Generalizing Instant Runoff Voting to Allow Indifferences

Théo Delemazure,Dominik Peters

2024-04-17

Abstract:Instant Runoff Voting (IRV) is used in elections for many political offices around the world. It allows voters to specify their preferences among candidates as a ranking. We identify a generalization of the rule, called Approval-IRV, that allows voters more freedom by allowing them to give equal preference to several candidates. Such weak orders are a more expressive input format than linear orders, and they help reduce the cognitive effort of voting.

Computer Science and Game Theory,Theoretical Economics

What problem does this paper attempt to address?

The problem that this paper attempts to solve is: how to extend Instant Runoff Voting (IRV) to situations where voters are allowed to express indifferent preferences among candidates. Specifically, the author proposes a new rule called "Approval - IRV" to solve the following problems:

1. **Reduce voting burden**: Traditional IRV requires voters to rank all candidates strictly, which will increase voters' cognitive burden when the number of candidates is large. Allowing voters to express the same preference for multiple candidates can reduce this burden.

2. **Improve expressive ability**: Allowing voters to express the same preference for multiple candidates makes the ballot more flexible and truly reflects the will of the voters.

3. **Reduce invalid ballots**: In some elections, voters may submit invalid ballots because they do not understand the complex ranking rules. Allowing weak ranking (i.e., allowing the expression of indifference) can reduce the occurrence of this situation.

4. **Align better with election campaigns**: Many voters decide how to vote based on election campaigns. In some cases, a campaign team may only suggest that supporters rank their candidate first and not provide ranking guidance for other candidates. Allowing voters to rank multiple candidates first equally can better reflect these situations.

5. **Better incentive mechanism**: In traditional IRV, voters may vote strategically to avoid wasting their votes. By allowing voters to rank multiple candidates first equally, this problem can be alleviated to a certain extent.

6. **As a compromise between ranked - choice voting and approval voting**: It combines the advantages of these two voting methods and may become a compromise solution in some cases.

### Specific content

The main contribution of the paper is to propose a new voting rule - Approval - IRV and prove that it has good properties when dealing with weak rankings. Specifically:

- **Independence of Clones**: Even if some candidates are replicated (i.e., multiple similar candidates are introduced), it will not affect the election result.

- **Respecting Cohesive Majorities**: If most voters rank a certain candidate at the top (even if tied with other candidates), then the final winning candidate must be one of the top - ranked candidates among these voters.

- **Indifference Monotonicity**: If a certain candidate is a winner, when a voter promotes this candidate from a lower position to a position tied with other candidates, this candidate should still win.

In addition, the paper also explores the problem of proportional representation in multi - round elections and proves that Approval - STV (the multi - round election version of Approval - IRV) retains the proportional representation property of linear - ranked STV.

In conclusion, this paper aims to improve the existing IRV rules by introducing Approval - IRV, making it more suitable for the needs of modern elections while maintaining its original good characteristics.