This program simulates an tournament-style election to Congress. In this simulation, candidates face off in a multiple rounds of elections until one candidate from the field wins (similar to the NCAA basketball tournament. The field of candidates is a random draw of surnames from the U.S. population (based on Census Bureau data and weighted so more common names are more likely to be represented). I used this election tournament simulation to explore ideas I tested empirically in my 2015 article Alphabetically Ordered Ballots and the Composition of U.S. Legislatures.

You can define the parameters of the election. You can simulate an election between 2 and 8 rounds. The number of candidates will be 2^(rounds). Candidates' expected vote share is normally distributed with mean 50% and user-define standard deviation.

This election simulation explores how different ballot cues in entry-level, low-information elections may affect the outcomes of subsequent elections to higher offices. You can specify the percentage advantage to candidates with more common names, whiter names, and names that come earlier in the alphabet. You can also specify how many final rounds are immune from these biases (presumably because they are high-information elections).

Tournament Settings:Voter Bias Settings:
Number of Rounds:
St. Dev. Voting:
No Satisficing in Final _ Rounds
Advantage to More Common Name:
Advantage to Whiter Name:
Advantage to Earlier in Alphabet: