Opinion Dynamics in a Social Conflict

We develop a simple computational model capturing the co-evolution of opinion formation, political decision making and economic outcomes to study how societies form opinions if their members have opposing economic interests. The model features two types of individuals, a small minority and a large majority with conflicting economic interests who form and update beliefs about which policy best serves their interest through three channels: social influence, exposure to costly advertisements, and stochastic access to unbiased outside information. Individuals receive a payoff based on a policy they decide democratically and can use their funds to influence other individuals by sending costly advertisements. Our model illustrates how a tiny, but well-informed minority can influence democratic processes in their favor - even in situations where it seems unlikely at first glance, due to a vicious cycle in which political and economic power gradually shift from one group to another. Our model offers two ways out of this misery: First, on the individual-level, bounded confidence - the tendency of humans to distrust opinions which are too different from their own - has a mitigating effect and leads to better societal outcomes. This is particularly interesting, as bounded confidence has famously been shown by Hegselmann and Krause (2002) to produce polarization which is generally considered harmful. However, bounded confidence can cause political chaos and its effectiveness can be reduced by strategic messaging. Second, on a societal level, better access to unbiased information sources can counter disinformation. Our model highlights the dangers that economic and information inequality can pose for democracies and contributes to debates on the causes of the decades-long increase in inequality in democratic countries and the persistent failure to adequately address climate change.