The impact of group membership on cooperation and norm enforcement: evidence using random assignment to real social groups

Due to incomplete contracts, efficiency of an organization depends on willingness of individuals to take non-selfish actions, e.g., cooperate when there is no incentive to do so, or punish inefficient actions by others. Organizations also constitute a social boundary, or group. We investigate whether this social aspect of organizations has an important benefit, fostering unselfish cooperation and norm enforcement within the group, but whether there is also a dark side, in the form of hostility between groups. Our experiment provides the first evidence without the confounding effect of self-selection into groups. Individuals are randomly assigned to different platoons during a four-week portion of officer training in the Swiss Army. We conduct choice experiments - simultaneous prisoner's dilemma games, with and without third-party punishment - in week three. Random assignment significantly increases willingness to cooperate with fellow platoon members. Assignment does not lead to hostility, in the sense of vindictive punishment of outsiders, but does affect norm enforcement, enhancing willingness to enforce a norm of cooperation towards fellow platoon members. This suggests that the social aspect of organizations motivates efficient behavior even when ordinary incentives fail, and helps explain practices designed to foster social ties or group identification within an organization.