CHAINS OF COOPERATION - THRESHOLD EFFECTS IN COLLECTIVE ACTION

Granovetter's threshold model of collective action shows how each new participant triggers others until the chain reaction reaches a gap in the distribution of thresholds. Hence outcomes depend on the network of social ties that channel the chain reactions. However, structural analysis is encumbered by the assumption that thresholds derive from changing marginal returns on investments in public goods. A learning-theoretic specification imposes less stringent assumptions about the rationality of the actors and is much better suited to a structural analysis. Computer simulations suggest that threshold effects may be the key to solving the coordination problem: When individual choices are contingent on participation by others, this interdependence facilitates the coordination of contributions needed to shift the bistable system from a noncooperative equilibrium to a cooperative one. Further simulations with low-density networks show that these chain reactions require bridges that link socially distant actors, supporting Granovetter's case for the strength of weak ties.