A recent poll by Public Policy Polling (PPP) on the state of the New Hampshire Republican Presidential Primary shows Donald Trump retaining a strong lead, but it also shows how a ranked choice voting (RCV) election would leave Trump in a neck-and-neck race with Senator Marco Rubio.
While many in the Republican Party have lamented Trump’s ability to stay well atop the polls despite being viewed unfavorably by many within the party, this is only possible because the support of those who oppose Trump is divided between a number of other candidates who could spoil one another’s chances at success. Without a single candidate who can unify the opposition to Trump, it is entirely possible that states using winner-take-all primaries could award all of their Republican delegates to a candidate that most of the party’s voters oppose.
This sort of unrepresentative outcome is an inherent flaw in many American elections in which earning a simple plurality of votes is sufficient for victory.
Ranked choice voting provides a solution to these problems, by allowing voters to express their preferences between a range of candidates, instead of merely selecting the candidate they like best. If a voter’s first choice candidate does not have enough support to win, that candidate is eliminated, and their vote instantly goes to the candidate they ranked next.
This process is repeated until only two candidates remain, at which point the candidate with the most votes is declared the winner.
The recent PPP poll from New Hampshire allows us to simulate an RCV election between four of the top candidates in order to reveal more about New Hampshire Republican Primary voters’ true preferences.
While we have previously lauded PPP for collecting data on voters’ second choice candidates, in addition to their first, the most recent poll expands on this information by also gauging what voters’ preferences would be if the field were limited to different groupings of three or four of the top candidates, making a more detailed simulation of a ranked choice voting election possible.
The table and chart below break down the results of the simulated RCV election. Among voters' first choices, Trump has a clear lead, with more than double the level of support of the second place candidate, Ted Cruz. However, as weaker candidates are eliminated, the bulk of their support flows not to Trump, but to other top candidates.
When the field is narrowed to four (Carson, Cruz, Rubio, and Trump), Trump's lead is cut considerably, but he still comes out on top with 36.4% of the vote, compared to 28.4% for Rubio, 20.5% for Cruz, and 14.8% for Dr. Ben Carson.
In the 2nd round, when Carson is eliminated, the votes of nearly 80% of his supporters transfer to Rubio or Cruz, and the former closes to within a few percentage points of Trump. When Cruz is finally eliminated in the 3rd round, a large majority of his supporters have their votes transferred to Rubio, leaving Rubio and Trump to split the vote 50%-50% in the final round.
Clearly, the simple sample of first choices taken in most polls, and indeed, most elections, is not sufficient to give an accurate picture of voters’ preferences. While Trump has a commanding lead in first choices, it appears that just as many voters would ultimately oppose his candidacy as those who would support him.
In the Republican primaries, it may take a number of other candidates dropping out for anti-Trump voters to unify around a single candidate. If they do not, a candidate that a majority of Republicans oppose could become the nominee.
We applaud PPP for collecting data that provides a fuller picture of voters’ preferences. As RCV elections illustrate, allowing voters to express a fuller range of preferences allows us to better assess voters’ sentiments and the majority’s will.
Editor's note: This post, written by Andrew Douglas, originally published on FairVote's blog and may have been modified slightly for publication on IVN.