The robust beauty of APA presidential elections: an empty-handed hunt for the social choice conundrum

Social choice theory in Economics and Political Science has highlighted that competing notions of rational social choice are irreconcilable. This established wisdom is based on hypothetical thought experiments, mathematical impossibility theorems, and computer simulations. We provide new empirical evidence that challenges the practicality of these discouraging predictions. We analyze the ballots from thirteen presidential elections of the American Psychological Association. We report on an empirical comparison of the Condorcet, the Borda, the Plurality, the Anti-Plurality, the Single Transferable Vote, the Coombs, and the Plurality Runoff rules. We find that these rules frequently agree both on the winner and on the social order. Bootstrapping reveals that the coherence among competing rules is a property of the empirical distribution of voters’ choices, and it is not specific to a particular sample. Our findings are highly robust to changes in the modeling assumptions that enter our analysis. These findings suggest many interesting open research questions for the emerging paradigm of behavioral social choice: Why do competing social choice procedures agree in real-world electorates? How broadly does the accumulated evidence against the social choice conundrum generalize to other electorates and other candidate choice sets