Single-Winner Delegative Voting

Bryan Ford - October 21, 2002

Introduction

Single-Winner Elections

Multiple-Seat Elections

(Copyright 2002 Bryan Ford)

How it works

Single-winner elections are used to elect candidates to offices that can inherently be occupied only by one person at a time, such as President, Governor, or Treasurer. Delegative voting in a single-winner election works as follows. Well in advance of the election, each candidate submits to the electoral authority a list of next-choice candidates, in which order is not significant. This list indicates which other candidates the candidate in question considers to be (most) qualified to take the office in his place, if he himself is not elected to that office. In essence, before the election each candidate publicly casts a "approval vote" among all the other candidates. The electoral authority publicizes the next-choice lists of all the candidates along with other public information it makes available to voters for consideration.

When the popular vote is taken, voting and ballot counting is done exactly as in plurality elections: each voter simply chooses one from the set of candidates, and the counting process merely involves totaling the number of votes cast for each candidate. The only difference between plurality and delegative voting is in how the winner or winners are determined after all the ballots have been counted and the raw totals established.

To determine the winner of a single-seat election, the list of results is sorted according to the number of votes each candidate received, and the candidate with the smallest number of votes is successively eliminated until only one remains. When a candidate is eliminated, his votes are not simply "lost", but are instead transferred to other candidates still in the race. The votes received by the eliminated candidate are divided evenly among all of the candidates on his next-choice list that have not already been eliminated. In effect, the votes of an eliminated candidate remain "in play" and can affect the outcome of the rest of the election, providing an identical "boost" to each of the next-choice candidates.

Advantages

In contrast with plurality systems, which often elect a candidate that is supported by far fewer than a majority of voters, delegative voting can ensure that the winning candidate always receives a majority of the valid votes, as long the candidates cast their next-choice "approval" votes intelligently (and it is generally in their own best interest to do so). The vote transfer mechanism eliminates the "spoiler" effect that plagues plurality systems, in which two closely-aligned candidates or parties lose to a third candidate that is the least popular overall because the votes for the first two are divided.

As an example, suppose there are three candidates in a close election: Left, Center, and Right. Left is more closely aligned with Center than with Right, and Right is similarly more closely aligned with Center than with Left; therefore Left and Right each list ("approve of") Center as their sole next-choice candidate. Center, not wanting to take sides between the other two candidates, might either include both Left and Right on its next-choice list, or leave the list empty. Suppose the popular vote is then taken: Left receives 32% of the votes, Center receives 33%, and Right receives 35%. In a conventional plurality system, the Right candidate would win this election despite having only a minority of support (35%) and despite probably being the least favored by up to 32% of the electorate. In a delegative election, in contrast, the Left candidate would be eliminated first and his 32% share of the votes transferred to Center, who subsequently wins with a solid 65% majority.

Delegative voting removes the effective "lock-out" on smaller parties inherent in plurality systems, enabling less popular candidates to influence the election even if they cannot win it and thus giving them a stronger incentive to run. This friendliness to secondary candidates increases the effective choice voters have because it allows them to vote their conscience - i.e., for the candidate they really prefer rather than for the least undesirable candidate they think can win - without wasting their vote or risking spoiling the election for the mainstream candidate who is most closely aligned to their views. Delegative voting also encourages greater cooperation between candidates and discourages negative campaigning, because mainstream candidates often must depend on receiving some transferred votes from other candidates in order to win a majority.

Relationship to Instant Runoff Voting

Those familiar with electoral systems will immediately notice that this system is similar to Instant Runoff Voting (IRV) in the way the winner is determined. Both alternative systems involve the successive elimination of candidates and the automatic transfer of the eliminated candidate's votes. In IRV, however, voters explicitly rank the candidates in order on their ballots, and the ballots of eliminated candidates are transferred according to the voters' rankings. Delegative voting in contrast gives this power to the candidates themselves, effectively making their political alignments a fixed part of their public "platform".

Although DV may seem to give voters less "choice" than IRV if the number of candidates is assumed to be fixed, DV nevertheless has several important advantages over IRV:

  • Simplicity: DV makes the ordinary voter's task simpler, easier to understand, and more familiar to voters accustomed to majority/plurality elections. Experience with IRV and other preferential systems demonstrates that a large percentage of voters in a preferential election either do not mark any second or higher choices at all, or simply follow their party's or first-choice candidate's recommendations anyway when marking their second and successive choices. Therefore most voters are unlikely to miss the preferential-ranking flexibility of IRV.
  • Better next-choice decisions: Candidates are often inherently in a better position than voters to make next-choice decisions anyway, because candidates are activists who have to know about and pay close attention to each other in order to campaign and compete effectively, whereas voters usually just pay attention to the one or two candidates they know best and agree with most, showing little knowledge or preference among their less-desirable choices. DV allows voters to focus their attention on their most important decision - picking the candidate who best represents themselves and their interests - without being distracted with ambiguous choices between candidates they don't like or simply don't know about.
  • Broad-spectrum candidacy: DV makes up for the loss of preferential ranking flexibility by making it feasible for a much wider range of candidates to run in an election than is practical even with IRV. While IRV can increase the typical effective voter choice from two or at most three competitive candidates in a plurality election to five or ten candidates in an instant-runoff election, the number of candidates in IRV cannot be allowed to grow much beyond this level without making preferential ballots impractically large and risking "voter overload". With an appropriate single-choice ballot design, however, such as the French system in which voters simply insert party- or candidate-supplied cards into a "ballot envelope", DV makes it practical for hundreds or even thousands of candidates to run in a single election.
  • Community representation: With a larger number of candidates running in a given election and using their next-choice lists to forge alliances, the candidates in a DV election can focus their popular campaigns on the particular communities with which they are most closely aligned, whether those communities are defined by geography, culture, religion, political ideology, or specific issues. Voters in turn can choose among candidates from their own communities who they know and trust directly, and can ignore the more distant candidates entirely. Candidates build strength and ultimately win office not through mass propaganda, but by building a solid local support base coupled with a strong network of political alliances with other candidates in similar or related communities.
  • Transparency: The inter-party relationships and coalitions that IRV encourages are made fully explicit, publicly transparent, and directly measurable in DV. Since smaller candidates directly control the transfer of their votes, mainstream candidates have a strong incentive to court the smaller candidates for inclusion on their next-choice lists and take account of the minority viewpoints they represent.
  • Cost: On the practical side, shifting responsibility for the next-choice lists from the voters to the candidates eliminates the need for expensive voting equipment upgrades or re-training of electoral administration staff to deal with the intricacies of preferential ballot counting.
  • Speed: The vote counting process itself can proceed much more quickly in DV, because ballots are only counted once rather than multiple times. The runoff elimination process to determine the actual winner only requires the raw totals from the vote count and the previously published next-choice lists of the candidates.
  • Security: The simpler vote counting process is less prone to tampering, making DV particularly attractive in unstable or corrupt democracies where the multiple preferential ballot re-shuffling steps involved in IRV elections may be seen as too difficult to implement and monitor effectively. The runoff elimination process in DV is fully independently verifiable because it only requires information that is already public anyway; therefore DV is no less secure than plurality voting.