Consent or Conflict: Coevolution of Coordination and Networks

Abstract

Societies are sometimes divided into groups that behave in different ways or have strongly opposing opinions. At other times, everyone seems to behave according to similar principles and opinions. It is likely that individual decisions on behavior or opinions are affected by social networks through influence and selection processes. However, the outcomes are not necessarily optimal for the society as a whole. Two types of problems might arise: (a) polarization of the society into two camps that do not reach consensus, possibly leading to conflict; (b) actors choosing suboptimal behavior, because changing behavior is too risky if done unilaterally. Simulations show that if a society is rather segregated initially, there exists a heightened probability that this situation will worsen. The effect of network density is twofold. First, density has a positive effect on reaching a uniform opinion and, therefore, decreases the likelihood of polarization. Second, density increases the likelihood that actors do not change their behavior, worsening the inefficiency of already suboptimal initial situations. * We thank Werner Raub, Tom Snijders, Arnout van de Rijt, Gerald Schneider, Thomas Plümper, anonymous journal reviewers, and many seminar participants for their constructive comments.

Network Structure and Strategic Investments: An Experimental Analysis

“…Goeree et al (2008) analyze a pure network formation game and find that equilibrium predictions fail completely with homogeneous agents, while heterogeneity fosters the formation of an efficient structure over time. Buskens et al (2007) find some evidence that subjects will form networks that lead to play of the efficient Nash equilibrium in a simple coordination game, if they can choose their network partners themselves.…”

What Drives Economic Policy Uncertainty in the Long and Short Runs? European and U.S. Evidence over Several Decades

“…First, higher levels of political polarization lead to greater economic instability and uncertainty (Alesina, et al 1989). Second, as people with similar beliefs find ways to self-segregate, the amount of conflict and its severity tends to rise unless offset by social networks that cut across group lines (Buskens, et al, 2008). In other words, recent increases in political polarization, especially after the onset of the recent financial-market crisis, would naturally lead to elevated levels of economic uncertainty as groups come to see other groups as posing a threat, whether it be natives versus immigrants, low-skilled versus high-skilled, or the poor versus the rich.…”

Alcohol use among adolescents as a coordination problem in a dynamic network

“…We then briefly summarize the simulation analyses by Buskens et al (2008), which provide the basis for our predictions. Subsequently we describe the data and methods used to test these predictions, and present the results of our analyses.…”

“…The analysis was performed separately for two different payoff structures: one where the payoff-dominant action Y is also risk-dominant, and one in which Y is Table 1. Selected logistic regression coefficients on the proportion of actors choosing payoff-dominant action 'Y' interactions computed with variables centred at their means (simulation results from Buskens et al, 2008 risk-dominated. The latter case corresponds to our model of drinking as presented above.…”

“…Third, there are interaction effects between initial behaviour and initial density and centralization, in the sense that these network properties can strengthen or weaken the effect of initial behaviour. For further detail on the setup of the simulations and the derivation of the above results, we refer the reader to Buskens et al (2008).…”