The voting method determines how voters will communicate their preferences on ballots, and how ballots are counted.
Which method should we use?
One of the first things to consider when preparing for a decision, which includes choosing a voting method, is how many options can win. Probabilistically speaking, voters are likely to be more satisfied with the results of a decision that selects multiple winners because there's a greater chance at least one of multiple winning options comes closer to their ideal than when a single winner is selected.
Say, for example, you want to decide on a few lunch spots for the upcoming week among colleagues at work. Instead of repeating a single-winner approval decision multiple times, which could result in the same winner over and over, you could run a multiple-winner approval decision, yielding a better variety and better optimizing satisfaction.
Here’s a list of each method Accord supports matched with an example scenario it’d be ideally suited to:
- Approval: Choosing a restaurant to go to among a group of people
- Single transferable vote: Electing officers
- Consensus: Approving a complex but important budget
- Score or STAR: Choosing whom to hire among a range of candidates
Plurality is intentionally not listed, as we’ll clarify in its section below.
In approval decisions, voters receive a ballot where they can put a checkmark next to each option they approve of. As a result, the option(s) with the greatest degree of approval will be selected.
Contrast this method with:
- Score, which could be argued to yield winning option(s) with the greatest degree of enthusiasm, not just approval, however it is a more complex ballot
- Plurality, which instead of letting voters indicate all options that would be satisfactory, asks voters to choose just one, encouraging strategic voting and risking a suboptimal result
In consensus decisions, voters receive a ballot where for each option they can indicate support, indifference, or principled objection. The option(s) which both did not receive a disqualifying number of principled objections and received the greatest support will win.
Using consensus ensures that no option that is sufficiently objectionable will win. Because of this, it's possible, even likely, that no option will win at first; this is by design. If you choose to use a consensus method, you should also expect to need to negotiate and then repeat it anew until the group achieves a positive result.
In score decisions, voters receive a ballot where for each option they can indicate a degree of support from indifference up to a configurable maximum number. Accord selects winning option(s) based on the sum of scores given on each vote’s ballot.
Among theoreticians, score voting is widely considered to optimize voter satisfaction best among all voting methods.
In STAR (“score then automatic runoff”) decisions, voters receive the same type of ballot as for score decisions. STAR has the same advantages as score voting but reduces the incentive to vote strategically by selecting winners in successive runoff contests among top scored options.
The STAR method is developed and promoted by the Equal Vote Coalition for civic elections.
Single transferable vote
Single transferable vote (STV) is a type of ordinal voting, so voters receive a ballot where they are asked to rank their choices. The ballot is similar to an approval ballot except instead of choosing a group of options they prefer to any degree, the voter arranges the options in the order of their preference.
STV is used in political elections in Ireland and Australia as a way to give voters an opportunity to support the candidates they prefer regardless of how popular they may be, since votes are redistributed to voters’ next preference when a candidate either wins or comes last in an iteration.
Because of the way STV works, it’s not recommended for decisions which should have only one winner; STV was designed for multiple-winner elections specifically.
In plurality decisions, voters receive a ballot where they choose the one option they think should win. The option(s) with the most votes win.
While this is the simplest voting method available, it is arguably the least likely to optimize satisfaction since it often causes these unwanted effects:
- Wasted votes, which are votes “cast for options which are virtually sure to lose, and votes cast for winning options in excess of the number required for victory.”
- Tactical voting, where “voters are under pressure to vote for one of the two options most likely to win even if their true preference is neither of them because a vote for any other option is unlikely to lead to the preferred option winning, but it will instead reduce support for one of the two most popular options whom the voter might prefer to the other.”
- The spoiler effect, where two similar options may split a large share of votes and cause an opposing option to win that would not have won had only one of the similar opposing options been available.
If your group is hesitant to use a method other than plurality, consider running plurality decisions in parallel with the same decision using a different method.