A Note on Detecting Dividing Lines in Turnout: Spatial Dependence and Heterogeneity in the 2012 US Presidential Election

US voters have been moving apart in the last twenty years. This paper analyzes how their voting participation has partitioned by looking at US counties in the 2012 Presidential elections. To tackle this question, we propose a methodology that jointly addresses spatial autocorrelation of the dependent variable and splits the sample according to the non-linearity in the covariates. We find that in different groups of US counties, obtained through a spatial lag regression tree procedure, some variables have different statistical significance (or lack of it), and sometimes different signs. This heterogeneity – which is a manifestation of the complexity of the political behavior - is obfuscated by traditional methods that extrapolate a single average relationship between the variables.