Wishocracy: Solving the Democratic Principal-Agent Problem Through Pairwise Preference Aggregation

Abstract: Politicians' votes have near-zero correlation with citizen preferences (Gilens and Page, 2014). Elite preferences predict policy outcomes. No mechanism connects citizen preferences to electoral consequences for representatives. RAPPA: Millions of citizens answer simple pairwise questions ("How would you split \$100 between these two budget categories?"). Geometric mean aggregation produces population-level preference weights from sparse individual responses. Unlike approval voting or ranked choice, RAPPA captures preference *intensity*, not just what people want, but how much they care. Compare aggregated preferences to each legislator's voting record. Publish Citizen Alignment Scores. Channel campaign resources to high-alignment candidates through Incentive Alignment Bonds. The mechanism achieves three properties no prior system combines: minimal cognitive load (~20 comparisons per participant yields statistical convergence), preference intensity capture, and approximate strategy-proofness. Summary: Representative democracy suffers from an inescapable principal-agent problem where elected officials' incentives diverge from citizen welfare. Wishocracy introduces RAPPA (Randomized Aggregated Pairwise Preference Allocation), which aggregates citizen preferences through cognitively tractable pairwise comparisons and creates accountability via Citizen Alignment Scores that channel electoral resources toward politicians who actually represent what citizens want.