Predicting personality and sociopolitical outcomes from shopping data

Personality traits are predictive of real world outcomes, however they are often measured through self-reports. Can we extract personality traits from mass observational data reflecting real world choices? Using shopping data from 20,550,952 customers of a large UK retailer, we calculated behavioural entropy or preference for variability in shopping choices. Areas with higher behavioural entropy scored higher on the personality trait openness, supporting the shared characteristics of embracing novel experiences between the two variables. Regional entropy was positively associated with the proportion of Remain votes and voter turnout. This research contributes to psychological theory by demonstrating that stable personality traits can be inferred from everyday consumer behaviours, and that these inferred traits systematically relate to sociopolitical patterns across regions. This finding bridges individual-level psychological constructs and collective outcomes, suggesting that personality operates not only at an individual level but also as a distributed, observable force in population-level behaviour.