A recent report submitted for a ranked-choice voting (RCV) lawsuit in Maine highlights the deleterious effects of the system and notes there is little evidence that the purported benefits of RCV have come to fruition.
The report’s author, Nolan McCarty, professor of politics and public affairs at Princeton University, found that RCV elections led to lower “full voter participation” rates in Maine in 2018 compared to the state’s plurality elections.
In RCV elections, full participation means ranking at least all but one candidate, and not overvoting (i.e. selecting more than one viable candidate as the highest rank). McCarty found that in Maine’s 2018 gubernatorial primary, 2nd congressional district election and congressional primary––all decided by RCV––35 percent, 38 percent and 47 percent of voters, respectively, “fully participated.”
Meanwhile, full participation in a plurality election is defined as simply casting one valid ballot for the race. In the non-RCV gubernatorial race and state senate elections in 2018, around 97 percent of Maine voters fully participated.
The lower participation rate in RCV elections is attributed to ballot exhaustion. A ballot may be exhausted if a voter overvotes, undervotes by skipping more than one column/ranking or ranks only candidates who are no longer viable in proceeding rounds.
Drawing on previous research conducted by the Maine Policy Institute, McCarty examined 98 RCV elections from 2006 to 2019 and found that, on average, 10.8 percent of ballots casted were considered exhausted by the final round.
Further, McCarty determined that as the number of candidates increased, so did the rate of ballot exhaustion. This finding, he argues, is consistent with the idea that ranking candidates is confusing for voters –– and that confusion only increases as more candidates are added to the ballot.
What’s more, the report shows that RCV leads to even less participation than the second rounds of majority run-off voting, which requires voters to make a second trip to the polls. Fall off participation in non-RCV elections, McCarty concludes, is often less than the 10.8 percent average of ballots exhausted in RCV elections.
The report also indicates that those most at risk of ballot exhaustion are elderly people (those over age 65) and less-educated people (those without a college degree).
In examining voting data at the town level and matching it with demographic data of the town’s voters, McCarty finds that the town with the most senior voters truncates ballots (i.e. discards those that are not fully participating) at a rate almost 9 percent greater than the town with the least senior voters. Similarly, the least educated town truncates ballots at a rate 14 percent greater than the most educated town.
In addition to demonstrating the clear issues RCV raises, McCarty also examines the purported “benefits” of the system. He finds that most are non-existent in races in which RCV has been used.
He dismantles the claim that RCV helps smaller parties, for example, by citing an Australia election for single member districts of lower parliament. In their 2019 election, decided by RCV, only 6 of 151 seats were won by candidates outside of major party blocs. Meanwhile, in the British parliament, which decides elections through a plurality vote, 75 of 650 seats are held by non-major parties.
Moreover, McCarty shows that RCV does not significantly improve voter turn-out and engagement. In addition to citing several studies that prove voter turnout drops in RCV elections, he also turns to Maine’s 2018 elections. Eight of the fourteen most-contested state senate races, decided by plurality vote, had lower rates of total undervotes than the RCV elections.
Finally, McCarty dispels one of the central arguments for RCV –– that it allows a candidate to win by a majority. He says this is incorrect because after ballots are exhausted, the number of valid ballots used to determine a majority is less than the number of votes cast.
Winners, therefore, often fail to reach a true majority. In the 98 elections examined, he found this was the case for over 60 percent of winners, a near-identical match to Maine Policy’s finding of similar election data in 2019.