Electoral systems exhibit macroscopic regularities that can be studied through macroecology by treating parties as species, votes as abundances, and polling stations as communities. We analyze Spanish elections from 1982 to 2023, focusing on the abundance fluctuation distribution (AFD), Taylor’s law, the mean abundance distribution (MAD), and pairwise party correlations. The empirical AFD varies with party mean abundance and resembles gamma or log-normal distributions; Taylor’s exponent fluctuates around 1.33; the MAD is well described by a logit-normal distribution; and party correlations are symmetric and heavy-tailed.
Because standard ecological models neglect the finite-size and compositional constraints of elections, we introduce constrained stochastic models. A constrained stochastic logistic model, with abstention acting as a compositional sink, reproduces the AFD of large parties but yields a Taylor exponent of 2. Adding multinomial sampling captures demographic noise, reproduces the AFD across parties, predicts a logit-normal MAD, and lowers the exponent to about 1.6. Finally, sparse stochastic interactions in a constrained stochastic Lotka–Volterra model recover the empirical correlation structure. These results lay the foundations for an ecological theory of voting.
Supervisors: Raúl Toral, Juan Fernández Gracia
Committee: Emilio Hernández García, Tobias Galla, Juan Fernández Gracia
This Master Thesis will be broadcasted in the following zoom link: https://us06web.zoom.us/j/89466064429?pwd=po9p99eAEYVPaNI8xIIGoOIz0hOqaF.1
Detalles de contacto:
Juan Fernández Gracia Contact form