Sunbelt XXXI International Network for Social Network ... - INSNA

Network Effects On Interest Rates In Online Social LendingLerner, Juergen; Brandes, Ulrik; Nick, BoboOnline Social NetworksFinancial Networks, Event Data, Dynamic Networks, prosper.comTHURS.AM1The Prosper Marketplace (prosper.com) facilitates peer‐to‐peer lending by matching potential borrowers with potential lenders using an auction mechanism.Borrowers can post loan requests and lenders can bid on a request specifying a minimum interest rate and an amount they are willing to lend. In this talk weanalyze the impact of social information on the probability of funding, on the final interest rate, and on the credit default risk. While externally determinedcredit grades strongly influence these outcomes, there is a large variation within classes. We hypothesize that potential lenders use social information ‐including group membership and endorsements from other users ‐ to trust some borrowers more than it is suggested by their credit grade. Furthermore weanalyze whether this social behavior is rational in the sense that it leads to lower credit default rates.Network Evolution And The Beauty AdvantageO'Connor, Kathleen; Gladstone, EricKnowledge NetworksStatus, Personality, Laboratory Experiment, Cognitive ScienceTHURS.AM1Current theories of network evolution presume that actors consciously seek to occupy advantageous positions in networks. We relax this assumption, andargue that status characteristics—e.g., attractiveness—are sufficient to generate favorable network positions without agentic behavior on the part of the actor.Specifically, in a series of experiments, we find that people believe that both male and female actors who are described as more attractive are more likely tooccupy advantageous positions in networks, irrespective of other important characteristics (e.g., competence). Moreover, our results indicate that whenpeople select actors to fill advantageous network positions, they are more likely to select more attractive rather than less attractive actors. Together, ourresults support our argument that actors need not be agentic in their pursuit of favorable network positions to achieve these positions. For some, centralityfinds them.