Social Loafing and Group Development: When "i" Comes Last

The present research examined social loafing in groups across different stages of group development. Individuals working in newly formed groups worked harder in the group setting than alone, although this difference was non-significant. However, individuals working in groups that were at the midpoint or end of their existence performed better when working by themselves. This latter finding is consistent with traditional findings on social loafing. Overall, the study suggests that social loafing is affected by the group's developmental cycle and suggests that early in a group's life when social identity is higherno social loafing will occur. INTRODUCTION One of the earliest focuses of research (Knight, 1924; Triplett, 1898) and a continuing concern of business and industry (Steers & Porter, 1979) is the effect that grouping people together has on productivity. Questions concerning this effect have taken on many forms. Are individuals more productive when working in groups than when working alone? Are certain types of groups more productive than other types? Are some tasks more effectively handled by groups than by isolated individuals? The group has alternatively been praised (Zajonc, 1965) and damned (Steiner, 1972) for its influences on productivity. Using the classic Ringelmann (1927) study as a foundation, the reputation of the group as a contributor to productivity has been severely tarnished by research over the last couple of decades. This research has shown that individual productivity declines when individuals work in groups (Ingham, Levinger, Graves & Peckham, 1974; Kerr & Bruum, 1981; Latane, Williams & Harkins, 1979). The effect has been demonstrated by measuring group output such as rope pulling, shouting, clapping, constructing paper moons, etc. This phenomenon has been given such unflattering labels as "social loafing" (Latane et al., 1979), the "free rider effect" (Kerr & Bruun, 1983) or the "sucker effect" (Kerr, 1983). Formally, social loafing is defined as "the tendency to reduce one's effort when working collectively compared with coactively on the same task (Karau & Williams, 1993, p. 683)." The most common explanation for social loafing is that people in groups are not identifiable, or more precisely, that individual production cannot be associated with specific individuals. Responsibility is diffused as individuals "hide in the crowd" (Latane et al., 1979). In support of this position, Williams, Harkins, & Latane (1981) found that when productivity was clearly associated with individuals, social loafing was reduced. In a meta-analytic review of the social loafing literature, Karau and Williams (1993) also found that social loafing decreased when evaluation potential was constant across individual and group working conditions. A second explanation for social loafing focuses specifically on the task. The position is that loafing occurs because people find the task unimportant, uninteresting, and uninvolving. The group offers them the opportunity to reduce their involvement in these tasks because there is little monitoring of individual efforts. Zaccaro (1984) reported that social loafing was reduced when the task was an attractive one. More directly, Brickner, Ostrom & Harkins (1986) reduced social loafing by using a task that was involving and personally relevant to group members. The reduction in loafing occurred even when individuals were not identifiable in the group. Similarly, Karau and Williams (1993) reported that social loafing was reduced when the task was of high valence. Based on their meta-analytic review, Karau and Williams (1993) offered a new model to explain social loafing – the Collective Effort Model (CEM). CEM rests on an economic-based expectancy value theory of effort (Vroom, 1964). CEM suggests that an individual's work level is determined by the perception of the instrumentality of one's personal efforts. Individuals work to the extent that they view (1) their efforts as benefitting group performance, (2) group performance as being translated into group outcome, and (3) group outcome as resulting in individual outcome. The studies included in Karau & Williams' (1993) meta-analysis generally involved having participants enter the lab and working on a single task as individuals and/or in a group, with characteristics of the group, task, or context being manipulated or measured. Much of the social loafing work has viewed groups as static units, and the research has concentrated on identifying a social loafing "effect" (Williams, Harkins & Latane, 1981). This research generally has involved a single measure of productivity from groups at a single point of time. Karau and Williams (1993) explicitly ackowledge this limitation by recommending that a future direction of research should be to examine the long-terms effects of working collectively. The purpose of the present paper, then, was to examine the effects of working in groups over time on social loafing. Long-term effects are important because there is growing evidence that groups are dynamic units, and that "effects" found at one point in a group's developmental cycle may not be found at other points. For example, Gersick (1988) observed a variety of work groups over time and found group productivity was highest during the midpoint of task work. Worchel, Grossman & Coutant (1994) found that minorities were able to influence group decisions during the latter stages of group life, while these same minorities were rejected during early stages of group development. These and other findings suggest that a more complete understanding of group phenomena and group dynamics will result from studying groups over an extended period of time. There have been several models (Moreland & Levine, 1988; Tuckman, 1965; Tuckman & Jensen, 1977; Worchel, Coutant-Sassic & Grossman, 1992) that suggest that groups develop through predictable stages, and that the dynamics of the group are influenced by the developmental cycle. In general, these models suggest that the initial focus of groups is on establishing a clear identity and ensuring group cohesion and uniformity. Only later do groups focus on productivity issues and meeting the needs of individual group members. According to the model proposed by Worchel et al. (1992), groups progress through a six stage developmental process. A period of discontent and a precipitating event lead to the development of the new group. The third stage actually the first stage for groups with prescribed membership and clear boundaries involves group identification. The group becomes very concerned with drawing clear ingroup-outgroup boundaries. A central dogma or theme for the group may be established. Group norms and structure are identified, and leadership is centralized. Competition and conflict with outgroups is invited, and the group avoids opportunities to cooperate or reduce intergroup conflict. Within the group, conformity is demanded and dissent is punished. The group may demand that members demonstrate their commitment to the group through personal sacrifices or initiation rites. At the individual level, members make public demonstrations of their loyalty to the group. The group becomes an important part of the individual's identity. As the group establishes its independent identity, attention is turned toward group productivity. Group goals and tasks are identified. Distinctions are made between members based on the ability to help the group achieve goals. Leaders become task oriented and less attention is paid to the socio-emotional climate of the group. Interaction with the outgroup becomes less antagonistic. Members realize that the group cannot exist in isolation, nor can group doctrine remain so extreme. During the previous two stages, the focus was on the group. During the individuation stage, attention shifts to the individual group member. Individuals begin to negotiate with the group to expand task efforts to meet personal goals. Individuals demand personal recognition; equity norms are favored. Group members base their satisfaction with the group on their personal views of what they deserve from the group. Individuals, rather than the group, may recruit new members who will help the individual achieve personal goals. Cooperative interaction with outgroups is desired. Eventually, these individualistic concerns lead the group to disintegrate during the final stage of decay. This model has been used to study diverse phenomenon such as minority influence (Worchel et al., 1994), perceptions of group homogeneity (Worchel et al., 1992), bonus preferences (Rothgerber, Worchel, Day, & Goodwin, 1995), and leadership preferences (Day, Worchel, Goodwin, Rothgerber, & Lamb, 1995), but it also has implications for social loafing. In the beginning of a group's life, when members focus on the group and having a positive social identity, group members should be motivated to work harder in a group setting and be more productive when working as a group than as individuals. Conversely, at the end of a group's life, members should be more individualistic, concerned with personal identity, and more productive when working as individuals. The traditional explanations of social loafing, reduced identifiability and task enjoyment, however, would not predict any differences in loafing over time. The idea that social identity and social loafing are inversely related is not original. Indeed, Worchel, Rothgerber, Day, Hart, and Butemeyer (1998) have argued that social loafing will decrease as social identity increases. This argument stems from Karau & Williams' (1993) CEM model that suggests that an individual's social identity may be enhanced to the extent that his or her group performs well. This suggests that individuals may resist the temptation to loaf when they perceive that their efforts can directly benefit the group. Worchel et al. (1998) conducted three experiments that supported the general prediction that group productivity would be enhanced (i.e. social loafing reduced) by factors that increase group categorization and the importance of the group to members' social identites. In the first experiment, undergraduates were divided into groups of four and worked on a task that involved manually making chain links from construction paper (see method section). Individuals worked in isolation and in a group setting, the order being counterbalanced. To manipulate the likelihood of group categorization and the importance of the group to members' social identitites, subjects were placed into a group that was either referred to as a collection of individuals, a group explicitly labelled a group expecting future interaction, or a group that worked for a bonus. Consistent with previous research where individuals work as a collection of individuals, a significant loafing effect was found in the collective condition. No significant loafing resulted from the other two conditions; in fact, there was tendancy for subjects in the reward condition to work harder in the group context. It was argued that no loafing occurred in the latter two conditions because expecting future interaction and working for a reward enhanced perceptions of being in a group and the importance of group membership. A second experiment used a similar methodology as the first except that individuals were placed into either a collective condition or into one of several interdependent groups. Results were consistent with the first experiment in that the collective condition produced more loafing than the interdependent groups presumably because the incentive (regardless of its type) served to enhance the classification of the members as an ingroup and increased the relevance of the group for the individual's social identity. Self-report measures of social identity were consistent with this explanation. The third experiment investigated the role of social identity on social loafing in a slightly different context. Subjects worked in a group in the presence or absence of an outgroup, and they wore or did not wear a group uniform. The presence of an outgroup was expected to enhance the prominence of the ingroup in the individual's social identity. Wearing a uniform was expected to make the group more salient and enhance group categorization. The authors reasoned that the effect of enhancing group categorization on productivity would depend on whether the group was a salient component of the individual's social identity. When the group is not a salient component of an individual's social identity, then, as predicted by deindividuation (Diener, 1979), loafing will increase when group categorization is enhanced. Conversely, when the group is a salient component of an individual's social identity, then, as predicted by social identity theory (Tajfel & Turner, 1986), loafing will be attenuated when group categorization is enhanced. In support of this reasoning, subjects in the group uniform condition increased their group productivity more than those in the no group uniform condition when the outgroup was salient, whereas this effect was reversed when the outgroup was not salient. Although Worchel et al. (1998) demonstrated that increasing social identity leads to a decrease in social loafing, the purpose was not to assess social loafing in a group developmental context. That is, only single measures of loafing were taken at one point in time. The purpose of the present study, then, was to examine whether social loafing varies over time in groups. In the present study, groups worked together on several tasks over three time periods. Social loafing was assessed in the beginning, middle, or end of a group's developmental cycle. It was predicted that loafing would be heightened at the end of a group's life and eliminated during the beginning of a group's life and furthermore, that social identity would decrease at the end of a group's life.