Exclusion Aversion – Evidence from a Minimal Social Participation Game

In an experimental study, participants are placed in a situation in which they must trade off efficiency for inclusion. Groups of three participants are formed from a student sample. The group must complete a task that requires only two group members. One group member is randomly selected to serve as the decision maker (DM) for the group. The DM automatically participates in solving the task. The DM’s task is to determine how many and which of the other two players (s)he wants to include in solving the task. Including a second group member is for free; the unnecessary inclusion of a third player comes at a certain price. Upon successful completion of the task, an amount of 90 experimental currency units (ECUs) minus the price for the inclusion of a third group member, if applicable, are distributed equally among all three group members, regardless of how many and which group members were included in completing the task. This rule ensures equality of payoffs. Before the actual inclusion task begins, a group creation phase takes place (Eckel & Grossman 2005). The three group members are introduced as the “orange group” and are encouraged to chat with one another for two minutes. Then, the group must solve an unincentivized coordination problem by communicating via chat. The coordination task consists of three questions, two historical and one numeric. The questions are tricky, but some participants may know the right answers. Please note that this task is not about finding the correct answers, but rather about coordinating a possible answer. After selecting one answer for each question, the groups will be forwarded to the next page. If a group does not coordinate within a certain time frame, we will track that groups as having not successfully solved the coordination game, but will still forward this group to the next page. The inclusion task is visualized as a two-sided scale with a weight of 60 kg on the right side. Each member of the group is represented by a game figure. The DM’s and another group member’s game figures are assigned a symbolic weight of 40 kg, while the weight of a third member’s figure depends on the specific treatment (either 10 or 30 kg). The game figure of the DM is positioned by default on the left side of the scale. To complete the task and receive a collective reward of 90 ECUs, the DM is instructed to distribute a minimum of 60 kg of weight on the left side of the scale, represented by the other group members’ figures. In all treatments of the experiment, it would be sufficient to include one additional group member, which would solve the task without additional costs. The DM may opt to include the third group member on the left side of the scale for a specified price depending on the treatment (either 3 or 15 ECU). Before making their decisions, all participants are informed of their randomly assigned roles (DM or “passive member”). Note that the third group member, depending on the treatment, either has a weight of 10 kg and therefore is useless for completing the task; or has a weight of 30 kg and therefore has the same value as the other 40-point player for completing the task, even though (s)he has a lower weight. To evaluate the effect of responsibility in the inclusion decision, we also ask the passive participants how they would make decisions if they were the DM. In addition to monetary incentives for completing the task, social participation is implemented in a minimalistic manner as a social activity, namely a variant of Pictionary (i.e. drawing and guessing terms). The social activity comprises the opportunity to play and chat for four minutes. However, before the task, participants are informed that there will be a social activity only for the included group members and they are not told about the specific details of the social activity (i.e., that it is Pictionary). The social activity involves no further monetary incentives (which would favor a certain group composition). Thus, the (non-)inclusion of the third group member cannot be attributed to the expectation of increasing group payoffs. Members not included in the completion of the task are excluded from the social activity and have to wait for four minutes. If the task has not been completed (i.e., the DM placed less than 60 kg on the left side of the scale), the whole group has to wait for four minutes. The monetary rewards (90 ECU - Price for inclusion of 3rd player if applicable) are distributed equally among all group members, regardless of the specific outcome. Consequently, the potential motivations for the inclusion decision are limited to the trade-off between efficiency and inclusion concerns. Apart from the binary inclusion decision, we also elicit the hypothetical willingness to pay for inclusion. Participants are also asked to state the motives behind their decisions (for passive participants: their hypothetical decisions). We elicit participants’ norms by asking about personal normative beliefs, normative expectations, and empirical expectations. Thus, we can track down potentially norm-driven behavior. Using neutral terms to describe the rules of our experiment and implementing an abstract, minimalist representation of social participation, we expect our results to be conservative as compared to inclusion needs present in real-world social contexts. We add a standard dictator game, in which participants distribute an amount of money between to either themselves or a non-profit organization supporting inclusion. This allows us to check if participants who are inclusive in our task are also willing to donate more money to inclusive charities than non-inclusive participants. The objective of our study is to examine the conditions under which individuals are willing to sacrifice their groups’ monetary rewards to avoid excluding someone. There are three between-subjects treatments: As compared to the baseline treatment, we vary the cost of inclusion (baseline: 3 ECU, high cost: 15 ECU) and the weight of the third group member (baseline: handicap: 10 kg; small handicap: 30 kg). Since all other motives are ruled out by design, costly inclusion of the useless/unnecessary third player is interpreted as exclusion aversion. Treatment effects are analyzed by comparing the share of participants who include the third group member for a certain price. We control for stated motives, norms, experiences with inclusion and sociodemographic data.