Should Irvine voters approve Measure D? (November 3, 2026 ballot)

What's this about?

People disagree about whether Irvine should use ranked choice voting for mayor and City Council races.

Voters would rank candidates from first choice to last choice.

What supporters say

• The measure adds rules for voting tools, costs, ballot mistakes, vote counts, and public reports.
• The winner could gain more than half the votes still in play after each round.
• Ranking backup choices could stop two similar candidates from splitting their shared support.
• Some results from other places challenge the claim that ranked ballots confuse most voters.

What critics say

• A cost limit does not show that the whole plan will cost little money.
• Ranked voting needs more voter teaching than the current system.
• Some ballots may stop counting if voters rank no more choices after their first pick loses.
• The system can sometimes create results that surprise voters.

How to read this

The number of points on each side does not show who is right; the strength of the evidence matters more.

Measure D would bring ranked choice voting to Irvine’s mayoral and City Council elections. The evidence shows plausible benefits, but it does not establish that those benefits would outweigh the city’s local costs and implementation risks.

The case for

Ranked choice voting could reduce spoiler effects. Because voters can rank backup choices, support for similar candidates would be less likely to split the vote and hand victory to another candidate simply because a smaller candidacy drew votes away. A municipal comparison and a compilation by FairVote support this possibility, though FairVote is an advocacy group and the comparative study provides limited methodological detail. The evidence therefore makes the benefit plausible, not conclusive. 1

The system could also produce a winner with majority support among the ballots still active in the final round. When the lowest-ranked candidate is eliminated, that candidate’s ballots can move to voters’ next available choices. This differs from plurality elections, where a candidate can win with less than half of the votes cast for candidates. An Alaska election report illustrates how these transfers work in practice (see Figure 1). 2

RCV gives voters more than one meaningful way to express their views. Someone could support a preferred candidate while also identifying acceptable alternatives. In New York City, nearly 90% of voters ranked more than one candidate in at least one ranked-choice primary, suggesting that many voters used the additional options, although that statistic does not show that they felt better represented or more satisfied. (see Figure 2) 3

Measure D also includes safeguards before implementation. It delays the change until conditions involving voting equipment, county election coordination and costs are met. It sets requirements for ranking capacity, write-in candidates, ballot errors, manual counting and public reporting. It also caps additional eligible costs at 0.23% of the applicable general-fund budget. 4

Research on voter confusion is mixed. One peer-reviewed study using real election data found that some concerns about complexity, strategic voting and exhausted ballots may be overstated. Another result summarized by FairVote found fewer invalid ballots and smaller demographic gaps in ballot errors under RCV than under a traditional single-choice ballot, though the underlying study is not directly included in the available evidence. 5

The case against

The strongest concern is that some ballots may not count in the final round. A ballot can become “exhausted” when all of its ranked candidates are eliminated or when no further preference is available. As a result, the final-round winner may have a majority only among continuing ballots, not among everyone who originally voted. 6

RCV also asks more of voters than choosing one candidate. Voters must understand rankings, skipped or repeated choices, transfers and when later preferences will be used. Alaska research and interviews with election administrators suggest that success depends on sustained voter education and a reliable information campaign. 7

Administrative mistakes could damage public confidence. Ranked tabulation requires careful testing, data management and communication. In New York City, test ballots left in the election system temporarily produced incorrect RCV results before officials corrected them. The episode shows how even a repaired error can undermine trust. 8

The cost limit does not prove that overall costs will be low. The available analysis does not estimate future spending on equipment, software, staffing, audits, voter education or county coordination. The cap is a ceiling on certain additional costs, not a full estimate of the system’s lifetime cost or a demonstration that its benefits will exceed them. 9

RCV can also produce counterintuitive results. Under some preference patterns, ranking a candidate higher can cause that candidate to lose, a mathematical problem known as a monotonicity failure. The evidence shows that this can happen, but not how often it would matter in Irvine. 10