Networked politics: political cycles and instability under social influences

Media, opinion leaders, co-ethnics, family members, and friends influence our political decisions. The ways in which these influences affect political cycles and (in)stability has been understudied. We propose a model of a networked political economy, where agents' choices are partly determined by the opinions of the individuals with whom they are connected in a fixed influence network. The model features two types of individuals: ideological individuals who never change their views and who seek to influence the rest of the society; and non-ideological individuals who have no political allegiance and do not influence anybody, but who can be influenced by ideological individuals with whom they are connected. We show that influence networks increase political turnout and cause non-ideological individuals who are subject to antagonistic influences to keep changing their political views. This in turn increases political cycles and instability in two ways: (1) by reducing the number of stable and popular political leaders; and (2) by worsening the tradeoff between political competition and the existence of a stable leader. We uncover a necessary and sufficient condition that characterizes all of the political technologies and network structures that guarantee political stability. This condition introduces a preference-blind stability index, which maps each pair of a constitution and an influence network into the maximum number of competing political leaders that a society can afford while remaining stable regardless of the extent of preference heterogeneity in its population. Our findings have testable implications for different societies. They shed light on the network origins of political cycles in two-party systems. They also imply that individualist societies are more politically stable than collectivist societies and societies organized around ethnic groups or characterized by a high level of homophilous behavior and influences. For ethnic democracies, we quantify the exact tradeoff between political competition and stability, and show that ethnic fragmentation increases stability. The findings further provide a rationale for using the "divide and rule" strategy to maintain political power. Finally, we find that cliques and multi-layer cliques maximize the competition-stability tradeoff, whereas star networks, lines and rings minimize it.