We propose a model on strategic formation of communication networks with (i) link specificity: the more direct links somebody maintains, the less she can specify her attention per link, the lower her links' value, while this negative externality was previously ignored in the communication context, and (ii) value transferability via indirect links for informational but not for social value from communication, while this positive externality was modeled uniformly before. Assuming only social value to isolate the impact of link specificity, the pairwise stable set includes many nonstandard networks under high or particular combinations of fully connected components under low link specificity. Allowing for social and informational value, the joint effect of link specificity and value transferability reduces the stable set to certain fragmented networks under high or the complete network under low link specificity. These extremes are beneficial for efficiency, whereas quite inefficient networks may arise for intermediate link specificity.JEL classification: A14, C79, D85, M31

We propose a model on strategic formation of communication networks with (i) link specificity: the more direct links somebody maintains, the less she can specify her attention per link, the lower her links' value, while this negative externality was previously ignored in the communication context, and (ii) value transferability via indirect links for informational but not for social value from communication, while this positive externality was modeled uniformly before. Assuming only social value to isolate the impact of link specificity, the pairwise stable set includes many nonstandard networks under high or particular combinations of fully connected components under low link specificity. Allowing for social and informational value, the joint effect of link specificity and value transferability reduces the stable set to certain fragmented networks under high or the complete network under low link specificity. These extremes are beneficial for efficiency, whereas quite inefficient networks may arise for intermediate link specificity.JEL classification: A14, C79, D85, M31

Network formation among individuals constitutes an important part of many OR processes, but relatively little is known about how individuals make their linking decisions in networks. This article provides an investigation of heuristic effects in individual linking decisions for network formation in an incentivized lab-experimental setting.Our mixed logit analysis demonstrates that the inherent complexity of the network linking setting causes individuals' choices to be systematically less guided by payof but more guided by simpler heuristic decision cues, and even stronger less motivated by the payoff for others. Furthermore, we show that the specific complexity factors value transferability and social tradeoff aggravate the former effect. These heuristic effects have important research and policy implications in areas that involve network formation.JEL Classification: A14, C25, C91, D85

No abstract

Network formation among individuals constitutes an important part of many OR processes, but relatively little is known about how individuals make their linking decisions in networks. This article provides an investigation of heuristic effects in individual linking decisions for network formation in an incentivized lab-experimental setting.Our mixed logit analysis demonstrates that the inherent complexity of the network linking setting causes individuals' choices to be systematically less guided by payoff but more guided by simpler heuristic decision cues, and that this shift is systematically stronger for social payoff than for own payoff. Furthermore, we show that the specific complexity factors value transferability and social tradeoff aggravate the former effect.These heuristic effects have important research and policy implications in areas that involve network formation.JEL Classification: A14, C25, C91, D85