FormalPara Key Points

We outline a social identity approach to sport psychology based upon social identity and self-categorization theories.

We argue that this approach has the potential to make a significant, unique, and powerful contribution to the analysis of an array of phenomena and behaviors that are at the heart of sport and sport psychology.

In particular, the approach offers a comprehensive analysis of the ways in which individual psychology both structures and is structured by the dynamics of group life; this involves seeing groups not simply as features of sporting contexts but rather as elements that can be, and often are, incorporated into a person’s sense of self and, through this, become powerful determinants of their sport-related behavior.

1 Introduction

The purpose of this article is to provide researchers and practitioners in the field of sport with a source of reference for a perspective that has over the past 4 decades become a dominant approach to the study of groups in social psychology. Incorporating social identity theory [1, 2] and self-categorization theory [3–5], the insights from this social identity approach have, until relatively recently, been largely overlooked within the domain of sport [6]. Nevertheless, as other authors have recently intimated [7, 8], we believe that the approach has the potential to make a significant, unique, and powerful contribution to the analysis of an array of sport-related behaviors. This is due primarily to the fact that the approach offers a comprehensive analysis of the ways in which individual psychology both structures and is structured by the dynamics of group life. We expand upon these claims in this review by first introducing the concepts of social identity and self-categorization and outlining a number of the key theoretical aspects of the social identity approach. From this we then abstract four key lessons for sport psychology: two fundamental, and two applied. We then conclude by outlining implications for a range of additional topics in sport to which the approach could reasonably be expected to make (and in some cases has already made) an important contribution.

2 The Social Identity Approach: Foundations and Hypotheses

2.1 Social Identity Theory

Social identity theory was developed in the 1970s with a view to understanding the psychology of intergroup relations, particularly as played out against the backdrop of large-scale social conflict. In the wake of World War II, dominant social psychological approaches sought to explain events such as the Holocaust with reference to individual-level factors (e.g., personality), yet for Henri Tajfel and colleagues [1, 2, 9, 10], theorizing that failed to take account of the role that social context and group memberships play in such phenomena was necessarily limited. Speaking to this point, the classic ‘boys camp studies’ of Sherif [11] and Sherif et al. [12] had previously shown that intergroup competition had the capacity to dramatically transform well-adjusted middle-class boys from good friends into vicious opponents. Yet Tajfel [9] wondered whether realistic competition was necessary in order for such transformations to occur. This issue was explored in the so-called ‘minimal group studies’, which sought to identify the minimal conditions that might lead members of one group to discriminate in favor of that group and against another.

Tajfel’s [9] studies were designed to exclude a range of factors previously considered to play an essential role in driving group behavior—for example, a history of cooperation or conflict, personal liking or animosity, and interdependence. Having been assigned to groups with no prior meaning, participants had to award points (signifying small amounts of money) to an anonymous member of their group (the ‘ingroup’) and to a member of another group (the ‘outgroup’). Individual self-interest and personal economic gain were ruled out because participants never assigned points to themselves. What transpired was that participants tended to deviate from a strategy of fairness by awarding more points to ingroup members. Furthermore, rather than maximizing absolute ingroup gain, the boys favored a strategy of maximizing relative gain by out-doing the outgroup.

The key finding from the minimal group studies was that, in even the most stripped-down of conditions, the mere act of categorizing themselves as group members was sufficient to encourage individuals to make ingroup-favoring responses [10, 13]. In subsequently formulating their social identity theory, Tajfel teamed up with John Turner [2] to propose that, after being categorized in terms of a group membership, and having then defined themselves in terms of that social categorization, individuals seek to achieve or maintain self-esteem by positively differentiating their ingroup from a comparison outgroup on some valued dimension. This drive for positive distinctiveness means that when people’s sense of who they are is defined in terms of social identity (as ‘we’ and ‘us’ rather than just ‘I’ and ‘me’) they strive to see ‘us’ as different from, and preferably as better than, ‘them’ in order to feel good about who they are and what they do [14–17]. This is evident in sporting rivalries, such as those between the Scottish and the English in soccer [18], as well as in fans’ motivation to engage in derogatory behaviors towards opposition hockey fans [19]. Among many other things, it also helps to explain how the New Orleans Saints football team’s 2009 Super Bowl-winning season galvanized the citizens of its city in the rebuilding that followed the devastation wrought by Hurricane Katrina in 2005 [15].

Importantly, though, social identity theory does not conceptualize ingroup favoritism as an inevitable consequence of group membership. Indeed, Tajfel and Turner [2] subsequently identified three key factors that bear upon the emergence of ingroup favoritism. These are (1) the extent to which individuals identify with an ingroup and internalize that group membership as an aspect of their self-concept (e.g., by identifying themselves as University of Cincinnati football fans [17]); (2) the extent to which the prevailing context provides ground for comparison between groups (e.g., Cincinnati’s performance and status can be gauged by their league performance relative to other teams, such as Notre Dame); and (3) the perceived relevance of the comparison outgroup, which itself will be shaped by the relative and absolute status of the ingroup (e.g., the relevance of Notre Dame for Cincinnati fans was heightened by the rivalry that developed following Cincinnati coach Brian Kelly’s move to Notre Dame in 2009). Individuals are therefore likely to display ingroup favoritism only when an ingroup is central to their self-definition, when a given comparison is meaningful, and when the outcome of that comparison is in some sense contestable.

More generally, social identity theory focuses on the role of three key elements in structuring social identification and hence group behavior: the perceived stability and legitimacy of an ingroup’s position in relation to other groups, and the perceived permeability of group boundaries [2, 20]. If status relations between groups are perceived to be insecure (because they are seen to be either unstable or illegitimate and hence open to negotiation), their members will typically engage in social competition with a view to defining themselves as superior on status-defining dimensions. This is a characteristic of most sporting competitions in which teams vie for superiority knowing that a higher-status position (e.g., as league champions) is possible. However, if relations are secure, high-status groups will tend to define their group on status-defining dimensions (e.g., “we are the best team”), and low-status groups will be more likely to embrace a strategy of social creativity in which they define themselves on status-irrelevant dimensions (e.g., “we may not be the best, but we best represent the true spirit of the game”). This latter behavior was demonstrated in a study of players of an ice hockey team in last place in a nine-team regional league (with one win and 21 losses) [21]. Although players saw their opponents as more skilled, they maintained positive distinctiveness, and compensated for their lower performance status by proclaiming their opponents as more ‘dirty’ than themselves (a finding that was particularly intriguing, as coaching staff observed that the team had the second-worst record for “good behavior” in the league). At the same time, if boundaries between groups are permeable (so that it is possible to move between groups), members of low-status groups will tend to engage in a strategy of social mobility such that they disavow social identity and seek to join the high-status group—a phenomenon evidenced by the vitality of the transfer market in professional sport [22], as well as in examples from amateur sport (e.g., the case of Olympic gold medalist Eric Lindros who in junior Canadian ice hockey refused to play for the lowly ranked Sault Ste. Marie Greyhounds).

2.2 Self-Categorization Theory

Self-categorization theory extends social identity theory by delving more forensically into the dynamic workings of the self. In the process, it generates three key insights. The first is that social identity is what allows group behavior to occur at all. As Turner [3] put it, “social identity is the cognitive mechanism that makes group behavior possible” (p. 21). More specifically, he argued that group behavior is underpinned by a process of self-stereotyping—which he termed “depersonalization”—through which the self comes to be perceived as categorically interchangeable with other ingroup members. Self-stereotyping leads those who define themselves in terms of a particular social identity (e.g., as “us Bruins” under coach John Wooden’s tenure [23]) both (1) to seek to discover the meaning of the category (e.g., that Bruins value integrity, reliability, and loyalty) and then (2) to strive to enact that meaning (so that, as a Bruin, “I value integrity, and strive to be reliable and loyal”) (see also Burns [15] and Whigham [18]).

A second core insight is that the self reflects the operation of a categorization process in which, depending on context, people see themselves as either sharing category membership with others (i.e., in terms of a shared social identity, ‘us’), or not [4]. As we discuss below, sport fandom is a particular form of social identity, wherein we see fellow fans as ‘us’ and opposing fans as ‘them’, leading to both ingroup and outgroup homogeneity effects; for example, of the form ‘all Yankees fans are alike’ (p. 174; [24]).

Whether, and which, social identities are activated is understood to be an interactive product of a particular categorization’s fit and a person’s readiness to use it [25]. For example, a young woman is more likely to define herself as a gymnast (sharing category membership with other gymnasts) if this self-categorization maps on to her understanding of the patterns of similarity and difference between gymnasts and non-gymnasts (e.g., in terms of skills and interests), and if this group has become meaningful to her (e.g., through participation in youth events [26]). For example, this process underpins women rugby players’ identity as rugby players [27], and athletes’ preferences for individual or team sports [28]. Similarly, Kopietz and Echterhoff [29] noted that Germans became more willing to embrace their national identity after this had been made fitting and available by Germany’s hosting of the FIFA 2006 Soccer World Cup.

Beyond this, though, a third insight is that shared social identity is the basis for mutual social influence [30]. When people perceive themselves to share group membership with other people in a given context (e.g., as players on, or fans of, the same team), they are motivated to strive actively to reach agreement with those other people and to co-ordinate their behavior in relation to activities that are relevant to that identity. Moreover, it is through individuals’ identification of and conformity to norms that are seen to be shared with others in a particular context that their potentially idiosyncratic views become socially organized and consensual. Indeed, at the most general level, categorization-based influence is central to the transformation of low-level individual inputs (e.g., a player’s personal skills or aspirations, a fan’s personal opinions) into higher-order group products (e.g., a team tactic, a culture of professionalism, shared beliefs about a player’s ‘star’ status). In effect, then, these processes provide the all-important psychological conduit between personal cognitions, perceptions, and behaviour and organized collective endeavor [31, 32].

These ideas feed into the analysis of an array of sporting phenomena. In particular, as we will see, they help to explain the dynamics of sports group behavior, sports group formation and development, sports group support and stress appraisal, and sports group leadership. A key point here is that people’s willingness to engage constructively with others—and to respond positively to such engagement—is structured by the self-categorization process. Moreover, such engagement is more likely to be effective when, and to the extent that, both parties share a sense of social identity, because this provides a basis for them to perceive and relate to each other as similar (as both part of self) rather than different (as self vs. other).

3 Four Key Lessons for Sport Psychology

3.1 Social Identity is the Basis for Sports Group Behavior

A foundational idea for research informed by social identity theorizing is that groups are not simply external features of the world that provide a setting for individual behavior. Instead, they shape and qualitatively transform psychology through their capacity to be internalized as part of a person’s sense of self. In this regard, the self is conceptualized as a context-sensitive process in which self-definition varies as a function of the prevailing social setting [33]—e.g., in the case of women’s rugby, so that a player sees herself as an athlete on the field of play, but as a woman off it [27].

This means that in some contexts a person’s sense of self is defined in terms of personal identity such that they see themselves, and behave, as a unique individual; but in many other contexts, self-definition and behavior is underpinned by social identity—the person’s sense that they share group membership with other members of their ingroup [28]. Thus, while psychological theorizing in the sporting domain generally tends to construe the self in purely personal terms (i.e., as referring to a person’s understanding of themselves as ‘I’ and ‘me’), social identity theorizing asserts that the self can be, and often is, defined in social terms (as ‘we’ and ‘us’). It is thus social identity that underpins people’s sense that they are part of a particular team, squad, club, or organization [20] and, more generally, social identities provide the psychological substrates for the translation of the contents of individual minds into the collective aspects of sport [34, 35].

Critically, too, it is people’s capacity to define the self (i.e., self-categorize) in terms of social identity that allows them to act as group members [3]. For example, in a game of rugby between England and France [26], it is the capacity of players to see themselves not simply as unique individuals (Sophie, Rochelle, Laura) but also—and more importantly—as representatives of the same team (e.g., England) that allows them to have a meaningful game. In particular, this is the basis for critical forms of context-specific knowledge (e.g., about whom to pass to and whom to tackle) and expectations (e.g., about who will likely pass to you, who will likely tackle you).

Another core assumption of social identity theory is that when the self is defined in terms of social identity, people strive to define the ingroup as positively distinct from comparison outgroups [2, 36]. As suggested above, how (and whether) they achieve this will depend on prevailing features of social context (in particular, whether boundaries are impermeable and whether group relations are secure). Indeed, differences in the degree to which soccer players and soccer fans see group boundaries as permeable—and in associated levels of social identification—help explain why players (and coaches) are far more likely to transfer their affections from one club to another than are fans (especially those of the ‘die-hard’ rather than the ‘fair weather’ variety [16, 17, 37, 38]). This was seen, for example, in the derogatory reactions of University of Cincinnati football fans to the aforementioned loss of their coach, Brian Kelly, after his departure to Notre Dame in 2009 [17]. More generally too, there is clear evidence that the ways in which teams understand (and market) themselves is informed by sensitivity to such factors [36, 39–45].

In this way too, we can see that the definition of social identity is both relational and comparative (p. 16; [2]), such that the meaning of a given ingroup membership is always partly defined in relation to some comparison outgroup. Thus, what it means to be ‘an Arsenal fan’ in soccer is defined by a motivation to positively distinguish that group from salient comparison outgroups (e.g., Chelsea fans). So, when compared with league champions Chelsea, an Arsenal fan might proclaim her team a ‘home-grown side’ that did not buy its way to success. But when compared with a team in a lower league (e.g., Bournemouth), she might proclaim Arsenal to be a well-resourced ‘Premiership team’. Likewise, in comparison with shot-putters, javelin throwers may define themselves as fast, but in comparison with sprinters they are more likely to define themselves as strong [46, 47].

3.2 Social Identity is the Basis for Sports Group Formation and Development

Of course, the idea that groups are a fundamental feature of human sociality is hardly new to sport [48]. Indeed, not only do we organize people in groups, teams, squads, and organizations, but there are grounds for arguing that all sporting activity is in some sense structured by group memberships. This is true, for example, even in the case of the lonely long-distance runner who indulges in this pastime only to experience ‘splendid isolation’ [49]. After all, she is likely to have learned to enjoy this pursuit through a process of group-based learning or coaching (e.g., at school); she is likely to have a sense of this as an activity others also find meaningful (e.g., her peers—noting that friends often sign up for and participate in long-distance races together, to create a sense of solidarity); and the shoes and clothes she wears are likely to be branded, sold, and bought in ways that affirm the meaningfulness and positivity of her social identity (e.g., as a runner or a woman) [39, 43, 50, 51].

Speaking to the ubiquity of group concerns, empirical and anecdotal evidence also suggests that researchers, practitioners, and sports coaches all recognize the importance of group cohesion in many (if not most) domains of sporting performance [52–54]. Indeed, to this end, teams that are interested in success have been observed to invest in ever-more elaborate programs devised to enhance group bonding [55].

But what exactly is it that makes a team cohere? According to the classical group dynamics literature, individuals cohere within a group when (and to the extent that) (1) they find other group members attractive, (2) they feel the group satisfies their affective needs, (3) they feel they are similar to other group members, or (4) they see the benefits of group membership as outweighing the costs [56, 57]. In short, according to this traditional view, individuals become, and remain, group members to the extent—and only to the extent—that they believe it is in their personal interests to do so. An important feature of this classical analysis is that it effectively renders the group superfluous, since groups are simply the aggregation of the personal motivations that bind individuals to them. Indeed, according to this literature, to the extent that groups no longer meet their members’ personal needs, they should simply disband and disintegrate.

However, in practice, people’s decisions to join or leave groups are not well predicted by the degree to which those groups satisfy their personal needs. Indeed, if this were the case it would be hard to explain why people make a point of being ‘die-hard’ fans [58–61] who stick with their team ‘through thick and thin’. How, for example, in the 2010–2011 season, did the Cleveland Cavaliers continue to sell out their home games, while racking up the longest losing streak in the National Basketball Association’s history (26 straight games)?

Speaking to such phenomena, for self-categorization theorists, the core process that binds group members to one another (whether team member to team member, player to coach, or fan to team) is depersonalization, through which the self comes to be understood in terms of a social identity that is shared with others [4, 62]—e.g., as in the case of the ‘Bad Blue Boys’, fans of a Croatia-based team in Australia’s soccer league [63]. Indeed, rather than seeing similarity, attraction, and satisfaction as determinants of group formation [64, 65], self-categorization theory sees them as outcomes of social identification. As we have noted, following the devastation from Hurricane Katrina in 2005, the City of New Orleans’s identification with its team’s rise from underperformers to triumphant 2009 Superbowl winners was then reflected not just in the struggles but also in the subsequent hopes and dreams of its citizens [15]. That is, rather than signing up for groups we feel similar to and good about, we feel similar to and good about the groups for which we (psychologically) sign up [66].

Social identity is not just a basis for joining groups, it is also a basis for ongoing group development and commitment to their causes. This proposition has received widespread empirical support within the fields of social and organizational psychology—in particular, from experimental studies in which support for a group and its activities is shown to increase following manipulations that serve to enhance participants’ sense of shared identity [34, 62, 66]. In the sporting domain too there is evidence that social identification rather than the satisfaction of personal needs is the key motivator of continued group support. In particular, Wann and Branscombe [61] examined commitment to baseball and basketball teams in the USA and observed that for fans who lived close to their team’s playing base, support was unrelated to the success or failure of the team. Indeed, for these fans, withdrawing support from their teams when it failed was simply not an option. Why? Because it was their team. As Green Bay Packers fan Steve Gay put it, “The Packers are like your children. You don’t love them because they’re good. You love them because they’re yours” [67].

Along related lines, subsequent research has shown that fans with high levels of team identification [40, 68–75] are likely (1) to attend more games, (2) to pay more for tickets, (3) to purchase more team merchandise, and (4) to buy more team sponsors’ products [43, 51, 76–78]. Interestingly too, Turner et al. [79] observed that failure to win a collective game can actually make group members more committed to the group, not less, as the personal interests approach would predict. Indeed, as Queenan [80] notes, the ultimate proof that one is a ‘true’ fan is not to be able to say that you were there for the team at its zenith (when the personal benefits—e.g., of being able to ‘bask in reflected glory’—are high [58, 81, 82]), but to be able to say that you were there at the team’s nadir (when the personal benefits are low [83, 84]). At the same time, though, it is precisely the fact that “they share their grief with millions of kindred spirits” that allows fans to “suffer the slings and arrows of outrageous fortune” (p. 108; [80]).

Building on these observations, one might also wonder why athletes not only stick with their team regardless of results, but in fact sacrifice their own self-interests for the ‘greater good.’ For example, why in 2012 did British cyclist Chris Froome forego his own opportunity for victory in the Tour de France to help his teammate, Bradley Wiggins, win? Why in 2013 did Gregory Campbell of the Boston Bruins throw himself into the path of a slapshot and then play on with a broken leg until his team (which was a player down at the time) returned to full strength? Such actions not only fail to meet personal needs, they run directly counter to them.

Of course, one might argue that such behavior is a natural part of professional sport—and that players are paid (often handsomely) to make the necessary sacrifices for their team. Again, however, evidence generally fails to bear this classical economic analysis out [85]. Indeed, it is not uncommon for the best-paid professional athletes to be criticized for failing to show the required amount of enthusiasm for their team. For example, despite having one of the best group of players at soccer’s 2012 Euro Championships, Holland’s early elimination from the tournament followed poor performances and reports of in-fighting [86]. Speaking more systematically to such observations, in a comprehensive study of US major league baseball [87], the presence of highly paid stars had a significant negative impact on both the financial and the on-field performance of the team for which they play. Additional research [88] has also noted that baseball teams that are built on a more even distribution of observed talent perform better than those with a mixture of highly able and less able players. Indeed, the conclusion from an analysis of basketball data is that “when teams need to come together, more talent can tear them apart” (p. 1; [89]).

Having the best players and paying them the highest wages is thus no guarantee of either the desire for group success or its achievement. Instead, it appears that people’s willingness to sacrifice personal interests for those of the group is driven primarily by identification with the group (and its perceived needs and goals) and that it is this shared social identity that helps turn a team of champions into a champion team [90–92].

3.3 Social Identity is the Basis for Sports Group Support and Stress Appraisal

A key focus in sport has been on the impact of stress on performance and on the ways in which this can be effectively managed [93, 94]. Much of this research has been informed by the work of Lazarus and colleagues [95, 96], in which the experience of stress is generally understood to be structured by both primary appraisal (a sense that stressors in the environment are in some way threatening to the self and to well-being) and secondary appraisal (a sense of the extent to which social support is available to help deal with that strain) [97–99].

Importantly, the social identity approach posits that both primary and secondary appraisals are structured by social identity and self-categorization processes [34, 100, 101]. In the first instance, this is because whether or not a stimulus is perceived as threatening to self depends in part on how the self is defined. For example, whether or not one perceives the poor performance of a soccer team as a source of stress depends on the degree to which one identifies with that team [102–105]. Indeed, among other things, this means that a loss suffered by a rival team can be a source of eustress (e.g., in the form of schadenfreude [18, 106, 107]) rather than distress.

Relatedly, and second, the likelihood of receiving effective support from others generally depends upon those others being (and being perceived to be) representative of a shared social identity. This means that we are more likely to offer help to people we perceive as belonging to an ingroup that is salient for us, and that we are more likely to receive help from those who perceive us as belonging to an ingroup that is salient for them. These points have been highlighted not only in a large body of work examining task performance and workplace stress [100, 108], but also in research examining bystander helping behavior [109].

Most relevant to the present discussion, work has explored the role that social identity plays in the helping behavior of soccer fans. For example, in research by Levine and colleagues [110, 111], fans of Manchester United were asked to think about their team before being asked (for a spurious reason) to go to another building. Then, on their way to that building, a man (a confederate) tripped and fell over in front of them. Significantly, shared social identity was manipulated by having the man in question wear either a Manchester United shirt, the shirt of a rival team (Liverpool), or a plain red t-shirt. As predicted, the actor’s apparel had a major bearing on the level of help participants gave him. When he was wearing a Manchester United shirt (i.e., so that he was an ingroup member), 92 % of participants stopped to ask if he needed help or helped him directly. Yet when he wore the Liverpool shirt or a plain shirt, on average, only 32 % of participants offered any form of help (the typical response being to fail to notice his accident). However, evidence of the context sensitivity of self-categorization, and of the significance of this for the provision of support, was provided in a follow-up study in which participants were induced to think of themselves not as Manchester United fans but as soccer enthusiasts. With the ingroup thus redefined, the Liverpool fan was now ingroup not outgroup, and the support he received was elevated to the same level as the fan in the Manchester United shirt. So, whereas in Experiment 1 a minority (30 %) of participants had offered him help, now a majority (70 %) did.

Such research highlights the point that there is a significant social dimension to stress. This accords with recent evidence that in addition to various personal sources of stress, organizations and groups can also be a source of considerable stress in sporting contexts (e.g., in the form of ‘personality clashes’ with team mates or coaches or problems with selection procedures and outcomes [112–114]). In this way, the experiences of stress in sport are generally observed to be bound up with the (intra- and inter-) dynamics of group life.

Yet, while groups are sources of stress, they can also be the key to overcoming it. In this regard too, it is apparent that, as its progenitors concede [115], a key limitation of the stress and coping approach is that its conceptualization of the stress process is inherently individualistic—seeing the stress sufferer as someone who makes essentially personal judgments about the status of the strains that he or she confronts (i.e., of the form, “Is this threatening for me?”, “Can I cope?”). Yet while an individual’s personal identity will sometimes be salient when making stress-related judgments, there will also be a range of situations—especially in sport—where a person’s sense of self will be informed primarily by their group membership and where the questions they ask will therefore be of the form “Is this threatening for us?”, “Can we cope?”.

Along these lines, a large body of research shows that processes of both primary and secondary appraisal are critically structured by the nature of salient social identity and levels of social identification. For example, Levine and Reicher [116] showed that whether female athletes found a given medical condition (a facial or a knee injury) stressful depended on whether they were induced to categorize themselves as women or as athletes. When thinking of themselves as athletes, they were more concerned by the knee injury; when thinking of themselves as women, they were more concerned by the facial injury. Significantly too, this also determined the likelihood of their seeking treatment for the injury in question.

Relatedly, research also shows that people’s experience of social support and their well-being is heavily structured by shared social identification. For example, among patients recovering from heart surgery, identification with their family was a powerful predictor of experiencing useful forms of support from family members and of post-surgical life satisfaction [108]. More generally, social identification has also been found to be a major determinant of a person’s ability to stave off depression in response to adverse life events [117, 118]. Speaking to the physiological underpinnings of such processes, Häusser et al. [119] also observed that the presence of others with whom we identify helps to attenuate neuroendocrine stress reactions, because there is a sense that “we are going through this together” (p. 976).

Work by Haslam and Reicher [120] also found that when shared social identity within a group declines, this leads not only to heightened intragroup tensions, but to elevated cortisol levels that are associated with increased burnout. On the other hand, in groups where shared social identity increases, individuals are not only buffered from such responses [121], but also report positive stress-related experiences that are associated with resilience, indomitability, and salutogenesis [122, 123].

3.4 Social Identity is the Basis for Sports Group Leadership

Although a wealth of evidence speaks to the fact that social identity is the basis for group behavior, formation and development, and support and stress appraisal, probably the key sports-related phenomenon to which these ideas have been applied elsewhere is leadership [66, 124–127]. This is clearly a central focus for sports research and one that reflects the significance of coaches, managers, and captains to sporting processes and outcomes [128]. Indeed, whenever a sporting group experiences considerable success, it is customary for researchers and commentators to try to abstract the lessons that might account for their success—whether in basketball [23], soccer [129], curling [130], or ‘big time college sports’ [131]. Speaking to the wisdom of these efforts, there is also considerable evidence that particular coaching behaviors do appear to promote positive psychosocial and behavioral outcomes among athletes [132–135].

Although a range of sport-specific models of leadership have been proposed—in particular, the Multidimensional Model of Leadership [136] and the Mediational Model of Leadership Behavior [137]—more recent work has focused on the importance of transformational leadership [138–143], with its distinct emphasis on the leader-follower relationship. Yet, despite the promise that transformational leadership might hold for sport—e.g., in advancing beyond simple transactional or exchange theories of leadership—critics have questioned both its focus on charismatic leadership [144] and its rather vague theoretical underpinnings [145]. This criticism relates not to the concept of charisma itself, but to the suggestion that effective leaders possess some specific set of leadership qualities that distinguish them from others—a focus that harks back to classical ‘Great Man’ theorizing [146, 147]. In contrast, social identity theorists have attempted to move beyond the individualism that is inherent in such perspectives, while also grounding their analysis of leadership in well-specified theoretical models [124].

The core claim of the social identity approach is that leadership is not, as commonly supposed, a property that resides in a person’s character as an individual, but rather one that results from a leader’s capacity to embody what a particular group means to its members in any given context (in formal terms, to be prototypical of a shared social identity [148]). It is then on this basis that leaders exert influence over other group members. In short, this is because it is only when leaders are seen as embodying ‘who we are’ and ‘what we want to be’ that their leadership engenders followership [111, 125, 149, 150].

In these terms too, it is a shared social identity that imbues a leader with charisma and the power to influence followers [124]. Thus, to the extent that leaders act in the interests of the group (with all other variables held constant), they are indeed perceived to be more charismatic [151, 152], as well as more trustworthy [153] and effective [154]. Thus, rather than simply trying to identify the ‘special stuff’ that makes leaders and coaches great, a social identity approach to leadership focuses on the processes through which leaders and followers come to see each other as part of a common ingroup [90, 152].

As one example of this theory being translated into action, the Australian field hockey coach Ric Charlesworth drew on empirical evidence that social identity in a team can be enhanced by having a random rather than an appointed leader [155] to develop and implement the concept of a leaderful (but captain-less) team [156]. Lending further support to the notion that effective sporting leadership is about cultivating a sense of ‘we’ and not ‘I’, research is now emerging in which these aspects of leadership have been explicitly addressed and tested. For example, in the context of the London 2012 Olympic Games, researchers [150] observed that successful leaders strove to mobilize team members by consistently articulating values in the media that communicated a positive, distinctive, and enduring sense of team identity. Relatedly, experimental research has shown that, because it enhances team members’ social identification, the confidence leaders express in a team is a powerful determinant of basketball players’ success in competition [90] (see also Fransen et al. [157]).

4 New Frontiers: Towards a Broader Application of the Social Identity Approach

The key claim of the present review is that the social identity approach has considerable merit as a framework for advancing our understanding of sport-related behavior. Specifically, we have shown that it offers a new perspective on sports group behavior, sports group formation and development, and helps us to better understand the group-related dynamics of stress appraisal, mutual support, and effective leadership.

However, there is a range of other important topics to which this approach could readily be (and in some cases has already been) applied. As we have noted, considerable research has applied this approach to the study of fans [73, 78, 158]. In addition, research has drawn upon social identity theorizing to understand how fans and players deal with changes to a club’s overall structure [159], how fans’ identification with teams is influenced by the ‘unscrupulous’ acts of athletes [75], and how identity is formed in teams [160–162]. Related research also points to the significance of athletic [163–170] and exercise [171–173] identity, and this has fed into research that draws on notions of social identity to understand both exercise behaviour [174] and health behaviour [175], as well as event participation (e.g., in the Gay Games [176]).

Along lines suggested above, the social identity approach can also be used to inform the current interest in personal [177] and collective [178] resilience in sport. Here, research has demonstrated that social identity and membership of multiple groups enhances resilience in the face of pain, and also aids heart rate recovery in athletes following winter sports training [179]. Furthermore, the extent to which athletes can bounce back from performance failures may be dependent on them receiving feedback from a person who is seen to be ‘one of us’ (i.e., an ingroup member) [180, 181]. In light of this evidence, and work demonstrating that social identity predicts better coping with transitions and life challenges [182, 183], one would also expect social identity to play a role in sport injury recovery [184], the dynamics of sporting burnout [185], and the negotiation of career transitions in sport [186]. This is because, provided a strong sense of social identity can be maintained in the context of dealing with these challenges, athletes should derive positive self-esteem, receive support, and collaborate with others to resist stressors, thus increasing the likelihood that they will be able to cope with (and adjust to) these various challenges. Evidence that social identity and self-categorization processes structure social influence within groups [30] also leads us to expect that the social identity approach could be used to better understand the dynamics of (1) the coach–athlete relationship (with outcomes depending on the evolving nature of their shared identity) [128], (2) coach–athlete attributional divergence and conflict [187, 188], (3) sports-related impression formation [189], (4) sports communication [190], and athletes’ and fans’ affective states and emotional contagion [191].

Finally, in the wake of recent work demonstrating that social identity underpins both pro- and anti-social behaviors towards team mates and opponents (e.g., reflecting intergroup dynamics and identity-related norms) [192], the approach could also be used to shed light on moral development and integrity in sport. Indeed, although social identity can be a motivational force for good, it can also lead to aggression, hooliganism, and violence towards sporting outgroups [102, 104, 105, 193–198]. Moreover, in this context, violent action toward outgroups can come to be seen as legitimate—such that, for example, injurious ‘tackles’ and slurs against opposition players are seen as ‘fair’ and ‘nothing personal’ (at the same time that these same acts by opponents are seen as entirely unreasonable). It might help understand, for example, why Chris Nilan (formerly an enforcer for the Montreal Canadians, Boston Bruins, and New York Rangers) can speak out forcefully against school bullying, but describe his own violent behavior in sport as legitimate (because “What is done on the ice and in hockey is a totally different environment than going to school” [199]). Extrapolating further, social identity processes can also be seen to contribute to the legitimization of cheating and other forms of abuse [17, 200–203]. For example, in the recent ‘Penn State Case’—where criticism over how an institution handled a scandal involving a former coach led to the dismissal of its current coach and resignation of its president—one sees that more highly identified fans showed more support for retention of both the coach and the president because they were more supportive of actions that defended their team [203]. Significantly too, research also points to ways in which sports-related problems can be successfully tackled [197, 198], as when England fans at Euro2004 created an identity within which hooliganism was considered non-prototypical, leading to the marginalization of those fans who sought to create disorder [198].

5 Conclusion

The social identity approach has broad applicability to issues and topics that are at the heart of sport and sport psychology. Having made this point in relation to key dimensions of sporting activity, we also highlighted an array of recent research developments to which the approach appears to have clear relevance—including fan behavior and identification, resilience, adaptation, and moral development and integrity. It is nevertheless true that, until relatively recently, the potential usefulness of the social identity approach had been largely neglected in sport research. One advantage of this is that, within sport, the insights the approach affords are both novel and fresh. Accordingly, they offer a new generation of researchers opportunities to embrace a powerful new paradigm for understanding sport-related behavior.

Our contention is that much of this power derives from the fact that the approach not only moves beyond the (often implicit) individualism of prevailing theory but also provides a coherent, integrated, and richly evidenced framework with which to structure both theoretical inquiry and practical application. At heart, this framework centers on the realization that sport is rarely, if ever, a purely personal pursuit. Rather, it is one that draws upon and enriches the human capacity for (and appreciation of) collective endeavor. It is for this reason that—whether as theorists or practitioners—we neglect the psychology of ‘us-ness’ at our peril. And it is for this reason that the social identity approach to sport psychology is so sorely needed.