This post is broken into a introduction on ranked choice voting followed by seeking to answer how ranked choice voting could have changed past U.S. elections. Feel free to skip the introduction if you’re already familiar with the topic.
Introduction to Ranked Choice Voting:
For those unaware ranked choice voting (also known as instant runoff voting (IRV) or the alternative vote (AV)) is a simple to understand method to hold elections in place of plurality voting (aka first past the post). It involves ranking the candidates in order of preference. Then when the election is counted the candidate with the fewest votes is dropped and those votes go to the those voters next choice until one of the candidates have 50+% of votes. It’s akin to automatically running a series of of runoff votes until one candidate has 50%. The reason this method is seen as better than simply electing than the U.S. and U.K. method of having only one round, is that it means no one wins without support of a majority of voters. Moreover, voting for a party outside the main two will not spoil an election. So it lends itself more readily to multiple parties which are more likely to collaborate with one another. See more information via Fairvote.
Now that everyone understands ranked choice voting, I want to pose the question what U.S. elections that have already occurred.
Of the soon to be 55 competitive U.S. presidential elections, 18 (very possibly 19 come November 9th) times a candidate won the election without securing a popular vote majority. It is thus tempting to ask which of those elections would have changed under a different voting system (namely directly electing the president with ranked choice voting). Of these 18, I estimate 6 of these elections selected a different winner than the majority preferred candidate (1844, 1876, 1880, 1888, 1912 and 2000). 4 of these six times in part because of a spoiler effect (the presence of third parties allowing candidates to win with less than 50% support) and 2 times because of the Electoral College (1876 and 2000). Note, if these elections were instead decided by plurality vote only 1876, 1888 and 2000 would have changed. Additionally, the four way election of 1824 would have selected the plurality winner but not the majority winner.
So how did I make these estimates? While we have only results of the first choice ballots, it’s possible to make inferences based on historical information on how these minor parties behaved through party actions, candidate statements, prior behavior among supporters and platform similarities. Naturally, we can’t know what would have changed had these elections explicitly been conducted via ranked choice voting, but we can at least make a certis paribus estimate of what would have happened.
Twice did the United States fail to elect the candidate who won the majority of the popular vote. This is the election of 2000 where Al Gore failed to best George Bush despite winning 51.3% of the vote. Additionally there is the election of 1876 where Republican Hayes won by one electoral vote despite the (Anti-Reconstruction) Democrat Samuel Tilden winning 50.9% of the vote.
This leaves us with four elections where I believe would have changed with ranked choice voting. In 1844, the newly minted Liberty Party pealed enough anit-slavery northerners from voting for Whig Henry Clay that Democrat James Polk won with 49.54% of the vote.
From 1880 to 1892, several minor parties existed. These parties mostly consisted of non-Southern populists affiliated the Democrats but broke away because of disputes over soft money policies (the Wizard of Oz debate between those who wanted a weaker currency (silver backed and greenback/fiat) and those who wanted stronger currency (gold backed). For example the Greenback party commonly and consistently ran fusion tickets with the Democrats. The Union-Labor party ran Democrat affiliated James Weaver for president. Therefore, I believe it’s reasonable to assume the majority of supporters of these populist parties (i.e. Greenback, Union-Labor and Populist parties) would have ranked the Democratic candidates as their second choices. They often ran on the same ticket! And so the Greenbacks (3.32%) prevented Winfield Scott Hancock (48.25%) from gaining more voters than James Garfield (48.27%) in 1880. Next in 1888, the Union Labor party (1.31%) could have put Grover Cleveland (48.63%) over Republican Benjamin Harrison (47.80%) in popular voters. However, this would not have changed the electoral results as only in Indiana did the Democrat lose by a smaller margin than the 3rd party vote share.
Fourth and last, in 1912 former President Roosevelt ran as a 3rd party against Republican Taft and Democrat Wilson. Roosevelt could have easily overtaken Wilson’s 41.84% vote share both in popular and electoral voters with only the help of Taft’s supporters let alone the minor parties.
There are other very close elections like 1960’s defeat of Nixon or 1884’s first win by Grover Cleveland. However, my best estimates point to that the actual winner would have also narrowly won the in a ranked choice voting format as well. Finally, there is the election of 1824 where Jackson won the plurality of popular and electoral voters but John Quincy Adams was selected by the House. It seems like had Quincy Adams ran instead of the other candidates like Clay, Adams would have defeated Jackson. And so the House incidentally choose the majority if not plurality preferred candidate (an outcome that is unlikely to repeat itself).
I wanted to highlight these historical elections to show where ranked choice voting could have changed the Presidential outcome. Furthermore, it is because of a combination of the Electoral College and presence of third parties that a probable six times in U.S. history the candidate without majority support became president. Additionally, the Electoral College does not favor either Republicans or Democrats for more than a few elections at a time. That its third parties regardless of ideology (Libertarians and Greens) are capable of spoiling an election. It was simply Al Gore’s narrow electoral loss which gives too many the false impression that the Electoral College intrinsically favors the modern Republican Party. Lastly, that ranked choice winners historically do not mean better presidents (see 1876!) — only I would argue presidents selected by fairer and more democratic elections. Hence if there is a call to action from this post, it’s for electoral reform because it does have practical benefits.