Distributive Preferences and Need-based Justice in Networks

Network structures affect the ability to articulate ones need claims and the power to enforce ones preferences. Integrating ideas from the theory of social exchange networks, the sociology of justice, and behavioral economics, subproject B1 Distributive Preferences and Need-based Justice in Networks studies variations in the recognition of needs in dyadic negotiations that permit allocations to network members outside the dyad. Using two three- node networks the triangle and the three-line as examples of equal- and unequal-power networks, we showed in the 1st funding period that distributions negotiated in dyads do actually include outsiders and that the willingness to do so depends on the subjects social value orientations and the power structure of the network. We have also shown that subjects do take into account heterogeneous need levels. In the 2nd funding period, we further elaborate dimensions of network transparency, that is, the amount of information available about a node, in order to study scope restrictions in need recognition. In the first step, we explore the effect of transparency in the network on the scope of need satisfaction by introducing heterogeneity in two ways: First, we study the recognition of needs if individuals contribute to a joint production task before they bilaterally negotiate the distribution of the output. Second, we use arbitrary criteria to form group identities before subjects are allocated to networks in which they bilaterally negotiate distributions. In the next step, we study the effect of varying the size and density of the network on the recognition of needs. A third experiment increases the conflict level by substituting resource scarcity for abundance, which implies that at least one need claim cannot be satisfied. In a last step, we explore the congruence of attitudes and behavior and study the generalizability of the laboratory findings by comparing them to survey data.

Decision about the distribution of collective resources require a foundation in justice principles. Starting from the premise of equal human dignity, the classical criteria of distribution according to equality or merit fall short of their promise if they result in the non-satisfaction of an individual need. This holds for all those who need more than or contribute less than average, which implies that these principles must be amended by needs. Needs are heterogenous, individual and subjective. Furthermore, what counts as a need in a specific society is often unclear. In consequence, the need principle is manipulable, vulnerable to attacks and dependent on public discourses influenced by power inequalities. The project studies by means of social scientific laboratory experiments the extent to which needs are considered in distributive decisions and the factors that influence the probability of need satisfaction. These questions are analyzed in networks consisting of three subjects who are connected by communication channels via which exchange is either possible or not, depending on the realized network structure. Subjects are assigned an individual need threshold which has to be fulfilled in order to participate in a subsequent stage of the experiment, in which additional income can be generated. This operationalization creates a condition which implements the concept of capabilities under laboratory conditions. A decision is taken if two of the three subjects agree on a specific distribution of the joint endowment, consisting of 24 units, over the three subjects. Each distribution realizes a specific principle of justice (e.g., equality 8-8-8). The experiments reveal that the probability of need satisfaction declines with increasing need level. In particular, need levels below the equal distribution are satisfied more often, while those above the equal distribution are satisfied less often. This effect is accentuated by transparency: For need thresholds above the equal distribution the probability of satisfaction drops toward zero if the threshold is unknown to others. In contrast, allocations are geared toward the threshold if their level is known to others, which results in lower allocations for low need thresholds compared to the opaque condition. A third factor is group membership: Outsiders' needs are particularly seldom covered if they are higher than the equal share. Further findings are that self-regarding subjects can enforce their preference on other-regarding subjects in the determination of the share allocated to third subjects and that needs are more likely satisfied if they are articulated and motivated.