While virtually all political polls are designed to determine who is currently leading in the race to become our next President, we thought it would be interesting to go inside the minds of undecided voters. After all, some pundits consider them the gold that the major candidates need to win over in order to win the race, right? The data for this article comes mostly from a survey*, conducted by AYTM, just days before Americans cast their ballots for the next president.
In the weeks leading up to the election, undecided voters appear to represent 7% of the electorate according to an early October YouGov/Economist Poll. Our survey also found that the incidence rate for remaining eligible, undecided, likely voters was 7% with nearly 6,000 people screened out of our survey for not fitting those criteria. Please note that the AYTM survey ran during the first days after the latest ‘October Surprise,’ the announcement by FBI Director Comey that the Clinton email investigation had been reopened.
So what did we learn about this group of undecided voters?
We found that 31% were registered Democrats, 33% were registered Republicans and 36% were not registered to either major party. According to a Gallup Poll performed last month, 27% of Americans considered themselves Republicans and 32% considered themselves Democrats. Even though Republicans appear to be in the minority in the U.S., there are slightly more of them than Democrats represented in our survey. Perhaps this indicates that Republicans are having a more difficult time accepting Trump than Democrats accepting Clinton.
Of those who are registered Democrats, their best prediction reported today is that 42% will vote for Clinton and 18% will vote for Trump. Of the registered Republicans, 15% predict they will vote for Clinton and 47% will vote for Trump.
Among females, 28% predict they will vote for Clinton and 22% predict that they will vote for Trump. Among males, 27% predict they will vote for Clinton and 38% predict that they will vote for Trump. Of the segment in our survey who are neither registered Democrats nor Republicans, 27% predict they will end up voting for Clinton and 19% will vote for Trump. Thus, Clinton appears that she may have the edge with the independents.
We found that 60% of the undecideds have been undecided all along. 40% had been decided at some point during the race among the remaining candidates, but have since become undecided. 17% had intended to vote for Clinton at some point, but have since flipped to undecided, while 15% had done the same with Trump.
When asked to make a prediction, as of this moment, for whom they will ultimately cast their ballot, 28% said Clinton and 28% Trump. Johnson and Stein combined for 20%. 25% chose “Other,” which could pose a very interesting issue if a deciding number of votes is needed to tip the balance in favor of Clinton or Trump.
91% of the respondents have voted in a prior Presidential election. Of these, 32% reported that they sometimes were undecided a month before the election and 12% said it was common for them to be undecided this close to the election. There is a whole world of psychology around the subject of indecisiveness, and no doubt some of the undecideds suffer from this malady in a more generalized sense in their daily lives and will always be undecided in the days leading up to the election. In fact, 63% of our respondents admitted that it is possible that they would be undecided right up to the day of the election. The apparent good news is that 71% of those who reported that possibility expect to face it by going to the polls anyway and finally choosing a candidate. 21% can’t predict what they will do in that event and only 4% felt that it would lead to them not vote.
When asked why they are undecided at this point, 65% reported that they dislike all of the candidates. No surprise there given that the Trump and Clinton unfavorability ratings have been tracking in the mid to upper fifty percent range. 26% said that they were still studying each candidate’s positions and not prepared to make a final decision. Given how little of this race has been centered on issues, it’s difficult to imagine what they are still studying.
Only 34% reported that they were following the campaigns and news very closely. This correlates with studies in past elections demonstrating that undecided voters tend to be politically less engaged.
83% felt that this campaign season has been uglier than usual. It certainly has been unprecedented. When have we ever seen a major candidate, during a Presidential debate, threaten the other with an investigation and jail once elected?
More than two thirds of the survey respondents report that the tone and spirit of this campaign has contributed to their indecisiveness. Perhaps that is a significant contributor and helps explain why there are many more undecideds.
Now that we understand the undecided voters better, what do we at AYTM believe is their importance in the outcome of the election?
Undecided voters barely a week before the Presidential election probably represent little more than 7% of the likely electorate. Most of them believe that they still cast a ballot, but history suggests otherwise: in 2012, half of the undecided voters ending up not voting. So, it is likely that less than 4% of the undecideds will bother voting, and they will probably be mostly split between Clinton and Trump—with at most a one-point advantage to the winner of this group. Thus, except in closely contested states, we don’t side with pundits who believe that undecideds are the key to winning the race. We believe that if the race is close on election day, the candidate with the better “get out the vote” ground game will win the election.
* On October 29-30th, AYTM (Ask Your Target Market) ran a survey of 400 random, undecided voters who had not yet cast a ballot in the 2016 race for the presidency. All participants in the survey had to pass screening questions that confirmed that they are registered to vote, had not yet voted, considered themselves likely to vote, and were undecided at the time of the survey. The survey was performed on AYTM’s proprietary online panels and was purposely not balanced to census, weighted or otherwise intended to be a scientific poll designed to determine which candidate was currently in the lead.