the option to punish is influenced by (in addition to his or her unobserved

Table 3-D Logistic regression results: Chicken.

Independent variables Cooperation

B Wald χ2 RSE p OR 95% CI OR ME

Enforcing leadership 2.39 30.21 0.43 <.001 10.87 [4.64, 25.44] 1.85 Recommending leadership 1.12 7.83 0.40 <.01 3.06 [1.40, 6.69] 0.58 Punishment 0.68 7.53 0.25 <.01 1.97 [1.21, 3.20] 0.14 Cooperation by other t–1 0.19 0.96 0.20 .33 1.21 [0.82, 1.79] 0.61 Punishment by other t–1 −0.18 0.51 0.25 .48 0.84 [0.52, 1.36] 0.44

Table 3-E Logistic regression results: Battle of the Sexes.

Independent variables Cooperation

B Wald χ2 RSE p OR 95% CI OR ME

Enforcing leadership 1.15 16.17 0.29 <.001 3.17 [1.81, 5.56] 1.03 Recommending leadership 0.71 5.40 0.30 .02 2.03 [1.12, 3.70] 0.59 Punishment −0.02 0.01 0.25 .93 0.98 [0.60, 1.61] −0.14 Cooperation by other t–1 0.90 18.54 0.21 <.001 2.45 [1.63, 3.69] 0.79 Punishment by other t–1 −0.59 4.60 0.28 .03 0.55 [0.32, 0.95] 0.04

Notes for Tables 3-A to 3-E: Regression parameters of the generalized estimating equations for each game with the dependent variable the decision to cooperate in the current round (yes/no). Independent variables include governance structure (with four categories) chosen in the current round, cooperation by the other player in previous round (yes/no), and punishment strategy used by the other player in previous round (yes/no). RSE=robust standard errors; OR=odds ratio; ME=marginal effects with transformed scale. GENLIN procedure in SPSS 23 was used.

Table 4 Game-wise group benefit from governance structures.

M (SD)

Trivial Stag Hunt Matching Prisoner's Dilemma Chicken Battle of the Sexes

Individual decision 0.99 (0.02) 0.67 (0.26) 0.49 (0.50) 0.57 (0.08) 0.44 (0.40) 0.42 (0.45) Punishment 0.67 (0.34) 0.30 (0.29) 0.35 (0.40) 0.29 (0.22) 0.41 (0.31) 0.33 (0.38) Recommending leadership 0.99 (0.03) 0.73 (0.28) 0.77 (0.43) 0.62 (0.12) 0.66 (0.31) 0.62 (0.42) Enforcing leadership 1.00 (0.00) 0.83 (0.26) 0.90 (0.30) 0.64 (0.15) 0.78 (0.13) 0.73 (0.35)

Notes: Average value achieved by a pair per round. Individual decision=individual decision making chosen by both, punishment=at least one side chooses the pun- ishment option (and no one leadership), recommending leadership=at least one side chooses a recommending leader (and no one enforcing leadership or pun- ishment), enforcing leadership=at least one side chooses an enforcing leader (and no one punishment). Maximum possible mean value=1.00.

Table 3-C Logistic regression results: Prisoner's Dilemma.

Independent variables Cooperation

B Wald χ2 RSE p OR 95% CI OR ME

Enforcing leadership 2.05 27.59 0.39 <.001 7.74 [3.61, 16.61] 1.11 Recommending leadership 0.79 3.76 0.40 .05 2.20 [0.99, 4.86] −0.16 Punishment 0.17 0.34 0.30 .56 1.19 [0.67, 2.12] −0.77 Cooperation by other t–1 −0.20 0.87 0.21 .35 0.82 [0.54, 1.24] −0.29 Punishment by other t–1 −0.29 1.34 0.25 .25 0.75 [0.46, 1.22] −0.33

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

9

Table 5A Logistic regression results: Trivial.

Independent variables Punishment

B Wald χ2 RSE p OR 95% CI OR ME

Punishment by other t–1 1.34 16.97 0.32 <.001 3.80 [2.01, 7.17] 0.10 Cooperation by other t–1 −0.86 2.25 0.57 0.13 0.42 [0.14, 1.30] −1.00 Cooperation by both t–1 0.45 0.75 0.52 0.39 1.56 [0.57, 4.30] −0.34

Table 5B Logistic regression results: Stag Hunt.

Independent variables Punishment

B Wald χ2 RSE p OR 95% CI OR ME

Punishment by other t–1 0.56 5.25 0.24 <.05 1.75 [1.08, 2.81] −0.43 Cooperation by other t–1 0.16 0.32 0.27 .58 1.17 [0.68, 2.00] −0.63 Cooperation by both t–1 0.04 0.01 0.31 .91 1.04 [0.56, 1.92] −0.69

Table 5C Logistic regression results: Matching.

Independent variables Punishment

B Wald χ2 RSE p OR 95% CI OR ME

Punishment by other t–1 0.70 11.31 0.21 <.01 2.02 [1.34, 3.04] −0.41 Cooperation by other t–1 0.02 0.01 0.23 .94 1.02 [0.63, 1.61] −0.75 Cooperation by both t–1 N/A N/A N/A N/A N/A N/A N/A

Table 5D Logistic regression results: Prisoner's Dilemma.

Independent variables Punishment

B Wald χ2 RSE p OR 95% CI OR ME

Punishment by other t–1 1.16 23.75 0.24 <.001 3.19 [2.00, 5.09] −0.07 Cooperation by other t–1 0.43 3.68 0.22 .06 1.54 [0.99, 2.38] −0.43 Cooperation by both t–1 −0.17 0.29 0.31 .59 0.85 [0.46, 1.56] −0.73

Table 5E Logistic regression results: Chicken.

Independent variables Punishment

B Wald χ2 RSE p OR 95% CI OR ME

Punishment by other t–1 0.62 7.77 0.22 <.01 1.86 [1.20, 2.87] −0.29 Cooperation by other t–1 −0.29 1.04 0.28 .31 0.75 [0.43, 1.31] −0.75 Cooperation by both t–1 0.24 0.66 0.30 .42 1.28 [0.71, 2.28] −0.48

Table 5F Logistic regression results: Battle of the Sexes.

Independent variables Punishment

B Wald χ2 RSE p OR 95% CI OR ME

Punishment by other t–1 1.08 14.63 0.28 <.001 2.94 [1.69, 5.11] −0.35 Cooperation by other t–1 −0.64 7.85 0.23 <.01 0.53 [0.34, 0.82] −1.21 Cooperation by both t–1 N/A N/A N/A N/A N/A N/A N/A

Notes for Tables 5A–5F Regression parameters of the generalized estimating equations for each game with the dependent variable punishment chosen in the current round (yes/no). Independent variables include decisions of previous round concerning punishment chosen by other player (yes/no), cooperation by other player (yes/no) and cooperation by both players (yes/no). RSE=robust standard errors; OR=odds ratio; ME=marginal effects with transformed scale. GENLIN procedure in SPSS 23 was used.

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

10

psychological makeup) the game's challenge and structure and the ac- tions of the other side in the previous round. Thus, we estimate the following regression per game, with the dependent variable the choice of punishment governance (as this is a 0–1 variable, this is a logistical regression that estimates a probability):

= tProbability (punishment option in round (yes/no) )

function (Punishment by other (yes/no) , cooperation by other (yes/no) , cooperation by both (yes/no) ).

t 1 t 1

t 1

Table 5 shows which variables are significant drivers of retaliatory behavior. The regression suggests that choice of the punishment governance

by the other side strongly increases an individual's propensity to also choose the punishment option; this influence is statistically significant in all games separately (adding whether the other player then actually did punish in last period no longer makes a statistical difference, thereforewe left this variableout). In addition, thepersonalso reacts to whether or not the other player defected last time, but only in the high- conflict games, PD and BS. Thus, we see that the punishment

TR CH MT

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Fig. 6. Details of individual behavior by round, in the Trivial (TR), Chicken (CH) and Matching (MT) games. The rows are pairs (from 1 to 25), the columns rounds (from 1 to 10). In each row, the upper-half represents decisions by player A and the lower-half decisions by player B. The color in each cell represents the medal outcomes (gold, silver, bronze; no color=no medal). The circles with various fillings are the governance options chosen by each individual: no filling=individual decision, quarter filling=punishment option, three-quarter filling=recommending leader, complete filling=enforcing leader. If a cell with a quarter circle is shaded on top of the color, this person executed punishment. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

11

governance with its threat indeed causes retaliation spirals and reduces cooperation; it is cooperation that breeds further cooperation.

The benefit of leadership: Further evidence for hypothesis 2 Wenowreturn to theevidence inTable4with respect toHypothesis

2. The payoff value obtained by the group is higher with each leader- ship type as compared with individual decision-making and punish- ment. The difference is significant (p≤.001) relative to punishment in all games. Leadership also performs better relative to individual deci- sion making in CH and MT (p≤.005) and in PD and BS (p < .05) (the exceptions are SH and TR only in the recommending leadership case). These statistics support Hypothesis 2. In SH, the players have the necessary information to make the de-

cisions, so they do not need leadership to recommend or enforce a higher-value solution. Yet the value generated by both types of lea- dership takencollectively is greater (p=.015) than thevaluegenerated through individual decision-making—leadership helps the individuals make better decisions. However, the means reported in Table 4 do not take into account

the cost imposed by the enforcing leadership. We set the enforcing leadership's deviation punishment action to no medal; if this punish- ment is taken into account, the pair's payoff is significantly reduced in thegameswith conflictof interest (inPDtheaverage reduces to0.35, in CH 0.51, and in BS 0.70). In MT, the lack of conflict of interest makes enforcement unnecessary; here, the enforcing leadership performs very well. SH has no conflict of interest but mutual distrust stemming from the fear that theotherplayermayavoid risk and thusdestroyone's own benefit expected from the risky strategy. Here, the enforcing leader does need to punish, as in the games with conflict of interest, and the average group benefit reduces to 0.69, the same as under individual decision making. In games with conflict of interest, it is recommending leadership which produces the highest value.13

Illustration: behavior at the individual level Fig. 6 shows the behavior of individual subjects over the ten rounds

for three games. The left-hand diagram shows the Trivial game, in which most subjects achieved gold medals. But only 8 of the 25 pairs achieved this acting independently of each other (with individual de- cision making). The other 17 pairs used punishment and/or leadership as well, although their payoffs were independent of one another. Why? Consider pair two: player A starts by punishment, to which player B

respondswithdemandinga leader, and then theyexchangepunishment twiceover the remainingperiods.Wesee similarpunishment exchanges in pairs 5 (once) and in pairs 6–8, 14, 15, 19, 21, and 23–25. Pairs 13, 20 and 22 achieve full gold medals throughout, but “encourage” it by demanding leadership. These observations illustrate what the above statistical analyses

suggest: the social dynamics take on a life of their own, subjects try to “correct” each other and get into retaliatory punishment exchanges. In 9 pairs, the mutual influencing dynamics actually prompt at least one player to deviate from the goldmedal-producing actionmore thanonce over the then periods—the social exchanges matter. The retaliation spirals demonstrated by the subjects may be an expression of the an- tisocial punishment (Herrmann et al., 2008). Second, the observations show the variations that stem from in-

dividual idiosyncrasies. Why does, for example, player A of pair 2 start with punishment? Maybe this player wanted to signal toughness to the other side, or he came into the experiment in a bad mood, or he was a curious experimenter who wanted to see what happened if he did something unexpected. The starting move was, thus, unpredictable, but the dynamic it set off was quite predictable: once someone attacks the

other side, or shows distrust by demanding a leader, the other side tends to respond in kind. People seemed to respond emotionally by exercising punishment and retaliating to punishment in a way yielding inferior outcomes. Subjects knew that they were independent in terms of outcomes, but they might have perceived the game in a normative way, trying to educate the other side in making the value maximizing choice. Now we turn to the other two games in Fig. 6. CH has an conflict of

interest, MT a higher coordination challenge. Indeed, there are 80 in- stancesof thepair gettingnomedal inMTversus49 inCH.On theother hand, the pairs in MT achieve shared gold much more often, reflecting the lack of a conflict of interest. Only two teams in CH (11 and 20) achieve the shared cooperative outcome (silver for both) for at least 4 consecutive periods, and both use leadership to do it. In MT, 13 pairs achieve gold for>4 consecutive periods, only one of them (pair 20) without using leadership. Punishment is used more in CH than in MT (156 versus 135 times),

and it fails to prompt cooperation in 40% of the times when it is used. For example, punishment directly causes a string of mutual gold medal outcomes to be disrupted in pairs 3, 6, 10, 14, 17 and 18 of the MT game. Similarly, it causes the pair to move from silver‑silver to gold- bronze in pairs 1–4, 6, 8–10, 17, 21, 22 and 25 in the CH game. In sum, both games demonstrate the dynamics that lie behind the statistics re- ported earlier.

The effect of outcome-dependent monetary payoffs Fig. 7 displays the governance structures chosen by subjects in

games played with an outcome-dependent monetary incentive. The trends over time are similar to the patterns observed in Fig. 5 (games without incentive), but monetary incentives influence the choice of governance structure, as Table 6 shows. Table 6 suggests that the key effect of outcome dependent payoffs

on the choice of governance is a reduction in punishment: in both games, punishment is chosen significantly less often when payoffs de- pend on the outcome, and in turn, the use of both individual decision making and leadership increases; their relative proportions do not changemuch. Inotherwords, the fact thatmoney is at stake focuses the minds of the subjects away from the previously seen mutually dama- ging retaliation spirals. However, subjects do not change the level of demand for leadership. Table 7 compares the effect of monetary incentives on the value

achieved by the group. In the Stag Hunt game, incentives have a sig- nificant positive effect (except in the enforcing leader condition): they focus the minds of the subjects on the value of coordination; adding incentives to the individual decision making governance offers a higher value increase to the pair than recommending leadership without in- centive (and higher than the enforcing leadership without incentive after punishment cost; it has a net mean value of 0.69). Incentives are, however, less effective in the Prisoner's Dilemma

game, which incorporates a strong conflict of interest. Indeed, in- centives do not achieve a statistically significant value increase in any of the governance structures in thePrisoner'sDilemmagame, except for punishment. These results represent support for Hypothesis 3.14

We again illustrate the findings of the aggregated analyses with observations of individual behavior. Fig. 8 shows all behavior details for SH and SHI (for space reasons, we do not show the PD and PDI games here; their details are shown in Appendix 2). The SH panel on the left of Fig. 8 is more silver than gold-colored:

many players understand the risk of taking the high-value decision and conservatively go for silver. Punishment, however, tends to degenerate into bouts of retaliation, just like in TR. For example, in several

13The design of study results in self-selection problem and does not permit causal conclusions. Hence the evidence presented here may be seen as con- sistent with the hypothesis 2 but not as causal evidence.

14A separate set of six studies only used the monetary incentives. The results are similar to the ones reported here. Please see Appendix 3 for details of those studies.

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

12

instances players used punishment even if they both had won gold medals in that particular round (for example, pairs 11, 13, 16, 18, and 20). Moreover, punishment is ineffective in eliciting a cooperative de- cision from the partner in majority of the cases. Those, in contrast, who chose leadership, were more likely to take the risk of cooperation and be successful at obtaining it: the enforcement, or at least re- commendation, of the leader allowed them to go for the risky higher- value outcome. SHI (right-hand panel in Fig. 8) overall features more gold than SH.

9 pairs achieve (almost) full cooperation with the help of leadership, and pairs 2 and 14 even manage to achieve full cooperation just with implicit coordination (individual decision making). The use of punish- ment is significantly reduced (from147occurrences inSHto82 inSHI), and although retaliation cycles still occur (e.g., groups 6, 8, 13, 21), they are fewer than in SH, and there are instances where one player slides into punishment but the other resists the temptation to retaliate (for example, players in pairs 1 and 22). These results illustrate the statistical differences shown in Tables 6 and 7. It seems that the presence of incentives focuses the players on the

high-value opportunity to be had (Crawford et al., 2008). This does not necessarily imply that these subjects are “greedy,” caring about mate- rial payoffs only (in the sense of neoclassical economic theory). The incentive provides a signal that makes the high-value cooperative equilibrium more salient. Thus, the players take a bit more risk to achieve the better outcome (rather than settle on the second-best low value equilibrium), and they are a bit less aggressive in controlling the other side and more tolerant of setbacks (such as getting punished) without falling into retaliation spirals (although thesedo still occur).As we saw in Table 7, this focus-on-value benefit of monetary incentives is strong in the Stag Hunt game because there is no intrinsic conflict of interest between the subjects. The benefit of incentives is much smaller in the PD game (see Appendix 2). Extant research shows that the players can coordinate their actions in a 2-person Stag Hunt game under specific conditions (Devetag & Ortmann, 2007). Our findings are consistent with the extant research and identify evolutionary processes at play while also providing results on the comparative effects of dif- ferent mechanisms involved.

Discussion and conclusion

Our discussion of the evolutionary roots of the core functions of leadership ended with the hypotheses that (1) problems of collective action in a group, particularly coordination difficulties and conflict of interest, drive thedemand for leadership, and (2) that leadership canbe more effective in solving these challenges than individual-level at- tempts to cooperate, so it is more valuable the higher the challenges of coordination and conflict resolution in the group. Thus, we examine leadership from the viewpoint of the followers—when do they want it, and when do they benefit from it? The strategic games provide clear evidence of both—the leadership

option was chosen more often in games with high coordination and conflict of interest challenges, and leadership made a significant dif- ference in helping the pair find high-value cooperative outcomes in games with more difficult challenge. Moreover, the games illustrated

Table 6 Governance structures in games without and with monetary incentive.

M (SD)

Stag Hunt Prisoner's Dilemma

Without incentive

With incentive

p-Value Without incentive

With incentive

p-Value

Individual decision

23.40 (4.45)

27.10 (2.60)

.036 20.70 (2.67)

23.10 (3.03)

.077

Punishment 14.70 (5.03)

8.20 (2.70)

.002 14.40 (3.75)

11.30 (3.27)

.064

Leadership 11.90 (3.28)

14.70 (2.67)

.051 14.90 (2.56)

15.60 (2.01)

.505

Note: Count of occurrences per round, out of 50 subjects' choices per round.

Table 7 Group benefit in games without and with monetary incentives.

M (SD)

Stag Hunt Prisoner's Dilemma

Without incentive With incentive p-Value Without incentive With incentive p-Value

Individual decision 0.67 (0.26) 0.78 (0.26) .008 0.57 (0.08) 0.59 (0.10) .249 Punishment 0.30 (0.29) 0.48 (0.33) .004 0.29 (0.22) 0.41 (0.22) .003 Recommending leadership 0.73 (0.28) 0.87 (0.24) .012 0.62 (0.12) 0.63 (0.14) .862 Enforcing leadership 0.83 (0.26) 0.77 (0.30) .386 0.64 (0.15) 0.67 (0.14) .280

Note: Average value achieved by a pair per round, by game and by governance chosen. Individual decision=individual decision making chosen by both, pun- ishment=at least one side chooses the punishment option (and no one leadership), recommending leadership=at least one side chooses a recommending leader (and no one enforcing leadership or punishment), enforcing leadership=at least one side chooses an enforcing leader (and no one punishment). Maximum possible mean value=1.00.

0 5

10 15 20 25 30 35

1 2 3 4 5 6 7 8 9 10

Prisoner’s Dilemma With Incen!ve Game Individual Decision Punishment Leadership

Leadership: M = 15.60; SD = 2.01 0 5

10 15 20 25 30 35

1 2 3 4 5 6 7 8 9 10

Stag Hunt With Incen!ve Game Individual Decision Punishment Leadership

Leadership: M = 14.70; SD = 2.67

Fig. 7. Number of times subjects chose the indicated governance over the ten consecutive rounds (count of occurrences per round, out of 50 subjects' choices per round).

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

13

that individual-level coordination and punishment actions are less ef- fective in producing cooperation. Individual-level decision making did not reliably find the mutually beneficial actions, and punishment trig- gered retaliatory spirals that destroyed rather than supported co- operation. The leader did not always need to use power in order to achieve

group cooperation—recommending leadership was almost as effective as enforcing leadership, and even more effective if the cost of en- forcement was taken into account. Finally, we saw the potential of outcome-dependent monetary payoffs (which a leader can often shape) to focus the minds of the subjects on the value they could produce, but to a significant degree only in addressing a coordination challenge, not a conflict of interest. Leadership was used even in situations where it was unnecessary,

suchas inTRandSH.This is consistentwith the evolutionary argument that leadershipdemand is anadaptation, that is, a cognitivemechanism that may get activated automatically also in situations that only re- semble coordination and cooperation challenges. Incentives may act as

an alternative mechanism for leadership intervention when conflict of interest in the group is not too strong. Of course, an experiment is only a stylized representation of reality,

thus, there are limitations to generalizability. First, not much was at stake for the participants, so they may have been less engaged than in real life group and leadership situations. On the other hand, the re- taliatorybattles thatwewitnessed suggest that subjectswereverymuch engaged. Second, participants were not allowed to talk during the games, they had to communicate implicitly through their actions. On the other hand, communication in social context is sometimes not credible if not connected to a commitment, or even misleading; thus explicit communication makes the situation more complicated, but not always better. Third, the leader in our experiment was provided for free; there was no downside associated with choosing the leadership optionexcept for thepunishment cost of theenforcing leader. In reality, leaders may take an additional “tax”; for example, they may claim a greater share of group resources, or have personal interests in conflict with group interests. On the other hand, group members may not know

SHISH

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Fig. 8. Details of individual behavior by round, in the Stag Hunt (SH), and Stag Hunt with Incentives (SHI) games. The rows are pairs (from 1 to 25), the columns rounds (from 1 to 10). The color in each cell represents the medal outcomes (gold, silver, bronze; no color=no medal). The circles with various fillings are the govern- ance options chosen by each individual: no filling=in- dividual decision, quarter filling=punishment option, three-quarter filling=recommending leader, complete filling=enforcing leader. If a cellwith aquarter circle is shaded on top of the color, then this person executed punishment. (For interpretationof the references to color in this figure legend, the reader is referred to the web version of this article.)

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

14

how heavy the leader's tax burden is and thus weigh it less heavily in their governance choice. This discussion suggests that our experiment's limitations have equivalents in real situations and may therefore be sufficiently robust to carry over into real leadership situations. However, more replication and field studies are required in order to confirm the results suggested here. A fruitful area of further inquiry maybe theuseofn-persongames (as one limitationof our studies is the use of two-person games) which may allow studying the emergence of coalitions as a signalingdevice andalternative to leadership and its role in control of retaliatory punishment spirals. Whereas the current experimental economics literature has in-

vestigated the effects of leadership in social dilemmas, thedesignof our studyallowedus to study theproblemfromaverydifferentperspective. In most current research, players in the game are selected randomly to act as leaders (Sahin et al., 2015). Our study explores a different pro- blem, namely the demand for leadership and the effects of leadership presenceonparticipants' behavior. The leader is our research is nothing but a programmed response known as such to the participants, yet its mere presence brings about behavioral change in the responses of participants, hence providing evidence of the evolved psychological preference for leadership and associated decision rules under co- ordination and cooperation challenges. Our results suggest that in presence of strategic challenges, people express preference for leader- ship which plays a facilitative role in letting people exercise the co- operative choices. In other words, presence of strategic challenges and availability of leadership (even if in name only) biases people to re- spond cooperatively. The presence of alternative mechanisms to lea- dershipprovides amore stringent test for theusefulness of leadershipas a mechanism to promote coordination and cooperation. The leader in our experiments always acted in the benefit of followers. Reality is different—yet it does not seem to reduce the demand for leadership! In future, experiments in which a leader may behave erratically and may not alwaysbenefit thegroupmaybeconducted to study thedemand for leadership and the behavior of the followers. “When most of us think of leadership, we think of one person doing

something to another person. … The word ‘leader’ itself conjures up visions of a striking figure on a rearing white horse who is crying ‘Follow me!’” (Manz & Sims, 1991, p. 18). But this heroic conception of leadership, articulated through modern theories of charismatic and transformational leadership, does not “explicitly incorporate leadership behaviors such as organizing, coordinating, team building, and facil- itating groupdecisions” (Yukl, 1999, p. 38). Turning leaders into “hero- makers” (Manz & Sims, 1991; Pearce & Manz, 2005) only perpetuates the myth and shifts the problem. Coordination problems and conflict of interest present serious challenges (Camerer, 2003; Schelling, 1960), and leadership intervention maybe necessary at times for a group tobe successful, but the intervention need not necessarily be heroic to be effective. Our results suggest a more inclusive view of leadership than the

charismatic leadership that is widely discussed. For example, Kets de Vries and Florent-Treacy (1999, p. xxii) state: “The first [role of the leader] involves how leaders envision, empower andenergize to inspire andmotivate their followers. (…)This charismatic, empowering role is

the essence of leadership. The primary obligation of a leader is to de- termine where a company needs to go and then build commitment among followers to go in that direction.” This “vision responsibility” is one way of achieving coordination (and of asserting superiority on the side), but certainly not the only way, and maybe not even the best: studies on strategy processes have repeatedly shown that top managers must involve other members of the organization in setting the “vision”, and indeed, there are examples where the leaders did not set a vision at all, but it emerged out of debating and arguing and responding to en- vironmental pressures, and suddenly the management realized that theyhada strategy (andmaybeavision), butno leader set it.Already in 1978, Mintzberg (1978) recognized this process of emergent strategy. If the results of our experimental studycarryover intopractice, they

have important implications for how we should regard leaders: the essential leader would be doing less than all the things that have been discussed in charismatic/transformational leadership literature, fo- cusing on coordination and conflict resolution. This is a more humble view of leadership: it describes a role that many more people could be capable of achieving than the “heroic” leader role that requires pro- ducing a vision, absorbing pressure and motivating by good example, while simultaneously suppressing rivals with Machiavellian alliances. Leadership is about co-creating common good. This humble role requires very different skills from the leader, es-

sentially skills of diplomacy. They include, for example, the ability to recognize conflict of interest, understand enough of the group's tasks and interdependencies to be able to separate coordination failures (where pointing them out may already help a lot) and conflicts, un- derstand where the common interest of the group lies (appealing to which may help feuding individuals to bury their differences), under- stand individuals enough to be able to estimate who may go aggres- sively after his/her interests and who may be able to compromise. This explanation of the role of the leader for the group allows se-

parating it from the things a leader needs to do in order to get into the position and keep it: the competition for the alpha position requires a hyperactivecompetitivegame that leaderhopefulshave toplay inorder to beat rivals—this competition is exclusive and selfish by definition (only one individual per group can get there). However, the role of the leaderoncehe (or she)gets theremustbe inclusiveandoriented toward the group, and maybe it is less “exclusive” and more achievable by dedicated and well-meaning “normal” people than leadership literature suggests: maybe many people are capable of helping a group to co- ordinate and resolve conflicts, and therefore many people can be useful leaders. The distinction between the activities performed to achieve a lea-

dership position and the activities required to be an effective leader suggests a weakness of human resource systems that are designed to identify and reward leadershippotential. The talents and skills thatmay help someone rise in the hierarchy to reach a leadership position may not, later on, facilitate the group coordination effectiveness of the leader. Identifying how leadership contributes to success and what are the essential leadership activities from followers' perspective is neces- sary for team and organizational effectiveness.

Appendix 1. Game instructions (on PD example)

This is an experiment about decision making in 10 rounds. Your decisions will be rewarded with medals (gold, silver, and bronze.) You will be matched with another randomly selected person unknown to you. You will interact with that same person anonymously over the 10 rounds; one of you will play role A (Player A) and the other person will play role B (Player B) for the entire duration. In each round your decision will result in a medal which could be gold, silver, bronze, or none (denoted by black). As everyone knows, getting a

gold medal is best, getting no medal worst. Getting the gold is nice, but the result depends not only on your decision, but also on the decision of the other person. In each round player A will make a decision by choosing between the two possibilities A1 & A2, and player B will choose between the two

possibilities B1 & B2. The following diagram shows how the choices will be presented to each player and the results of decisions taken.

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

15

The result depends upon the decision of the other player

No medal

Player A receives Player B receives

Gold medal

Silver medal

Choice of player A

Player B chooses B1

Player B chooses B1

Player B chooses B2

Player B chooses B2

Choice A1

Choice A2

Gold medal

Bronze medal

Bronze medal

Silver medal

No medal

For example, if you are player A and you choose A1, then you get a silver medal if player B chooses B1, and you receive no medal if player B choosesChoiceB2. Similarly, if youareplayerBandyou selectChoiceB2, thenyouget a goldmedal if playerA selectsChoiceA1, andyou receive a bronze medal if player A chooses A2. So, while making decisions, both players should consider how the other player is going to decide. Before takingeachdecision, youwill have theopportunity to choose the conditionsunderwhichyoumakedecisions. Youwill chooseONEof the

following conditions before each round:

A. You take your decision individually. B. You takeyourdecision individually, but inaddition, afterhaving taken thedecision, youhave theoption to request allotmentof an inferiormedal to the other player.

C. You want a leader* to recommend a decision to you and the other player. Note that you and the other player may or may not follow the recommendation – there is no enforcement by the leader.

D. You want a leader* to recommend a decision to you and the other player and enforce the recommendation by allotting an inferior medal to the player who does not follow the recommendation. Note that the leader recommends and enforces the recommendation when any one player selects this option.

The typeof circumstances selectedbybothplayerswill appearon the computer screenalongwith the table containingdecisionchoiceswhenyou makeactualdecisions.Thecircumstances selected, recommendationsof the leader,decisionsmade, andmedals collectedbyeachplayer, and inferior medals allotted to each player will be displayed on the computer screens of both players at the end of each round. Youwill receive afixed sumof5€ in shapeof vouchers for your participation in this experiment. Pleasenote that during thewhole experiment –

fromstudying the instructions till leaving the lab –youarenot allowed to communicate throughanymeanswithother participants of this study.We will be forced to terminate the study in case of violation. * Note: The computer is pre-programmed to respond as a leader.

Appendix 2. Remaining detailed game observations

Fig. A-1 shows the behavior details of the BS, PD and PDI games.

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

16

PDBS PDI

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

Fig. A2-1. Details of individual behavior by round, in the Battle of the Sexes (BS), Prisoner's Dilemma (PD) and Prisoner's Dilemma with Incentive (PDI) games. The rows are pairs (from 1 to 25), the columns rounds (from 1 to 10). The color in each cell represents the medal outcomes (gold, silver, bronze; no color=no medal). The circles with various fillings are the governance options chosen by each individual: no filling=individual decision, quarter filling=punishment option, three- quarter filling=recommending leader, complete filling=enforcing leader. If a cell is shaded on top of the color, this person executed punishment. (For inter- pretation of the references to color in this figure legend, the reader is referred to the web version of this article.) The coordination and conflict of interest challenge of the BS game is illustrated by pair 2 on the left of Fig. A-1. They start out with successful

coordination on player B's preference, but then they fail to coordinate; then player B gets his/her preference again, then player A gets his/her preference twice, and they fail to coordinate again. Then they figure out how to share the benefits by taking turns, but they only manage with the help of a recommending leader, and the arrangement breaks down again in the last round. Only twopairs (7and8)manage to take turns (maximizingvalue inanequalway)over theentire10 rounds, andbothusea leader toachieve this

feat. There are many coordination failures (cooperation often breaks down because one player got his/her preference several times, violating fairness), and punishment is used 111 times, as often as not failing to produce cooperation. The center table in Fig. A-1 illustrates thePD game. Pair 6 on the left illustrates its conflict resolution challenge: playerA wins gold at the cost of

player B, then they manage to cooperate in the high-value option (with both getting silver), but only for one round before player A free-rides on the other; repeated attempts are thwarted by player A, before they then trade free riding in the last three rounds. The conflict of interest challenge is defection to the individual gold at the cost of the other player. As a result, bronze occurs more frequently than silver, there are many “no medal” outcomes, and punishment is used 144 times. In both the BS and PD games, punishment tends to lead to retaliating spirals, failing to bring about group cooperation, while leadership is much

more effective.Wealso see the importanceof fairness—subjects in thePD refuse to accept “nomedal”while theother side gets the gold, and subjects in the BS refuse to repeatedly have the other get his/her preference; thus, both situations are unstable. An impartial leader can help the PD group to avoid the temptation of free-riding (defecting to telling on the other side) to maintain silver, and taking turns to share the enjoyment of getting first preference in the BS game. Finally, we compare the PDI with the PD game: the outcome-dependent incentives do have an effect, specifically, several groups manage to stick

to a collaborative equilibrium solution for several periods in a row (pairs 1, 2, 9, 17 and 24). However, the occurrence of bronze is not significantly

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

17

changed, and retaliatory punishment is still widespread (for example, in pairs 6, 7, 10, 16 and 21). This shows why the benefit of the incentives is smaller in the PD game, with its conflict of interest, than in the MT game with its coordination challenge (as shown in Table 7). Responses to a post- game survey question showed that participants did not take turns to win the medals in the two versions of the Prisoner's Dilemma game.

Appendix 3. Studies with monetary incentives

In six additional studies with direct monetary incentives, the instructions of the original studies (with medals as incentives) were replicated with only one change: the instructions showed monetary values of each decision choice, instead of the medals. For example, instead of showing a gold medal, € 0.50 was written; similarly, instead of a silver medal, € 0.40 was written (see figure in Appendix 1). The monetary incentives used in original studies were utilized. For the six studies with direct monetary incentives, 300 additional participants were recruited from a similar pool of participants as those in the earlier reported eight studies. The results of original studies were replicated, as reported below.

0

5

10

15

20

25

30

35

40

1 2 3 4 5 6 7 8 9 10

Trivial Game Individual Decision Punishment Leadership

Leadership: M = 9.20; SD = 2.96

0

5

10

15

20

25

30

35

40

1 2 3 4 5 6 7 8 9 10

Chicken Game Individual Decision Punishment Leadership

Leadership: M = 14.40; SD = 3.58 0

5

10

15

20

25

30

35

40

1 2 3 4 5 6 7 8 9 10

Prisoner's Dilemma Game Individual Decision Punishment Leadership

Leadership: M = 15.70; SD = 2.45

0

5

10

15

20

25

30

35

40

1 2 3 4 5 6 7 8 9

Stag Hunt Game Individual Decision Punishment Leadership

Leadership: M = 15.80; SD = 2.14 0

5

10

15

20

25

30

35

40

1 2 3 4 5 6 7 8 9 10

Ba!le of the Sexes Game Individual Decision Punishment Leadership

Leadership: M = 20.10; SD = 3.14

0

5

10

15

20

25

30

35

40

1 2 3 4 5 6 7 8 9 10

Matching Game Individual Decision Punishment Leadership

Leadership: M = 19.40; SD = 4.57

Fig. A3-1. Number of times agivengovernancewas chosenper gameandover10consecutive rounds (count of occurrences per round, out of 50 subjects' choicesper round). “Leadership” combines the recommending and enforcing leadership structures, their separation is shown in Table A3-1.

Table A3-1 Leadership usage.

M (SD)

Trivial Stag Hunt Matching Prisoner's Dilemma Chicken Battle of the Sexes

Recommending leadership 6.40 (2.58) 10.80 (1.40) 10.00 (2.57) 8.20 (2.18) 8.00 (1.73) 10.00 (2.10) Enforcing leadership 2.80 (2.09) 5.00 (2.10) 9.40 (3.20) 7.50 (2.46) 6.40 (2.54) 10.10 (3.73)

Notes: Count of occurrences for 50 subjects over ten rounds.

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

18

Table A3-2a Logistic regression results: Stag Hunt.

Independent variables Cooperation

B Wald χ2 RSE p OR 95% CI OR ME

Enforcing leadership 1.79 12.63 0.50 <0.001 5.99 [2.23, 16.09] 1.67 Recommending leadership 1.00 13.54 0.27 <0.001 2.73 [1.60, 4.67] 0.89 Punishment 0.50 2.86 0.30 0.09 1.65 [0.92, 2.96] 0.38 Cooperation by other t–1 0.46 4.72 0.22 0.03 1.58 [1.05, 2.40] 0.94 Punishment by other t–1 0.29 0.92 0.30 0.34 1.34 [0.74, 2.41] 0.85

Table A3-2b Logistic regression results: Matching.

Independent variables Cooperation

B Wald χ2 RSE p OR 95% CI OR ME

Enforcing leadership 0.98 9.32 0.32 <.01 2.68 [1.42, 5.04] 1.02 Recommending leadership 0.71 4.27 0.34 <.05 2.03 [1.04, 3.97] 0.75 Punishment 0.18 0.39 0.29 .53 1.20 [0.68, 2.12] 0.22 Cooperation by other t–1 1.54 69.91 0.18 <.001 4.68 [3.26, 6.73] 1.28 Punishment by other t–1 −0.15 0.35 0.25 .55 0.86 [0.52, 1.41] 0.43

Table A3-2c Logistic regression results: Prisoner's Dilemma.

Independent variables Cooperation

B Wald χ2 RSE p OR 95% CI OR ME

Enforcing leadership 2.18 34.88 0.37 <.001 8.85 [4.29, 18.25] 1.14 Recommending leadership 0.86 4.81 0.39 <.05 2.36 [1.10, 5.09] −0.18 Punishment 0.26 0.66 0.32 .42 1.29 [0.70, 2.40] −0.78 Cooperation by other t–1 −0.21 0.94 0.22 .33 0.81 [0.53, 1.24] −0.32 Punishment by other t–1 −0.36 1.64 0.28 .20 0.70 [0.40, 1.21] −0.40

Table A3-2d Logistic regression results: Chicken.

Independent variables Cooperation

B Wald χ2 RSE p OR 95% CI OR ME

Enforcing leadership 2.23 34.99 0.38 <.001 9.28 [4.44, 19.42] 1.68 Recommending leadership 1.20 9.75 0.38 <.01 3.30 [1.56, 6.99] 0.65 Punishment 0.83 9.53 0.27 <.01 2.29 [1.35, 3.87] 0.28 Cooperation by other t–1 0.20 0.94 0.20 .33 1.22 [0.82, 1.82] 0.61 Punishment by other t–1 −0.06 0.05 0.26 .83 0.94 [0.57, 1.58] 0.49

Table A3-2e Logistic regression results: Battle of the Sexes.

Independent variables Cooperation

B Wald χ2 RSE p OR 95% CI OR ME

Enforcing leadership 1.05 13.83 0.28 <.001 2.86 [1.64, 4.98] 0.90 Recommending leadership 0.48 3.51 0.26 .06 1.61 [0.98, 2.66] 0.33 Punishment 0.06 0.06 0.26 .80 1.07 [64, 1.78] −0.08 Cooperation by other t–1 0.95 20.26 0.22 <.001 2.59 [1.71, 3.93] 0.73 Punishment by other t–1 −0.62 5.21 0.27 <.05 0.54 [0.32, 0.92] −0.06

Notes for Tables A3-2a to A3-2e: Regression parameters of the generalized estimating equations for each game with the dependent variable the decision to cooperate in the current round (yes/no). Independent variables include governance structure (with four categories) chosen in the current round, cooperation by the other player in previous round (yes/no), and punishment strategy used by the other player in previous round (yes/no). RSE=robust standard errors; OR=odds ratio; ME=marginal effects with transformed scale. GENLIN procedure in SPSS 23 was used.

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

19

Table A3-3 Game-wise group benefit from governance structures.

M (SD)

Trivial Stag Hunt Matching Prisoner's Dilemma Chicken Battle of the Sexes

Individual decision 0.99 (0.04) 0.67 (0.26) 0.51 (0.50) 0.57 (0.08) 0.45 (0.40) 0.42 (0.45) Punishment 0.55 (0.32) 0.21 (0.23) 0.30 (0.36) 0.27 (0.21) 0.38 (0.28) 0.31 (0.36) Recommending leadership 0.99 (0.03) 0.76 (0.28) 0.72 (0.45) 0.62 (0.12) 0.64 (0.32) 0.56 (0.44) Enforcing leadership 1.00 (0.00) 0.84 (0.26) 0.86 (0.35) 0.64 (0.15) 0.76 (0.17) 0.70 (0.37)

Notes: Average value achieved by a pair per round. Individual decision=individual decision making chosen by both, punishment=at least one side chooses the pun- ishment option (and no one leadership), recommending leadership=at least one side chooses a recommending leader (and no one enforcing leadership or pun- ishment), enforcing leadership=at least one side chooses an enforcing leader (and no one punishment). Maximum possible mean value=1.00.

Table A3-4a Logistic regression results: Trivial.

Independent variables Punishment

B Wald χ2 RSE p OR 95% CI OR ME

Punishment by other t–1 2.20 34.18 0.38 <.001 8.98 [4.30, 18.75] 0.56 Cooperation by other t–1 −1.40 3.70 0.73 .05 0.25 [0.06, 1.03] −1.23 Cooperation by both t–1 0.42 0.40 0.66 .53 1.52 [0.42, 5.55] −0.32

Table A3-4b Logistic regression results: Stag Hunt.

Independent variables Punishment

B Wald χ2 RSE p OR 95% CI OR ME

Punishment by other t–1 0.34 1.48 0.28 0.22 1.41 [0.81, 2.46] −0.89 Cooperation by other t–1 0.36 1.38 0.31 0.24 1.44 [0.78, 2.64] −0.88 Cooperation by both t–1 −0.13 0.20 0.29 0.65 0.88 [0.50, 1.54] −1.13

Table A3-4c Logistic regression results: Matching.

Independent variables Punishment

B Wald χ2 RSE p OR 95% CI OR ME

Punishment by other t–1 0.40 1.79 0.30 .18 1.50 [0.83, 2.71] −0.98 Cooperation by other t–1 −0.02 0.01 0.25 .93 0.98 [0.60, 1.60] −1.19 Cooperation by both t–1 N/A N/A N/A N/A N/A N/A N/A

Table A3-4d Logistic regression results: Prisoner's Dilemma.

Independent variables Punishment

B Wald χ2 RSE p OR 95% CI OR ME

Punishment by other t–1 0.97 11.39 0.29 <.01 2.64 [1.50, 4.62] −0.40 Cooperation by other t–1 0.13 0.28 0.24 .60 1.14 [0.70, 1.84] −0.82 Cooperation by both t–1 −0.17 0.25 0.34 .62 0.84 [0.43, 1.64] −0.97

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

20

Table A3-4e Logistic regression results: Chicken.

Independent variables Punishment

B Wald χ2 RSE p OR 95% CI OR ME

Punishment by other t–1 0.30 1.22 0.28 .27 1.36 [0.80, 2.33] −0.76 Cooperation by other t–1 −0.36 1.43 0.30 .23 0.70 [0.38, 1.26] −1.10 Cooperation by both t–1 0.46 2.16 0.31 .14 1.58 [0.86, 2.89] −0.69

Table A3-4f Logistic regression results: Battle of the Sexes.

Independent variables Punishment

B Wald χ2 RSE p OR 95% CI OR ME

Punishment by other t–1 1.11 12.23 0.32 <.001 3.04 [1.63, 5.66] −0.44 Cooperation by other t–1 −0.74 8.86 0.25 <.01 0.48 [0.30, 0.78] −1.37 Cooperation by both t–1 N/A N/A N/A N/A N/A N/A N/A

Notes for Tables A3-4a to A3-4f: Regression parameters of the generalized estimating equations for each game with the dependent variable punishment chosen in the current round (yes/no). Independent variables include decisions of previous round concerning punishment chosen by other player (yes/no), cooperation by other player (yes/no) and cooperation by both players (yes/no). RSE=robust standard errors; OR=odds ratio; ME=marginal effects with transformed scale. GENLIN procedure in SPSS 23 was used.

References

Abele, S., & Stasser, G. (2008). Coordination success and interpersonal perceptions: Matching versus mismatching. Journal of Personality and Social Psychology, 95(3), 576–592.

Alexander, R. D. (1974). The evolution of social behavior. Annual Review of Ecology and Systematics, 5(1), 325–383.

Antonakis, J., & Day, D. V. (2017). The nature of leadership (3rd ed.). Thousand Oaks, CA: Sage publications.

Ariely, D., Bracha, A., & Meier, S. (2009). Doing good or doing well? Image motivation and monetary incentives in behaving prosocially. American Economic Review, 99(1), 544–555. https://doi.org/10.1257/aer.99.1.544.

Arvey, R. D., Rotundo, M., Johnson, W., Zhang, Z., & McGue, M. (2006). The determi- nants of leadership role occupancy: Genetic and personality factors. The Leadership Quarterly, 17(1), 1–20.

Aureli, F., & de Waal, F. B. M. (2000). Natural conflict resolution. Univ of California Press. Axelrod, R., & Hamilton, W. D. (1981). The evolution of cooperation. Science, 211(4489),

1390–1396. Bargh, J. A., & Chartrand, T. L. (1999). The unbearable automaticity of being. American

Psychologist, 54(7), 462. Bennis, W. (2007). The challenges of leadership in the modern world: Introduction to the

special issue. American Psychologist, 62(1), 2–5. Bligh, M. C. (2011). Followership and follower-centered approaches. In A. Bryman, D.

Collinson, & K. Grint (Eds.). The sage handbook of leadership (pp. 425–436). London: Sage Publications.

Boehm, C. (1999). Hierarchy in the forest: The evolution of egalitarian behavior. Cambridge, MA: Harvard University Press.

Boone, J. L. (1992).Competition, conflict, and thedevelopmentof socialhierarchies. InB. Winterhalder,&E.A. Smith (Eds.). Evolutionary ecology and human behavior (pp. 301– 337). .

Bowles, S. (2009). Did warfare among ancestral hunter-gatherers affect the evolution of human social behaviors? Science, 324(5932), 1293–1298. https://doi.org/10.1126/ science.1168112.

Brown, D. E. (1991). Human universals. Philadelphia, PA: Temple University Press. Burke, C. S., Stagl, K. C., Klein, C.,Goodwin,G. F., Salas, E.,&Halpin, S.M. (2006).What

type of leadership behaviors are functional in teams? A meta-analysis. The Leadership Quarterly, 17(3), 288–307.

Calvert, R. (1992). Leadership and its basis in problems of social coordination. International Political Science Review, 13(1), 7–24.

Camerer, C. F. (2003). Behavioral game theory: Experiments in strategic interaction. NY: Russell Sage Foundation.

Carsten, M. K., Uhl-Bien, M., West, B. J., Patera, J. L., & McGregor, R. (2010). Exploring social constructions of followership: A qualitative study. The Leadership Quarterly, 21(3), 543–562. https://doi.org/10.1016/j.leaqua.2010.03.015.

Chagnon, N. A. (1997). The Yanomamo (5th ed.). (Fort Worth: Wadsworth). Chance, M. R. (1967). Attention structure as the basis of primate rank orders. Man, 2(4),

503–518. Chaudhuri, A. (2011). Sustaining cooperation in laboratory public goods experiments: A

selective survey of the literature. Experimental Economics, 14(1), 47–83. https://doi.

org/10.1007/s10683-010-9257-1. Clutton-Brock, T. H., & Parker, G. A. (1995). Punishment in animal societies. Nature,

373(6511), 209. Conger, J. A. (1990). The dark side of leadership. Organizational Dynamics, 19(2), 44–55. Conradt, L., & Roper, T. J. (2005). Consensus decision making in animals. Trends in

Ecology & Evolution, 20(8), 449–456. Cosmides, L., & Tooby, J. (1992). Cognitive adaptations for social exchange. In J. H.

Barkow,L.Cosmides,&J.Tooby (Eds.). The adapted mind: Evolutionary psychology and the generation of culture (pp. 163–228). .

Crawford, V. P., Gneezy, U., & Rottenstreich, Y. (2008). The power of focal points is limited: Even minute payoff asymmetry may yield large coordination failures. American Economic Review, 98(4), 1443–1458. https://doi.org/10.1257/aer.98.4. 1443.

Dawkins, R. (2006). The selfish gene. NY: Oxford University Press. De Cremer, D., & vanKnippenberg, D. (2002). How do leaders promote cooperation? The

effects of charisma and procedural fairness. The Journal of Applied Psychology, 87(5), 858–866.

De Cremer, D., & Van Knippenberg, D. (2005). Cooperation as a function of leader self- sacrifice, trust, and identification. Leadership & Organization Development Journal, 26(5), 355–369.

Devetag, G., & Ortmann, A. (2007). When and why? A critical survey on coordination failure in the laboratory. Experimental Economics, 10(3), 331–344.https://doi.org/10. 1007/s10683-007-9178-9.

Earle, T. K. (1997). How chiefs come to power: The political economy in prehistory. Stanford University Press.

Fehr, E., Fischbacher, U., & Gächter, S. (2002). Strong reciprocity, human cooperation, and the enforcement of social norms. Human Nature, 13(1), 1–25. https://doi.org/10. 1007/s12110-002-1012-7.

Fleischer, R. A., & Chertkoff, J. M. (1986). Effects of dominance and sex on leader se- lection in dyadic work groups. Journal of Personality and Social Psychology, 50(1), 94–99.

Foss, N. J. (2001). Leadership, beliefs and coordination: An explorative discussion. Industrial and Corporate Change, 10(2), 357–388.

Frackenpohl, G., Hillenbrand, A., & Kube, S. (2016). Leadership effectiveness and in- stitutional frames. Experimental Economics, 19(4), 842–863.https://doi.org/10.1007/ s10683-015-9470-z.

Friedman, J. W. (Ed.). (1992). Problems of coordination in economic activity. Boston: MA: Kluwar Academic Publishers.

Gat, A. (2006). War in human civilization (1st ed.). Oxford: Oxford University Press. Giessner, S.R.,&Schubert,T.W. (2007).High in thehierarchy:Howvertical locationand

judgments of leaders' power are interrelated. Organizational Behavior and Human Decision Processes, 104(1), 30–44. https://doi.org/10.1016/j.obhdp.2006.10.001.

Gino, F., & Mogilner, C. (2014). Time, money, and morality. Psychological Science, 25(2), 414–421. https://doi.org/10.1177/0956797613506438.

Gintis, H., Smith, E. A., & Bowles, S. (2001). Costly signaling and cooperation. Journal of Theoretical Biology, 213(1), 103–119.

Glowacki, L., & von Rueden, C. (2015). Leadership solves collective action problems in small-scale societies. Philosophical Transactions of the Royal Society: B, 370(1683), 20150010. https://doi.org/10.1098/rstb.2015.0010.

Glowacki, L., Wilson, M. L., & Wrangham, R. W. (2017). The evolutionary anthropology

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

21

of war. Journal of Economic Behavior & Organization. https://doi.org/10.1016/j.jebo. 2017.09.014.

Gneezy, U., Meier, S., & Rey-Biel, P. (2011). When and why incentives (don't) work to modify behavior. Journal of Economic Perspectives, 25(4), 191–210. https://doi.org/ 10.1257/jep.25.4.191.

Goodall, J. (1986). The chimpanzees of Gombe: Patterns of behavior. Cambridge, MA: Belknap.

Güth, W., Levati, M. V., Sutter, M., & Van Der Heijden, E. (2007). Leading by example with and without exclusion power in voluntary contribution experiments. Journal of Public Economics, 91(5–6), 1023–1042.

Hamilton, W. D. (1964). The genetical evolution of social behaviour I. Journal of Theoretical Biology, 7(1), 1–16.

Hamilton, W. D. (1975). Innate social aptitudes of man: An approach from evolutionary genetics. Biosocial Anthropology, 53, 133–153.

Harcourt, A. H., & Stewart, K. J. (2007). Gorilla society: Conflict, compromise, and co- operation between the sexes. Chicago: The University of Chicago Press.

Hardin, G. (1968). The tragedy of the commons. Science, 162(3859), 1243–1248. Heifetz,R.A. (1994). Leadership without easy answers. Cambridge,MA:HarvardUniversity

Press. Herrmann, B., Thöni, C., & Gächter, S. (2008). Antisocial punishment across societies.

Science, 319(5868), 1362–1367. https://doi.org/10.1126/science.1153808. Hobbes, T. (1651). Leviathan. Routledge. Homans, G. C. (1974). Social behavior: Its elementary forms. Oxford, England: Harcourt

Brace Jovanovich. Hooper, P. L., Kaplan, H. S., & Boone, J. L. (2010). A theory of leadership in human

cooperative groups. Journal of Theoretical Biology, 265(4), 633–646. https://doi.org/ 10.1016/j.jtbi.2010.05.034.

Johnson, A. M., Vernon, P. A., Harris, J. A., & Jang, K. L. (2004). A behavior genetic investigation of the relationship between leadership and personality. Twin Research and Human Genetics, 7(1), 27–32.

Johnson, A. M., Vernon, P. A., McCarthy, J. M., Molson, M., Harris, J. A., & Jang, K. L. (1998). Nature vs nurture: Are leaders born or made? A behavior genetic investiga- tion of leadership style. Twin Research and Human Genetics, 1(4), 216–223.

Judge, T. A., Piccolo, R. F., & Kosalka, T. (2009). The bright and dark sides of leader traits: A review and theoretical extension of the leader trait paradigm. The Leadership Quarterly, 20(6), 855–875. https://doi.org/10.1016/j.leaqua.2009.09.004.

Keeley, L. H. (1997). War before civilization. Oxford University Press. Kenrick, D. T., & Griskevicius, V. (2013). The rational animal: How evolution made us

smarter than we think. AZ: Basic Books. Kets de Vries, M. F. R. (2007). Decoding the team conundrum: The eight roles executives

play. Organizational Dynamics, 36(1), 28–44. Kets de Vries, M. F. R., & Florent-Treacy, E. (1999). The new global leaders. San Francisco:

Jossey-Bass. King, A. J., Johnson, D. D. P., & Van Vugt, M. (2009). The origins and evolution of lea-

dership. Current Biology, 19(19), R911–R916. https://doi.org/10.1016/j.cub.2009. 07.027.

Komai, M., Stegeman, M., & Hermalin, B. E. (2007). Leadership and information. American Economic Review, 97(3), 944–947.

Komorita, C., & Parks, C. D. (1996). Social dilemmas. Boulder, CO: Westview Press. Kotter, J. P. (1996). Leading change. Boston: MA: Harvard Business School Press. Kouchaki, M., Smith-Crowe, K., Brief, A. P., & Sousa, C. (2013). Seeing green: Mere ex-

posure to money triggers a business decision frame and unethical outcomes. Organizational Behavior and Human Decision Processes, 121(1), 53–61.https://doi.org/ 10.1016/j.obhdp.2012.12.002.

Lawrence, P. R., & Lorsch, J. W. (1967). New management job: The integrator. Harvard Business Review, 45(November–December), 142–151.

Lewis, H. S. (1974). Leaders and followers: Some anthropological perspectives. Reading, MA: Addison-Wesley.

Lord, R. G. (1977). Functional leadership behavior: Measurement and relation to social power and leadership perceptions. Administrative Science Quarterly, 22(1), 114–133.

Lord, R. G., De Vader, C. L., & Alliger, G. M. (1986). A meta-analysis of the relation between personality traits and leadership perceptions: An application of validity generalization procedures. Journal of Applied Psychology, 71(3), 402–410.

Lowie, R. H. (1948). Some aspects of political organization among the American abor- igines. The Journal of the Royal Anthropological Institute of Great Britain and Ireland, 78(1/2), 11–24. https://doi.org/10.2307/2844522.

Manz,C.C.,&Sims,H.P. (1991). Superleadership: Beyond themythofheroic leadership. Organizational Dynamics, 19(4), 18–35.

Markóczy, L., & Goldberg, J. (1998). Management, organization and human nature: An introduction. Managerial and Decision Economics, 19(7–8), 387–409.

McNamara, P., & Trumbull, D. (2007). An evolutionary psychology of leader-follower rela- tions. Nova Publishers.

Meder, A. (2007). Great ape social systems. In W. Henke, I. Tattersall, & T. Hardt (Vol. Eds.), Handbook of paleoanthropology. Vol. 2. Handbook of paleoanthropology (pp. 1235–1271). Berlin: Springer.

Megargee, E. I. (1969). Influence of sex roles on the manifestation of leadership. Journal of Applied Psychology, 53(5), 377–382.

Meggitt, M. J. (1967). The pattern of leadership among the Mae-Enga of New Guinea. Anthropological forum. Vol. 2. Anthropological forum (pp. 20–35). Taylor & Francis.

Messick, D. M., Wilke, H. A. M., Brewer, M. B., Kramer, R. M., Zemke, P. E., & Lui, L. (1983). Individual adaptations and structural change as solutions to social dilemmas. Journal of Personality and Social Psychology, 44(2), 294–309.

Mintzberg, H. (1978). Patterns in strategy formation. Management Science, 24(9), 934–948.

Mulder, L. B., Van Dijk, E., De Cremer, D., & Wilke, H. A. (2006). When sanctions fail to increase cooperation in social dilemmas: Considering the presence of an alternative

option to defect. Personality and Social Psychology Bulletin, 32(10), 1312–1324. Nicholson, N., & White, R. (2006). Darwinism—A new paradigm for organizational be-

havior? Journal of Organizational Behavior, 27(2), 111–119. Nowak,M.A.,&Sigmund,K. (2005).Evolutionof indirect reciprocity.Nature, 437(7063),

1291–1298. Nyquist, L. V., & Spence, J. T. (1986). Effects of dispositional dominance and sex role

expectations on leadership behaviors. Journal of Personality and Social Psychology, 50(1), 87–93.

O'Connor, J., Mumford, M. D., Clifton, T. C., Gessner, T. L., & Connelly, M. S. (1995). Charismatic leaders and destructiveness: An historiometric study. The Leadership Quarterly, 6(4), 529–555.

O'Gorman, R., Henrich, J., & Vugt, M. V. (2009). Constraining free riding in public goods games: Designated solitary punishers can sustain human cooperation. Proceedings of the Royal Society of London B: Biological Sciences, 276(1655), 323–329. https://doi. org/10.1098/rspb.2008.1082.

Pearce, C. L., & Manz, C. C. (2005). The new silver bullets of leadership: The importance of self- and shared leadership in knowledge work. Organizational Dynamics, 34(2), 130–140. https://doi.org/10.1016/j.orgdyn.2005.03.003.

Potters, J., Sefton, M., & Vesterlund, L. (2005). After you—Endogenous sequencing in voluntary contribution games. Journal of Public Economics, 89(8), 1399–1419.

Potters, J., Sefton, M., & Vesterlund, L. (2007). Leading-by-example and signaling in voluntary contribution games: An experimental study. Economic Theory, 33(1), 169–182.

Price, D. (1981). Nambiquara leadership. American Ethnologist, 8(4), 686–708. https:// doi.org/10.1525/ae.1981.8.4.02a00020.

Rutte, C. G., & Wilke, H. A. M. (1984). Social dilemmas and leadership. European Journal of Social Psychology, 14(1), 105–121.

Rutte, C. G., & Wilke, H. A. M. (1985). Preference for decision structures in a social dilemma situation. European Journal of Social Psychology, 15(3), 367–370.

Sahin, S. G., Eckel, C., & Komai, M. (2015). An experimental study of leadership in- stitutions in collective action games. Journal of the Economic Science Association, 1(1), 100–113. https://doi.org/10.1007/s40881-015-0010-6.

Samuelson, C. D., & Messick, D. M. (1986). Inequities in access to and use of shared resources in social dilemmas. Journal of Personality and Social Psychology, 51(5), 960–967.

Samuelson, C. D., Messick, D. M., Rutte, C., & Wilke, H. A. M. (1984). Individual and structural solutions to resource dilemmas in two cultures. Journal of Personality and Social Psychology, 47(1), 94–104.

Schelling, T. C. (1960). The strategy of conflict. Cambridge, MA: Harvard University Press. Shane, S. (2009). Introduction to the focused issue on the biological basis of business.

Organizational Behavior and Human Decision Processes, 110(2), 67–69. https://doi.org/ 10.1016/j.obhdp.2009.10.001.

Shapira, Z. (Ed.). (1997). Organizational decision making. Cambridge, MA: Cambridge University Press.

Silk, J. B. (2005). The evolution of cooperation in primate groups. In H. Gintis, S. Bowles, R. Boyd, & E. Fehr (Eds.). Moral sentiments and material interests: On the foundations of cooperation in economic life (pp. 43–73). Cambridge, MA: MIT Press.

Silk, J. B. (2007). The strategic dynamics of cooperation in primate groups. Advances in the Study of Behavior, 37, 1–41.

Smith, J. E.,Gavrilets, S.,Mulder,M.B.,Hooper, P. L.,Mouden,C. E.,Nettle,D., ... Smith, E. A. (2016). Leadership in mammalian societies: Emergence, distribution, power, and payoff. Trends in Ecology & Evolution, 31(1), 54–66. https://doi.org/10.1016/j. tree.2015.09.013.

Smith, J. M. (1982). Evolution and the theory of games. Cambridge, UK: Cambridge uni- versity press.

Sober, E., & Wilson, D. S. (1998). Unto others: The evolution and psychology of unselfish behavior. Cambridge, MA: Harvard University Press.

Sorauren, I. F. (2000). Non-monetary incentives: Do people work only for money? Business Ethics Quarterly, 10(4), 925–944. https://doi.org/10.2307/3857840.

Spisak,B.R.,O'Brien,M.J.,Nicholson,N.,&VanVugt,M. (2015).Nicheconstructionand the evolution of leadership. Academy of Management Review, 40(2), 291–306. https:// doi.org/10.5465/amr.2013.0157.

Trice, H. M., & Beyer, J. M. (1986). Charisma and its routinization in two social move- ment organizations. Research in Organizational Behavior, 8, 113–164.

Trivers,R. L. (1971). Theevolutionof reciprocal altruism. The Quarterly Review of Biology, 46(1), 35–57.

Uhl-Bien, M., Riggio, R. E., Lowe, K. B., & Carsten, M. K. (2014). Followership theory: A review and research agenda. The Leadership Quarterly, 25(1), 83–104. https://doi. org/10.1016/j.leaqua.2013.11.007.

Van Lange, P. A., Balliet, D. P., Parks, C. D., & Van Vugt, M. (2014). Social dilemmas: Understanding human cooperation. Oxford University Press.

Van Vugt, M., Hogan, R., & Kaiser, R. B. (2008). Leadership, followership, and evolution: Some lessons from the past. American Psychologist, 63(3), 182–196. https://doi.org/ 10.1037/0003-066X.63.3.182.

Van Vugt, M., Jepson, S. F., Hart, C. M.,& De Cremer, D. (2004). Autocratic leadership in social dilemmas:A threat to group stability. Journal of Experimental Social Psychology, 40(1), 1–13.

Vohs, K. D., Mead, N. L., & Goode, M. R. (2008). Merely activating the concept of money changes personal and interpersonal behavior. Current Directions in Psychological Science, 17(3), 208–212. https://doi.org/10.1111/j.1467-8721.2008.00576.x.

von Rueden, C., & Van Vugt, M. (2015). Leadership in small-scale societies: Some im- plications for theory, research, and practice. The Leadership Quarterly. https://doi. org/10.1016/j.leaqua.2015.10.004.

Vroom, V. H., & Jago, A. G. (2007). The role of the situation in leadership. American Psychologist, 62(1), 17.

Wang, L., Zhong,C.-B.,&Murnighan, J.K. (2014). The social andethical consequences of

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

22

a calculative mindset. Organizational Behavior and Human Decision Processes, 125(1), 39–49. https://doi.org/10.1016/j.obhdp.2014.05.004.

Weber, R., Camerer, C., Rottenstreich, Y., & Knez, M. (2001). The illusion of leadership: Misattribution of cause in coordination games. Organization Science, 12(5), 582–598.

Wilson,R.K.,&Rhodes,C.M. (1997). Leadershipandcredibility inn-personcoordination games. Journal of Conflict Resolution, 41(6), 767–791.

Wit, A. P., &Wilke, H. A.M. (1988). Subordinates' endorsement of an allocating leader in a commonsdilemma:Anequity theoretical approach. Journal of Economic Psychology, 9(2), 151–168.

Wit, A. P., Wilke, H. A. M., & Van Dijk, E. (1989). Attribution of leadership in a resource management situation. European Journal of Social Psychology, 19(4), 327–338.

Witt, U., & Schwesinger, G. (2013). Phylogenetic footprints in organizational behavior. Journal of Economic Behavior & Organization, 90, S33–S44. https://doi.org/10.1016/j. jebo.2012.12.011.

Woodburn, J. (1982). Egalitarian societies. Man, 17(1), 431–451. Wrangham, R. W. (1980). An ecological model of female-bonded primate groups.

Behaviour, 75(3), 262–300.

Wrangham, R. W., & Peterson, D. (1996). Demonic males: Apes and the origins of human violence. NY: Houghton Mifflin.

Yamagishi, T. (1986). The provision of a sanctioning system as a public good. Journal of Personality and Social Psychology, 51(1), 110–116.

Yoo, Y., & Alavi, M. (2004). Emergent leadership in virtual teams: What do emergent leaders do? Information and Organization, 14(1), 27–58.

Yukl, G. (1999). An evaluation of conceptual weaknesses in transformational and char- ismatic leadership theories. The Leadership Quarterly, 10(2), 285–305.

Yukl, G. (2013). Leadership in organizations (8th ed.). NY: Pearson. Zaccaro, S. J., Kemp, C., & Bader, P. (2004). Leader traits and attributes. In J. Antonakis,

A. T. Cianciolo, & R. J. Sternberg (Eds.). The nature of leadership (pp. 101–124). . Zaleznik, A. (1977). Managers and leaders: Are they different. Harvard Business Review,

55(5), 67–78. Zhang,Z., Ilies,R.,&Arvey,R.D. (2009).Beyondgenetic explanations for leadership:The

moderating role of the social environment. Organizational Behavior and Human Decision Processes, 110(2), 118–128. https://doi.org/10.1016/j.obhdp.2009.06.004.

M.G. Ahmad and C. Loch The Leadership Quarterly xxx (xxxx) xxx–xxx

23

View publication statsView publication stats