Ranked-Choice vs Approval Voting
Australia changed its federal lower-house ballot in 1918 after the rise of the Country Party created a non-Labor vote-splitting problem, according to the Australian Electoral Commission. Fargo, North Dakota amended its city charter in 2018 so voters could approve any number of candidates in city elections. Both systems attack the spoiler problem; they disagree on the meaning of a vote.
Ranked-choice voting says a voter carries an ordered list: first, second, third. Approval voting says the useful signal is a threshold: acceptable or not acceptable. That difference looks small on a ballot. It changes the politics of campaigning, counting, and strategic voting.
The Split
In single-winner ranked-choice elections, often called instant-runoff voting in U.S. debates, the lowest first-choice candidate is eliminated round by round. Ballots transfer to the next ranked candidate still alive. The winner needs a majority of active ballots at the final count.
Approval voting keeps one count. Each voter may approve any number of candidates. The candidate with the highest approval total wins.
| Question | Ranked-choice voting | Approval voting |
|---|---|---|
| Ballot signal | Ordered preference | Acceptability threshold |
| Count shape | Multiple rounds | One round |
| Australia federal House use | Since 1918 | No |
| Fargo city use | No | Adopted 2018 |
| Main promise | Less fear of wasting a first choice | Less fear of supporting more than one candidate |
| Main weakness | Can fail monotonicity | Cannot show preference strength between approved names |
The hidden philosophical question is this: does democracy need to know your ranking, or only your consent boundary?
How The Count Moves
Ranked-choice voting lets a voter rank a minor candidate first and a major candidate second without casting a dead ballot. That is the Australia logic after 1918: keep related parties from splitting a bloc.
Approval voting protects breadth. A voter can approve a favorite and a compromise candidate at the same time. Brams and Fishburn’s 1978 paper argues that this can identify candidates with wide acceptability in a crowded field.
Neither system removes strategy. Ranked-choice voters may still care about elimination order. Approval voters must decide where to place the approval cutoff: only the favorite, every acceptable candidate, or the viable compromise too.
What Is Contested
The ranked-choice side argues that equal approval marks flatten real preferences. If I approve A and B, the count cannot see that A is my first choice and B is only tolerable. In close elections, that feels like losing information.
The approval side argues that ranked-choice can hide a broadly acceptable candidate if that candidate is eliminated before later preferences activate. It also points to the simplicity of a one-round count: Fargo’s charter tells voters to vote for all names they approve of.
The contest comes from social choice theory. Arrow’s 1951 impossibility theorem did not say all systems are equally bad. It said no ranked system can satisfy every fairness condition at once under broad assumptions. The practical task is not finding a perfect ballot; it is choosing which failure mode a city can live with.
Why This Touches Other Realms
Voting systems look like civics, but they are also measurement devices. Ranked-choice measures order under pressure. Approval voting measures a boundary. The same problem appears in concept fermi paradox: one model asks where the civilizations are; another asks whether our detection threshold is wrong.
That is why election design belongs beside engineering pages like mission voyager 1. A counting rule is a machine. Its output depends on what the instrument is built to detect.
An Open Question
If a city wants less spoiler fear and less strategic ballot marking, should it ask voters for more information or ask them one simpler question? The answer may depend less on mathematics than on how many candidates voters can actually evaluate before the ballot reaches their hand.
Abhishek's take
I like approval voting because it treats politics like a threshold problem, not a personality ranking exercise. I like ranked-choice because it preserves the voter’s first story before forcing compromise. The trade is clean: ranked-choice records order, approval records consent, and both lose something the voter may have meant.
Tags: #voting-systems #social-choice #democracy #election-design #collective-decision