Word of mouth and universal voting behaviour in proportional elections

Human societies exhibit many examples of nontrivial collective behaviour generated by the interaction of a large number of individuals. To characterize these emergent phenomena and understand their origin it is crucial to discover regularities in the macroscopic patterns of data describing collective social phenomena. Elections are a prominent example of collective social phenomenon, and a quantitative statistical analysis has begun. Here we show that, in proportional elections, the distribution of the number of votes received by candidates is universal, i.e. it is the same function in different countries and years, when the number of votes is rescaled according to the strength of the party each candidate belongs to. A simple dynamical model for the behaviour of voters, based on the spreading of word of mouth, reproduces well the universal distribution. This finding reveals the existence in the voting process of a general underlying dynamics that does not depend on the specific historical, political and/or economical context where voters operate. This suggests that the collective features of phenomena emerging from the opinions of large groups of interacting individuals (fads, rumours, consumer behaviour) can be understood quantitatively in terms of simple agent-based models