Five hard critiques, answered.
Every voting system has trade-offs. Score voting does too. Before the book went out, it was tested against the five strongest critiques of score voting. Here they are, with the answers and the trade-offs the book accepts.
Score voting is the author’s argued choice. It is not settled research, and you won’t find it sold that way here. Where the evidence is thin, this page says so.
The critique
“It doesn’t follow majority rule.”
That’s true. A candidate can be the first choice of most voters and still lose. Runoffs and ranked choice are built to find a majority winner. Score voting is built to find the candidate with the most support overall.
Score voting weighs how much voters care, and how much they object. So a broadly acceptable candidate can beat one that nearly half the voters strongly oppose. Here is a race with 100 voters.
| Voters | Avery | Blake | Cruz |
|---|---|---|---|
| 51 voters | 10 | 7 | 0 |
| 49 voters | 0 | 8 | 10 |
| Total | 510 | 749 | 490 |
Avery is the first choice of 51 voters, and the other 49 gave Avery a 0. Blake is nobody’s favorite, but every voter scored Blake a 7 or higher. Blake wins with 749, ahead of Avery’s 510 and Cruz’s 490. The book argues that is the better result: a winner nearly everyone can live with.
Majority rule isn’t gone, though. If those 51 voters care enough to give Blake a 0, Avery wins, 510 to 392. A majority that feels strongly can still win by saying so on the ballot.
The trade-off the book accepts
A mild majority can lose to a candidate almost everyone likes. The book counts that as a strength. You may not, and that is a fair place to disagree.
The critique
“Equal opinions should get equal power.”
Picture two voters who agree. Both prefer Avery to Blake. One scores Avery 10 and Blake 0. The other scores them 7 and 5. On that pair, the first voter moved the result five times as much. Critics say careful voters get drowned out by voters who push to the ends of the scale.
That gap is real. The answer is about the floor. Every voter gets the same ballot and the same range. The strongest ballot anyone can cast, a 10 for one candidate and 0 for the rest, is the pick-one ballot that plurality forces on everyone today. Nobody is denied it.
So the worst an honest voter can do is what plurality already gives every voter. Nobody gets less voice than today, and most get more, because they can also say something about every other name on the ballot. A voter who scores 7 and 5 is telling the count something true: those two are close.
The trade-off the book accepts
A voter who uses the full range on two candidates moves that pair more than a voter who doesn’t. The book accepts that. Everyone gets the same range, and the floor is today’s ballot.
The critique
“Ranked choice has a track record. Score voting doesn’t.”
Also true. Ranked choice runs elections in dozens of U.S. cities and counties and in two states. Full 0 to 10 score voting has never run a U.S. government election. Its simplest form, approval voting, has. Fargo, North Dakota, adopted it with 63.5 percent of the vote, and St. Louis with 68 percent.
Source: Fargo Measure 1, Nov 2018; St. Louis Proposition D, Nov 2020
The track record also shows where ranked choice goes wrong. It keeps eliminations, and the eliminations cause its problems. In Burlington, Vermont, in 2009 and Alaska in 2022, the candidate who beat every rival head-to-head was cut before the last round. Ballots that rank nobody left in the race get set aside. And a candidate can lose by gaining first-choice votes. All of that happens in the rounds.
These failures are rare, and the book says so. One study of 182 competitive U.S. ranked-choice races from 2004 to 2022 found the head-to-head winner eliminated in two of them.
Source: Graham-Squire and McCune, “An Examination of Ranked-Choice Voting in the United States, 2004–2022,” Representation 61, no. 1 (2025)
Score voting has no rounds. It is simple addition. Each precinct can add up its own totals and post them, and anyone can add the precincts together. Ranked choice can’t be counted that way. Each round depends on every ballot in the race, so all the ballot data has to come to one place before the rounds can run.
The trade-off the book accepts
Score voting asks you to trust a method with less real-world use. The book bets that a simpler count will earn trust faster. That is an argument, and the elections that would prove it haven’t happened yet.
The critique
“Ranking feels more natural.”
People rank things all the time: a first choice, a second choice. Ranking asks only for order, and some voters find that easier than putting a number on a person.
Ranking shows order only. Scoring shows strength. A ranking of Avery, then Blake, then Cruz reads the same whether Blake is nearly as good as Avery or barely better than Cruz. A score ballot keeps that gap, and the count uses it.
To be fair to ranked choice, it doesn’t make you rank anyone you can’t stand. You can rank one candidate, or three, and stop. The difference is what your ballot can say. On a score ballot you can also give two candidates the same score. Most U.S. ranked ballots treat two candidates at the same rank as an error.
The trade-off the book accepts
A number asks a little more of you than an order does. The book thinks the extra information is worth it.
The critique
“A blank shouldn’t count as a zero.”
Leave a candidate blank and score voting counts it as a 0. But a blank often means “I don’t know this person.” It doesn’t always mean “I don’t want them.” Critics say that is unfair to candidates who are less well known.
A blank does count as zero, and that is deliberate. The other option is to average only the voters who scored a candidate. Then a candidate scored by a handful of fans could beat one scored well by almost everyone. A winner should have support from across the electorate.
The fix is in what gets reported. The author proposes that election offices publish blanks separately from zeros. Then every result shows how many voters rejected a candidate and how many didn’t know them. A candidate with many blanks and few zeros has a clear message: get known. A candidate most voters scored 0 has a different one.
The trade-off the book accepts
Candidates nobody knows start behind. They already do under every voting system. Publishing blanks makes the reason visible, so a newcomer knows the fix is getting known.
What Kenneth Arrow said, exactly.
Kenneth Arrow shared the 1972 Nobel Prize in economics. In voting, he is known for his impossibility theorem. It shows that once there are three or more candidates, no ranked voting method can meet a short list of basic fairness tests every time. Arrow proved what ranked methods can’t do. He did not invent a voting method.
In 2012 the Center for Election Science asked him which voting methods reformers should push. His answer, word for word:
“Well, I’m a little inclined to think that score systems where you categorize in maybe three or four classes probably (in spite of what I said about manipulation) is probably the best.”
The “in spite of” matters. Earlier in the same interview, asked about a 0 to 10 scale, he warned that “the incentive to misrepresent your vote may be high,” and said, “You probably need experience rather than theory.” He also favored three or four levels, not eleven. The book takes his answer as encouragement. It is not proof, and he didn’t offer it as proof.
Source: Kenneth Arrow, interviewed by Aaron Hamlin, Center for Election Science podcast, recorded Oct. 6, 2012 (transcript at electology.org)
Weigh the whole case.
The book makes these arguments at full length, with the real races, the numbers and the sources. Then you decide.