dc.contributor.author | Stensholt, Eivind | |
dc.date.accessioned | 2020-06-23T11:14:47Z | |
dc.date.available | 2020-06-23T11:14:47Z | |
dc.date.issued | 2020-06-23 | |
dc.identifier.issn | 1500-4066 | |
dc.identifier.uri | https://hdl.handle.net/11250/2659146 | |
dc.description.abstract | Struggles over the single-seat preferential election method IRV (Instant Runoff Voting) go on in public arenas and scientific journals, with focus on two “anomalies”. “Monotonicity failures” are preference distributions that allow a startling strategic voting called Pushover or its reverse. Analysis shows how a Pushover action works, and why it is hard to predict and exploit an opportunity. While not rare, monotonicity failures should be seen as a minor nuisance.
“No-Show paradoxes” are alarms. The IRV tally eliminates a very clear Condorcet winner in a realistic, but somewhat unusual preference structure, too close to Duncan Black’s Single- Peak condition: Too many YXZ-ballots let Z win instead of a very clear Condorcet-winner X who is eliminated; this harms IRV’s legitimacy. Baldwin’s elimination rule when three candidates remain is a suggested remedy.
Preference distributions with the same IRV-tally are grouped together and analyzed with “pictograms” as a tool. That allows a generalization of Black’s Single-Peak condition; real cases are close to “Perfect Pie-sharing”, which explains why Condorcet cycles are rare. | en_US |
dc.language.iso | eng | en_US |
dc.publisher | FOR | en_US |
dc.relation.ispartofseries | Discussion paper;6/20 | |
dc.subject | Instant Runoff Voting | en_US |
dc.subject | Condorcet methods | en_US |
dc.subject | Duncan Black’s Single-Peak condition | en_US |
dc.subject | Baldwin’s elimination rule | en_US |
dc.title | Anomalies of Instant Runoff Voting | en_US |
dc.type | Working paper | en_US |
dc.source.pagenumber | 40 | en_US |