23.5 Score voting

Score voting is a reasonable approach when there is a need to democratically pick one from several options (for example, the operator that wins a tender).

Voters give each option a numerical score (e.g. a number between 0 and 10), and the option with the highest average score is picked.

Pros of Score Voting

• Expressiveness and nuance: Voters can express the strength of their preference for each candidate using a scale, rather than a simple choice or rank order. This provides more information than other methods.
• Reduced spoiler effect: Score voting satisfies the "independence of irrelevant alternatives" criterion, meaning less popular candidates are less likely to act as "spoilers" by drawing votes away from a similar more popular candidate.
• Encourages broader appeal: Promoters of a candidate are incentivized to appeal to a wider range of voters to gain higher scores, rather than just focusing on their base, potentially leading to more civil campaigns and consensus candidates.
• Simplicity and speed: Marking a score for each candidate can be faster for voters than complex ranking systems, and the tabulation is straightforward: sum the scores.
• Higher voter satisfaction efficiency: Simulations often show that score voting leads to high average voter satisfaction because it generally elects the candidates with the strongest overall support across the people.
• No "less-is-more" paradox: A voter can never hurt their preferred candidate's chances by giving it a higher score or helping it by giving it a lower score.

Cons of Score Voting

• Vulnerability to strategic voting: This is the primary criticism. Voters are incentivized to give their favorite candidate the maximum possible score and all other candidates the minimum possible score, regardless of their true nuanced opinion, to maximize their impact on the outcome. This is known as "bullet voting" or "approval-style" strategy.
• Potential for undemocratic results via strategy: If one segment of the population votes honestly while another segment votes strategically, the strategic voters gain a disproportionate advantage, which could lead to results that do not reflect the true overall preferences of the people.
• Doesn't guarantee a Condorcet winner: Score voting does not always elect the "Condorcet winner" (the candidate who would beat each losing candidate in a one-on-one race), although this outcome is rare in practice when voters behave strategically.
• Potential for complication: While simple to use, the ability to assign scores within a range might be slightly more complicated than a simple approval vote.
• Favors moderates (debated): Some critics argue that the system tends to favor middle-of-the-road candidates and discourages more novel or extreme solutions, pushing towards the status quo.

Good Properties:

• no less-is-more paradox (raising an option's rating can never hurt its chances of winning)
• participation criterion (an option can never lose because more voters support it)
• independence of irrelevant alternatives (it is not affected by spoiler effects)
• not subject to Arrow's theorem

Bad properties:

• It does not satisfy the Condorcet criterion, that is, the method does not always agree with the majority rule.

Uses of n-way voting

• choose operator that wins a tender
• choose zoning of land

Uses of median value voting

• values that appear in tenders. E.g.
  • Number of rooms required when building a hospital
  • prices for fixed price contracts