Rational Reconstructions of Society
Arnout, van de Rijt
The dissertation contains four stand-alone studies, chapters 2 through 5. In chapter 1, I highlight commonalities among the studies. In chapter 2, I consider the principle of structural balance - "The friend of a friend is a friend, the enemy of a friend is an enemy, the friend of an enemy is an enemy, and the enemy of an enemy is a friend." I consider Harary's (1954) result that this principle can only be satisfied in a world consisting of two inimical friendship cliques. And I consider recent studies that show that when individuals in a structurally imbalanced world change ties one by one following the principle, they do not necessarily end up in a structurally balanced world. I prove that if multiple ties can be changed simultaneously, then a structurally balanced world is guaranteed. In chapter 3, I consider Burt's (1992) argument of "structural holes" that unconnected parts of a social network are niches for brokerage. I consider Burt's suggestion that those aware of brokerage benefits end up occupying structurally advantaged network positions. I show how this statement crucially depends on the unawareness of these benefits by others. If everyone strives for structural holes, no one ends up with a structural advantage. In chapter 4, I consider the extensive laboratory evidence on the relationship between the structure of small exchange networks and expected exchange rates. I consider a theory that reasonably predicts this relationship in a handful of networks. I show that if individuals add ties that increase expected earnings from exchange more than they cost and delete all other ties, then networks emerge that distribute exchange benefits equally. In chapter 5, I consider the old immigrant assimilation model of a monotonic process. I consider recent work in the direction of an alternative model. I propose an alternative model that follows up on this work and adds minimal complexity to the old model. In this model, quite assimilated migrants further assimilate, while not so assimilated migrants reverse-assimilate. Using longitudinal survey data, I show that the model is empirically competitive. In chapter 6, I propose four follow-up studies.
social networks; exchange; rationality; structural holes; structural balance; immigration
dissertation or thesis