A proposed model of between-group helping: an identity-based approach.

Researchers have taken note of Katz and Kahn's (1978) argument that performance is multidimensional and includes "innovative and spontaneous" behaviors that transcend formal role requirements (Borman and Motowidlo, 1993; Brief and Motowidlo, 1986; Organ, 1997; Smith et al., 1983). Whereas previous authors have examined extra-role behaviors performed by individuals, we develop a new, group-level construct called between-group helping (BGH), defined broadly as helping behaviors that work groups display toward other groups.

The growing popularity, frequency of use, and potential benefits associated with work teams (1) suggest that applying helping behavior concepts to groups could be useful. Such an application not only acknowledges the importance of discretionary behavior for organizational success, but also addresses the fact that teams have become more important in organizations and do not operate in a vacuum, but instead coexist with other teams. As groups and teams become increasingly common in our organizations, we should try to learn how to elicit beneficial outcomes such as teams working together to achieve organizational effectiveness.

Considering helping behavior as a group-level phenomenon heeds the argument made by meso-level theorists to consider whether phenomena that are typically examined at the individual level of analysis have isomorphic or discontinuous counterparts at the group level (e.g., House et al., 1995; Kozlowski and Klein, 2000; Rousseau, 1985). For example, individual-level mental models have been studied as shared or team mental models (Klimoski and Mohammed, 1994) and the application of individual self-efficacy has been broadened to include a team's collective efficacy (e.g., Bandura, 1982; Pescosolido, 2001). Like other constructs that are functionally different across levels, group-level BGH is not merely the sum of individual members' helping behavior.

Although previous researchers have studied the topic of intergroup helping, we conceptualize the phenomenon differently and place it specifically in the context of team-based organizations. For example, scholars have investigated factors related to conflict and cooperation between groups of children (Sherif et al., 1961), high school students (e.g., Gaertner et al., 1994), ad hoc laboratory groups (Dovidio et al., 1995; Dovidio et al., 1998; Gaertner et al., 1990) and managers (Blake and Mouton, 1961). We build on the theoretical principles contained within this body of research and focus on how group identity works with intragroup and organizational-level factors to predict BGH. Consequently, the addition of BGH to the extra-role performance literature complements previous work and builds upon it by enlarging the work performance domain.

A MODEL OF BETWEEN-GROUP HELPING (BGH)

Definitions and Boundary Conditions

We define BGH as intentional, discretionary acts of helping group members perform in order to assist one or more other groups. The range of relevant behaviors that teams could display can fall into one of two broad categories: information sharing (e.g., communicating about work strategies, suggesting how to avoid pitfalls) and assistance with task completion (e.g., providing materials, directly helping another group meet its goals). A key element of BGH is that work group members direct time and energy away from their own activities and apply them toward helping another team accomplish its tasks. Such an investment can be small (e.g., making a suggestion about task strategy), moderate (e.g., taking on a portion of another group's workload) or require considerable effort (e.g., helping a group redefine its operating procedures). Whatever the scope of the behavior, BGH is extra-role, meaning that it is not an explicit part of a team's assigned goals and is performed without the expectation of explicit compensation by the organization or reciprocation by other teams.

We argue that the actual helping will be carried out by one or more members of the group. However, for helping behaviors to be classified as BGH, members must be acting on behalf of their group, that is, as group representatives. When members engage in BGH, they are doing so with their group's consent. Such consent can be explicit and develop after members decide how shifting their efforts away from the group's own workload will impact their goal achievement. For example, a team may need to consider whether it can afford to invest valuable member resources to provide intergroup assistance in the face of upcoming deadlines. Alternatively, the group's consent for displays of BGH can be implicit. If norms for cooperative behavior develop within a group, members can act knowing that teammates consider BGH behaviors to be acceptable, or even preferable. To the extent that such cooperation norms are strong, consent from group members will be more easily inferred.

Because the act of BGH demonstrates a group's shared willingness to provide help to another group, it is of no great concern which team member or members actually carry out the behavior. BGH results from the understanding that any member of the team could perform the behavior. What is important is how the group members view themselves and the target when providing the help. If they view themselves as representatives of their group who are helping one or more representatives of another group, the behavior would be classified as BGH. However, if they see themselves and the target as individuals without ties to their respective teams, the behavior would be classified as individual-level helping behavior. Because BGH is an intergroup phenomenon, it falls under Sherif's definition of intergroup behavior: "Whenever individuals belonging to one group interact, collectively or individually, with another group or its members in terms of their group identification, we have an instance of intergroup behavior" (1966: 12). Therefore, the appropriate level of theory and analysis for BGH is the group and it is meaningful to discuss between-group variation in displays of BGH behaviors.

BGH is likely to occur between groups that meet a number of criteria. First, teams must be intact, continuing work units with interdependent members who interact with one another regularly. Their membership should be relatively stable, full-time and well-defined. These characteristics serve to facilitate the necessary within-team processes that lead to a strong group identity (see model below), a necessary precursor to BGH. Second, groups must have the autonomy to decide how and when they will meet their objectives. Members of teams whose operations are tightly controlled by external sources (e.g., team leader, management) will lack the ability to shift their effort away from tasks set for them and work to assist other groups. Third, groups must operate within an environment that allows for interactions with other groups. If a work team is isolated and does not come in contact with other groups, that group will have no opportunity to direct team-based helping behaviors at appropriate targets. Given these constraints, an ideal context for BGH to occur would be an organization staffed by self-managed work teams.

Social Identity in Groups

Social identity refers to those aspects of an individual's self-concept that are based on membership in emotionally significant social categories or groups (e.g., gender, nationality, social class, work team) (Turner, 1985; Turner et al., 1987). Three social psychological theories that have guided work on the role of social identity in group phenomena are relevant. The first, social identity theory (SIT; Hogg and Abrams, 1988; Tajfel, 1982; Tajfel and Turner, 1986), uses the concept of social identity to explain group processes and intergroup relations. A basic premise of SIT is that individuals derive self-worth (i.e., a social identity) from their group memberships; because "social identities have important consequences for the way individuals evaluate themselves and others, groups and their members are motivated to adopt strategies for achieving or maintaining intergroup comparisons that favor the in-group, and therefore the self" (Hogg, 1996: 67). This is accomplished through two mechanisms. The first, categorization, is a basic cognitive process people use to make sense of their surroundings. It occurs when people draw distinctions between stimuli and assign those stimuli to relevant categories (Hogg, 1996; Terry et al., 2000). Individuals, including the self, are assigned to these categories based on different criteria, such as group membership. The categorization process serves to clarify intergroup boundaries by emphasizing the differences between in-group and out-group members. The second mechanism, self-enhancement, builds on the distinctions that are drawn from categorization and involves an individual's need to make favorable comparisons of the self to others. Analogous to ethnocentrism, self-enhancement entails creating and maintaining group norms and stereotypes that favor the in-group and enhance one's social identity in relation to relevant out-groups (Hogg, 1996).

The second theory based on social identity, self-categorization theory (SGT; Turner et al., 1987), places greater emphasis on the cognitive categorization process group members experience (Messick and Mackie, 1989). Hogg and Terry (2000) described SCT as an extension of SIT that outlines the relationship between self-concept and group behavior, as well as the social cognitive processes that generate social identity effects. Whereas SIT deals with the effects of social identity on intergroup processes, SCT considers its effects on intragroup phenomena.