Altruism with Social Roots: An Emerging Literature
Pablo Branas-Garza y Maria Paz Espinosa
(see Hoffman et al., 1996). Recently, there has been a stream of papers
seeking a complementary result (that which relates social integration
with greater altruism). To our knowledge, there are very few papers
focussing on the link between latent4 social networks and altruism5.
As we will show in this section, there are two key ingredients in this
literature: the network elicitation procedures and the matching proc-
ess between subjects. The latter is controlled in order to explore so-
cial issues: integration, friendship, favoritism, etc.
Let us introduce some concepts.
Definition 1 (Fi) Network of friends is the set ({Fi}i∈N, N). In words,
each participant i declares who his/her friends are within the set N.
We denote by Fi the set of friends and by fi = |Fi| the number of friends.
The members of N not in Fi will be called strangers6 for i; the set of
strangers for i is thus: Si = N \ Fi.
Example 1 F1 = {2, 3}, F2 = {1}, F3 = {1, 4}, F4 = {3}, F5 = φ, F6 = {2, 3}.
Figure 1 represents this network.
Two additional definitions will be useful: the first connects the exist-
ing literature on the dictator game with this new literature7 and the
second is based on network measures.
4 These experiments do not create or promote social networks between participants; they
just elicit the social network existing prior to the experiment.
5 We do not include here designs of the “come to the lab with your friend” type (see Reuben,
2006 for an extensive discussion). Observe that these designs do not elicit a complete
network.
6 Note that Si contains the set of friends of i’s friends not in Fi, or more precisely the set of
k > 1 neighbors for i, while Fi includes all k = 1 neighbors. In many cases, k = 2 for
instance, these neighbors are not strangers at all (see Vega-Redondo, 2005).
7 See Dufwemberg and Muren (2006) for a discussion.
248