Open Primary Reform Simulator

An extension of the Atkinson, Ganz & Mantus (2024) 1D agent-based election model, using real per-state voter distributions from the 2020 CCES to compare choose-one plurality vs. approval open primaries, and top-2 vs. top-k general-election advancement.

Voter distribution & primary draw

Candidates are plotted directly on the voter-distribution curve: green = advanced to the general, gray = eliminated in the primary, ring = general-election winner, diamond = consensus candidate. The bar at x=0 shows the state's left/right voter split.

Election results

Headline comparison

Plurality top-2 (current CA/OH/WA default) vs. approval top-2 vs. plurality top-3, expected value of each metric over the Monte Carlo run, with change vs. the plurality top-2 baseline.

Metrics as a function of k

Each panel sweeps k ∈ {2,3,4,5} for all three primary rules. VSE points with no basis (denominator ≈ 0) are shown as a gap in the line, not a zero.

Metrics as a function of M

Each panel sweeps the primary candidate slate size M ∈ {5,10,15,20,25,30,40,50,61}, comparing plurality top-2 (baseline) vs. approval top-2 vs. plurality top-3 (mean-threshold approval, not fixed τ), with approval top-3 as an optional fourth line (see the checkbox in the controls panel).