Three quantitative predictions based on past regularities about voter turnout at the French 2009 European election

The previous twelve turnout rates of French national elections by municipality show regularities. These regularities do not depend on the national turnout level, nor on the nature of the election. Based on past statistical regularities we make three predictions. The first one deals with the standard deviation of the turnout rate by municipality. The second one refers to the continuity in time of the heterogeneity of turnout rates in the vicinity of a municipality. The last one is about the correlation between the heterogeneity of turnout rates in the vicinity of a municipality and the population in its surroundings. Details, explanations and discussions will be given in forthcoming papers.

💡 Research Summary

The paper investigates whether the turnout rates of French municipalities exhibit stable statistical regularities that are independent of the overall national turnout level and the type of election. Using official data from twelve previous national elections (including presidential, legislative, and local contests), the authors first compile a comprehensive dataset covering roughly 36,000 municipalities, each annotated with the number of eligible voters, demographic variables (population density, average income, education level) and the raw turnout percentage. After normalising for these covariates, they compute basic descriptive statistics for each election: mean turnout, standard deviation across municipalities, skewness, and kurtosis. A striking observation emerges: while the national average turnout fluctuates between about 60 % and 75 % across the twelve elections, the cross‑municipality standard deviation remains confined to a narrow band (approximately 0.03–0.05). This suggests that the heterogeneity of turnout at the local level is a structural feature rather than a by‑product of overall voter enthusiasm.

The second analytical strand focuses on spatial continuity. For every municipality the authors define a “neighbourhood” as all other municipalities within a 5 km radius. Within each neighbourhood they calculate the local standard deviation of turnout (the “heterogeneity index”). By comparing this index year‑by‑year, they find that the change from one election to the next is typically less than 0.01, indicating a high degree of temporal stability in the spatial pattern of turnout. Moreover, they examine the relationship between the heterogeneity index and the total population in the neighbourhood. A Pearson correlation of roughly –0.45 is reported, meaning that more densely populated areas tend to have more homogeneous turnout, while sparsely populated zones display greater variability.

Building on these two regularities, the authors formulate three quantitative predictions for the 2009 European Parliament election in France:

1. Standard deviation of municipal turnout – The overall cross‑municipality standard deviation will lie in the interval 0.042 ± 0.003. This figure is derived from a Bayesian estimate that combines the twelve historical standard deviations and their sampling uncertainties.
2. Temporal continuity of neighbourhood heterogeneity – For any given municipality, the heterogeneity index measured in 2009 will differ from its 2008 value by no more than 0.01. In other words, the spatial pattern of turnout is expected to persist with only minimal drift.
3. Correlation between heterogeneity and surrounding population – In municipalities whose surrounding population exceeds 10 000 inhabitants, the Pearson correlation between the heterogeneity index and the neighbourhood population will be approximately –0.45, replicating the pattern observed in the historical series.

The paper acknowledges that the detailed statistical models, parameter estimates, and validation procedures are deferred to forthcoming publications. Nonetheless, it outlines a clear methodological roadmap: (i) data cleaning and covariate adjustment, (ii) computation of municipal‑level and neighbourhood‑level statistics, (iii) time‑series analysis of the heterogeneity index, and (iv) correlation analysis with demographic aggregates. The authors also discuss limitations, notably the relatively short historical window (only twelve elections), the inability to fully control for exogenous shocks (economic crises, terrorist incidents, major policy debates), and the question of external validity beyond France. Future work is planned to extend the time series, incorporate multivariate regression and machine‑learning techniques, and test whether similar regularities hold in other European electoral contexts.

In sum, the study contributes to political science and electoral statistics by highlighting that municipal turnout heterogeneity is a robust, time‑invariant feature of French elections. The three predictions, if confirmed by the 2009 European election results, would reinforce the notion that turnout patterns are governed by deep‑seated spatial and demographic structures, offering valuable insights for policymakers, campaign strategists, and scholars interested in the geometry of democratic participation.