Peer Reviewed Research

"We find that compared to Plurality Voting, all methods tested substantially reduce the need for voters to consider candidate electability but that STAR Voting, Smith/Minimax, the Condorcet method included, and Approval Top Two go further to ensure an equally weighted vote. We conclude that STAR Voting in particular offers a viable, competitive, and compelling proposal for more representative elections."

"We find that Instant Runoff Voting incentivizes candidates to appeal to a wider range of voters than Plurality Voting, but that it leaves candidates far more strongly incentivized to appeal to their base than to voters in opposing factions. In contrast, we find that Condorcet methods and STAR (Score Then Automatic Runoff) Voting provide the most balanced incentives;"

"The results of this election displayed a well-known but relatively rare phenomenon known as “Condorcet failure:” Nick Begich was eliminated in the first round despite being more broadly acceptable to the electorate than either of the other two candidates. ... We find that under the best assumptions available about voter behavior, it is likely—but not at all certain—that Peltola would still have won the election under Approval Voting, while Begich would almost certainly have won under STAR Voting."

"This research provides empirical estimates of rates of improper marking and vote rejection, and compares them to those rates on non-ranked offices (particularly single-mark, ‘choose-one-candidate’ offices). The data show that in a typical ranked choice race, nearly 1 in 20 (4.8%) voters improperly mark their ballot in at least one way. We argue that these improper marks are consistent with voter confusion about their ranked ballot, and find evidence that this mismarking rate is higher in areas with more racial minorities, lower-income households, and lower levels of educational attainment. We further find that votes in ranked choice races are about 10 times more likely to be rejected due to an improper mark than votes in non-ranked choice races."

"From the perspective of social choice theory, ranked-choice voting (RCV) is known to have many flaws. RCV can fail to elect a Condorcet winner and is susceptible to monotonicity paradoxes and the spoiler effect, for example. We use a database of 185 American ranked-choice elections for political office from the years 2004 to 2022 to investigate empirically how frequently RCV’s deficiencies manifest in practice. Our general finding is that RCV’s weaknesses are rarely observed in real-world elections, with the exception that ballot exhaustion frequently causes majoritarian failures. ... even if one finds majoritarian failures unimportant, given the high rate of such failures (even in non-truncated elections) it is likely advisable that RCV advocates adjust their rhetoric around RCV. For example, (Lair 2022) states that “the majoritarian principle is an axiom of democratic government” and uses this statement (among others) to justify the adoption of RCV. ... Our results suggest that RCV does not live up to such statements in practice. ... whether RCV’s benefits outweigh its costs is, in our view, still an open question"

"It has long been recognized that Instant Runoff Voting (IRV) suffers from a defect known as nonmonotonicity, wherein increasing support for a candidate among a subset of voters may adversely affect that candidate’s election outcome. .... We conclude that monotonicity failures in three-candidate IRV elections may be much more prevalent than widely presumed (results suggest a lower bound estimate of 15 % for competitive elections).In light of these results, those seeking to implement a fairer multi-candidate election system should be wary of adopting IRV."

The table compares the Condorcet performance of non-Condorcet voting methods. Condorcet performance measures how often the method will elect the candidate who wins all head to head match ups. Despite this paper being framed around the benefits of strictly Condorcet systems, STAR Voting still performs remarkably well.

In Study 2, STAR Voting disagreed with Condorcet in 19% of trials (18,602 / 100,000), whereas Ranked Choice Voting (Hare) disagreed with Condorcet in 48% of trials (48,495 / 100,000). 

These trials are far more competitive than any election we'd expect to see in the real world, so these numbers should be interpreted as a relative comparison of each voting method's performance instead of being predictive of real world performance.