How Voters May Shift Before Election: Debates Likely Last Chances to Sway Voters, but Undecideds Unpredictable

Barring any kind of major “October surprise,” terrorist attack, disaster, or domestic unrest, the VP debate and especially the two remaining presidential debates are the last major things between now and Election Day that have the chance to sway voters to move away from their current leanings. If things continue as they have been recently, among voters leaning towards anyone we can expect Trump to lose some support, Clinton to gain some support, and third-party candidates to lose some support to give Clinton an overall slight edge; the real question is what will the undecideds choose to do, because there are enough of them to crown either candidate the victor, but they remain unpredictable and, frankly, strange in their thinking and habits. Despite Clinton’s edge, then, this election outcome will likely remain unpredictable until its final days.

 Originally published on LinkedIn Pulse October 3, 2016  

By Brian E. Frydenborg (LinkedInFacebookTwitter @bfry1981) October 3rd, 2016

Pool/Getty Images

AMMAN  These debates worry and perplex me because I never know how the American public and the media will react. I know how rational people who aren’t blind ideologues will react. And at this point, all rational people and non-ideologues are backing Clinton. The almost 6 in 10 Americans that are irrational and/or ideologues constantly seem to prefer and encourage the atrocious. What horrifies normal, decent, civil, thoughtful people delights these “deplorables” to the tunes of shrieks and howls of delight, often with more than just a thin veneer of racism, misogyny, or some other form of bigotry and hate. So when by any objective measure Clinton easily won the debate and Trump more or less imploded, that is little comfort in trying to gauge the reaction of the mob and the can’t-see-the-forest-for-the-trees media.

Getting Inside Voter’s Minds

For the first two minutes, Trump seemed the most articulate I’ve ever seen him. For the first half-hour, he managed to not look terrible talking about trade and the economy. And then he demonstrated he had the attention span of an ADHD Millennial in giving us an hour of unfocused, incoherent rants in which he demonstrated no ability to exercise self-control as he was successfully baited every time Clinton tried to bait him to talk about unflattering and off-topic items. He acted unpresidential, demonstrated a solid lack of both understanding and information on pretty much everything, and was unable to fill his time without repeating himself over and over again and without parroting the same anecdotes he’s been peddling for over a year (we heard about Carrier, but I’m surprised he didn’t bring up Komatsu and Caterpillar as well…). Clinton was careful not to appear too aggressive and stayed calm, composed, seemed to be enjoying herself, and demonstrated a mastery facts and issues that has come to be a hallmark of her as a public servant.

Regardless of with whom you agreed more on this issues, this is what happened. And regardless of your thoughts on various issues and your party affiliation, it is clear that Donald Trump has a dangerous temperament, lack of self-control and focus, and a stunning ignorance that makes him unfit to be president and Commander-in-Chief. 

Any rational person can and did see and understand this.

And that is what frightens me, because I fear we rational people are outnumbered in this country. Along with irrational people who can’t see these obvious truths and hateful, childish, spiteful, narcissistic ideologues who either don’t care about the effects of their vote or are willfully blind to them (and I am convinced there are far more of the former than the latter), they may for a majority of voters. And they may not. We will see.

In the end, this debate and the ones to come, much like the election itself, are simply about math.

To Consider Impact of Debate(s), Divide Voters Into Parts of an Equation

In fact, it’s something of a long equation with various parts. Two parts of the equation are immutable: those who are definitely voting Trump and Clinton. Let’s call them Td and Cd. The there are those who are leaning but not certain, and who can be divided by how strongly they lean—high (Tlh and Clh), medium (Tlm and Clm) and low (Tll and Cll). The same thing can more-or-less be done for Johnson and Stein. Then we would have undecideds (U). 

I will cherry pick Ann Selzer’s outfit’s latest poll conducted before the first debate, as she is regarded by FiveThirtyEight as the “best pollster in politics,” to set the landscape: 43% of voters said they will vote Trump, 41% Clinton, 8% Gary Johnson, 4% Jill Stein, 3% were not sure, 2% “don’t want to tell” (I suspect these people are almost all Trump supporters), and 1% said they were not voting for a presidential candidate (if you prefer, feel free to do this exercise with the Real Clear Politics current average, which is very similar but with Trump’s and Clinton’s numbers switched and Stein significantly less than 4%, but I would venture that Selzer’s Bloomberg poll is likely more accurate). 

Now, the first main question is, whose minds can be changed and whose mind are past that point of being able to be changed?

Trump & Clinton supporters 

At this point I’d say a strong majority of Trump and Clinton supporters are going to stick with their candidate through hell and high water. I would say that all of Td and Tlh are going to turn out for him, and the same with Clinton’s Td and Tlh. That leaves medium leaners, light leaners, and the undecideds (I’d include those saying they won’t vote for president in this category) that I’d consider as people whose minds could possibly change. I think clear, overwhelming majorities—let’s say including 2/3-of the medium leaners—are going to definitely stick with their candidates in the cases of Trump and Clinton. That leaves roughly 1/3 of the medium Clm and 1/3 of the medium Tlm to be combined with the light leaners (Cll and Tll) for each to form the body of persuadable voters.

As I indicated in my last article, while there is an enthusiasm gap in favor of Trump meaning in theory that there is more possibility of Clinton’s people switching, I think that this is less likely because Clinton’s people are so anti-Trump and relatively practical that weak enthusiasm for Clinton does not meant there is a good chance that they will support Trump or someone else. If Johnson was performing better I would think he’d have a better chance of picking off unenthusiastic Trump supporters, but with the climate of fear in America intensifying, I think that effect will be mitigated. If anything, Clinton has a chance to pick up Republicans who are both rational and worried about Trump’s ability to conduct foreign policy, but I am not sure many of those people exist. So I’m sticking with only 1/3 of medium leaners being up for grabs for both candidates, in addition to all light leaners. 

As far as quantifying these, I would think that by far most Trump and Clinton voters are not light leaners, and I would think that she has more medium leaners than he does and that he has more high leaners, but I would say that the vast majority of supporters of both Clinton and Trump are definites. Including heavy leaners that I think have pretty much zero chance of leaning anywhere else, I don’t think all leaners combined for either candidate exceed 20% of their support and may be as low as 15%. I would with 100% confidence say that the 1/3 of the medium and light leaners who could actually change their minds together would not be as high as 10% and could be lower than even 5%; for simplicity’s sake let’s say it’s 5% of their support for each.

Stein & Johnson people

I think most Stein people are fanatics at this point, and selfishly care more about “feeling good” about their vote than whether or not Trump destroys many mores of democratic custom and does gods know what in terms of foreign policy. I also tend to think of Stein people as liberals who are either super into her or really just disgusted by Clinton and seek an alternative candidate, without many people in the middle, so, in other words, there aren’t a lot of Stein medium-leaners (Slm), so we’ll divide her support into Sd Slh and Sll, and I’d think her Sd and Slh to be about 2/3 of her support, with only 1/3 Sll.  Plus, Stein’s Green Party (as well as Johnson’s Libertarian Party) are drawing a lot of brand new support from people who have never voted third-party, so there is little party-loyalty and those voters disenchanted with Democrats flirting with Stein (and others disenchanted and flirting with Johnson) may, after a brief love affair, also become disenchanted with their new lover(s) the more they are exposed to her (or him); in fact, this is likely.

Johnson is more complicated than stein, though: he is attracting roughly half his support from people who would otherwise vote Clinton (I discussed this in my previous article), which is counterintuitive because as the Libertarian Party candidate he is far to the right on my issues, but is deceptively alluring because he is a pretty cool/fun guy, is very straightforward, and is cool on things Millennials really care about like weed and gay rights and is also stridently anti-foreign-interventionist. I would imagine most of his liberal support, other than misogynistic Bernie Bros, is actually up for grabs, and I think a good portion of his conservative support that doesn’t come from actual libertarians is also up for grabs because he has demonstrated himself to be atrociously and lazily uninformed and unengaged on issues of foreign policy. He will keep the conservatives whose biggest concerns are the size and scope of government, regulation, the debt, the Fed, etc., but conservatives who care a lot about foreign policy and are not isolationist may find themselves slowly moving to Clinton’s camp, especially after Johnson’s “What is Aleppo?” debacle when he demonstrated that had no idea about or had even heard of the Syrian city of Aleppo, site of some of the worst fighting of the ongoing civil war there, and his more recent seeming inability to name a single foreign leader he liked, despite being given many chances to do so and then mocking this himself as “having another Aleppo moment.” So, for Johnson, I think between 2/3 and 3/4 of his liberal support is up for grabs and about but I would think that 2/3 of his conservative support are solid libertarians or Republicans focused on libertarian issues, and since polling inferences show a strong likelihood that his support is split in half between liberals and conservatives, I’d say this means about half of his support is up for grabs, maybe even slightly more. Basically, the Jlh aren’t up for grabs with Johnson either, which I’d say with Jd are about half his support. Jlm, on the other hand, as with Stein, are virtually nonexistent so I would say about half his are weak Jll who I would bet are not only possible switchers but likely ones.

So at this point, for Trump’s “definite” support, I’d have an equation saying that = Td + Tlh + 2/3Tlm, and for Clinton, Cd + Clh + 2/3Clm.

For Stein, it would be Sd + Slh, and Johnson Jd + Jlh

Up for grabs, we have 1/3Tlm +Tll +1/3Clm +Cll + Sll + Jll

Now, before I continue, I will just point out that these following numbers are wholly non-scientific, and that this is more an exercise in thought and deductive reasoning. Now, going back to the Selzer poll, this means (again roughly) that Trump, with 43%, has 40.85% that is rock-solid; for Clinton, with 41%, 38.95% is rock-solid; for Stein, with 4%, 2.66% is rock-solid; and with Johnson, with 8%, 4.0% is rock-solid. Clearly, Johnson has the most potential to lose support, for the reasons I discussed, while it would seem Stein would have the most loyal supporters. I have a pretty high confidence on this admittedly rough analysis as being close to numbers that will resemble the outcome in November. This very roughly quantifies to [1/3Tlm + Tll2.15%] + [1/3Clm + Cll2.05%] + [Sll1.33%] + [Jll4.0%] = 9.53% that could change candidates plus another 4% equaling the undecideds/won’t vote people. Of those 2% who did not want to share, let’s give 1.0% to Trump as definite Td and 0.5% as Tll, and we’ll increase to 5.5% the number of persuadable leaners for Trump since they were reluctant to declare their support; we will imagine that the other 0.5% of this 2% could be for anyone and would also be up for grabs, though I suspect many of them are Trump people; to be more cautious that 0.5% will be added to undecideds for the sake of simplicity. 

A Loosely Instructive Equation

So, the full revised equations would look like this:

Adjusted Selzer #s: Trump T=44.5% Clinton C=41% Johnson J=8% Stein S=4% undecided U=4.5%

[Definites T + C + S + J]           + [Persuadables T + C + S + J + U]         = 100%

[Td +Tlh+2/3Tlm + Cd+Clh+2/3Clm + Sd+Slh + Jd+Jlh] + [1/3Tlm+Tll +1/3Clm+Cll + Sll + Jll + U] = 100%

[42.05%T + 39.95%C +2.66%S +4.0%J] + [2.45%T + 2.05%C + 1.33%S + 4.0%J + 4.5%U]  = 100%

85.67%                                                 + 14.33%                                         = 100%

Now, I realize this may come off as stupid and artificial and possibly pointless. Maybe that’s even fair. And I did mention that this is rough. However, I do think this does actually gives a logical and roughly-relatively-precise idea of what’s going on here. Basically, close to 86% of the electorate will not be changing their minds under any non-wild-card circumstances, and the remaining 14ish% are up for grabs, on a spectrum ranging from medium-leaners to those saying they will vote but not vote for president. I don’t think all these 14%+ people will change their leanings. In fact, at this point, if I had to guess, some of the medium-leaners (many in Clinton’s case) won’t be changing unless something dramatic happens, like a major terrorist attack on American soil or some sort of serious proof of sexual abuse or rape by Trump, etc., etc.. Let’s say that the 1/3 of the moderate leaners represent 1/3 overall of the persuadables for both Trump and Clinton; that would leave 2/3 of their persuadables with a moderate-to-high chance of being persuadable. The question is: where would they go?

Potential for Change Among Candidate’s Current Supporters


How Clinton could lose support

I think it highly unlikely any Clinton people would go to Trump or Johnson or Stein; she’s too cautious a candidate to do something that would cause people leaning her way to move to such dramatically different candidates; unless there is a major terrorist attack that would drive the weakest supporters of Clinton to Trump’s tent out of fear, I can’t see any of her people switching sides at this point except for maybe a tiny fraction of Millennials out of peer pressure, and I mean a tiny fraction because Millennials are so strongly supporting Stein and Johnson anyway. So I’d say at most 0.15% go combined to Johnson or Stein, probably evenly, but if this does happen it will like be less than 0.1% if it happens at all. Yes, there is the possibility of more Assange Wikileaks releases that could hurt Clinton, but the people who are likely to be animated enough by that to switch candidates away from Clinton are probably all already saying now that they are voting for that someone else.  

Likely Clinton losses: none

How Clinton could gain support

Where could she steal support? I think a tiny number of Republicans might finally realize that Trump is truly awful and does not have the temperament to command the world’s largest military and nuclear arsenal or to conduct negotiations that are going to be far trickier than any business deals he’s been part of, but not many people overall. Sill, with people like the very well-respected WWII and Korea veteran, former Republican Sen. Of Virginia, and former secretary of the Navy John Warner publicly endorsing Clinton, and with former President George H. W. Bush privately choosing Clinton, there could still be some Trump voters that might switch in these final weeks not just to Johnson or abstaining, but to Clinton if they are more conscientious of the bigger picture.  I also think that people who wanted to support Johnson but actually give a big damn about foreign policy will move to Clinton, and my hope is that some of the liberal Millennial support for him will also go to Clinton, but that remains to be seen. I do think she can pick off some of the persuadable Johnson and (particularly) Stein people, especially since most of their supporters are newcomers, as I discussed before.

Likely Clinton Gains: +1.33% from Stein, +1.5-3% from Johnson (+2.25% average), +0.25-0.75% from Trump (+0.5% average)

Likely Net Clinton: +4.08% average up to 45.08%


How Trump could lose support

Unlike Clinton, I think there is much higher chance of Trump losing not only his light leaners but also the medium leaners. I would think that the light leaners are looking for any reason to bolt that they can, while the moderate leaners are looking and hoping he will improve. But throughout this campaign, even when it seems like he is improving, he always seems to stumble again in ways that suggests he is incapable of consistently applying or internalizing any lessons over time in any kind of deep way. Thus, I think there is a high likelihood that Trump will lose most or all of his persuadable voters, at least if he has more performances that resemble his self-destructive first debate performance and continues to do things like tweet at 5:30 AM about a supposed sex tape of Miss Universe 1996, Alicia Machado, whom Trump weight-shamed repeatedly and who has come out supporting Clinton, something Trump did three days ago. It’s hard to imagine any of Trump’s people bolting to Stein, meaning they will divide between Clinton and Johnson and just not voting for president out of disgust.. I think the more they listen to him, read about him, and see him, the more not only the light persuadables but also medium leaners will be turned off and lose hope that Trump can be an adequate candidate.

Likely Trump losses: -0.25-0.75% to Clinton (-0.5% average), -0.75-1.25% to Johnson (-1.0% average), -0.5-1% to no-vote (-0.75% average)

-2.25% overall average loss

How Trump could gain support

Wild-cards like terrorist attacks, racial unrest, and an economic downturn could all help Trump and hurt other candidates, though to what degree would be hard to predict. But leaving wild-cards aside, Trump is likely to gain some support from Johnson as some conservatives become disillusioned with Johnson and begin to worry about the effects of a Clinton presidency on the conservative movement. I think this will offset the Johnson people who leave Trump for Johnson.

+0.75-1.25% from Johnson (+1.0% average)

Net average loss of -1.25% down to 43.25%


How Johnson could lose support

At this point, with his major foreign affairs gaffes and just plain goofiness, I don’t think it’s a question of if Johnson loses support, but how much support he loses. I think him only losing 2% and keeping 6% might be middle-of-the-road estimate, given that about half of his support is liberal and/or would go to Clinton in a two-way race. I think he loses between 3-4% more from his liberal wing of support, but with some Johnson people breaking late who are particularly concerned about the Supreme Court and would be terrified by Clinton’s judicial picks, thus pushing them to vote for Trump, he will lose some of his current conservative base as well; as mentioned, though, these will likely be offset by people hoping Trump will improve as a candidate who will leave him when he doesn’t.

Likely losses: -1.75-2.75% to Clinton (-2.25% average), -0.75%-1.25% to Trump (-1% average)

-3.25% average overall loss

How Johnson could gain support

Johnson is proving himself an inept candidate, but, as with the beginning of his candidacy, even as he will likely lose support in it its twilight, the best thing to happen to his poll numbers and support levels is Donald Trump. Anecdotally, a number of conservative intellectuals are saying that since they (and if you) don’t live in a swing state, they will (and you should) vote Johnson. The message is more muddled from those and to those who live in swing states. Still, a lot of the conservatives that will lose hope in Trump if does not improve his performance may switch to Johnson on “conscience” votes, but this group will be largely offset by people leaving Johnson.

Likely Johnson Gains +0.75-1.25% from Trump (+1% average)

Net -2.25% average loss down to 5.75%


How Stein could lose support

Stein’s core of support is a group of fanatics blind to reality. But some of her newer supporters who were seeking something better than Clinton will find the reality that Stein is anything but: a bunch of empty talk with no actual plans to move them forward and with positions so far to the left she makes Bernie Sanders look like a Republican (or, perhaps more accurately, a moderate Democrat). Thus, after exposure to her, it is hard to see any non-extremist Democrats sticking with her, so you can expect her persuadables to abandon her en masse as they realize that Clinton and Trump are so close that they will realize what the consequences of voting for Stein will actually be, and will return to Clinton as unenthusiastic Clinton voters. It’s also hard to see any of Stein’s people switching to Trump or even Johnson.

Likely Stein losses: -1.33% to Clinton

How Stein could gain support

Perhaps there is a slight chance that liberal Johnson voters will realize how conservative Johnson actually is on many issues and switch to Johnson as a second Clinton alternative, but I am skeptical of this, as any liberal supporting Johnson are not particularly aware or informed people to begin with, and the Bernie Bros won’t likely support a woman over a man.

Likely Stein gains: none

Net -1.33% loss down to 2.66%

Likely final scenario after persuadables settle: so base levels of support if things continue as they have been will probably boil down to a poll showing about 45% Clinton, 43% Trump, 6% Johnson, and 2-3% Stein: a slight edge for Clinton over Trump and diminished support for third party candidates that benefits Clinton. 

The real question will be how the undecideds will break…

Conclusion: Those Pesky Undecideds Could Decide the Election

You may have noticed that I avoided discussing the undecideds. That’s because as to undecideds, damned if I know. If Trump’s year of outrages aren’t enough to make him a clear non-choice to the remaining 4.5% or so of undecideds, I do not pretend to understand the psychology of such a person who looks at both Trump and Clinton and finds that a difficult choice to make. They could split evenly or go 2/3-1/3 or go almost all to one candidate or another. Perhaps some are torn between Stein and Clinton and Trump and Johnson, and I would think that fear of the other side (liberal/vs conservative) would drive those people towards the two major party candidates. Maybe some conservatives in particular who hate Trump so much but don’t like Clinton either will stay with Johnson. The Republican I-hate-Trump-but-not-enough-to-vote-Clinton-vote is a lock for Johnson, and that group is likely larger than the Democratic I-hate-Clinton-but-not-enough-to-vote-Trump that will be Stein’s core. This is not difficult to comprehend, even as it is also not too difficult to condemn.

But I honestly cannot fathom the mentality or even the existence of people who are unable to choose between Trump and Clinton. I can easily picture the ignorance—willful or otherwise—and/or hate that motivates people to vote for Trump because I have encountered that hate and ignorance far too often in human, corporeal form, and I can easily picture the idealistic pragmatism that animates the most passionate of Clinton supporters because in that, I am talking about myself; likewise, I can picture the liberals who dislike or even hate Clinton but who are voting for her anyway, because they care about still advancing a liberal agenda and because Trump; I can picture the moderate Republican veteran who sees Trump as wholly unfit to be Commander in Chief voting Clinton; I can picture conservatives obsessed with the size of government and young Bernie Bros voting Johnson, and I can picture the far-lefties voting Stein; I can’t for the life of me picture the person who is undecided between Trump and Clinton. And I make no predictions regarding how they will break, and I can’t make any predictions about if the undecideds are torn between Clinton and Trump or one of them and a third party candidate; it’s just one big giant mystery.

And an even bigger mystery is how they will view, and react to, the debates.

One thing I will say is that if my analysis is even close, Trump will have to win significantly more undecideds to beat Clinton and can’t afford to split them. And again, this analysis depends on a trajectory that continues as is without any big surprises in the homestretch.

Blue Nation Review

And if we’re trying to gauge what could be game changers for any voters, that’s where the debates come in. They’re pretty much the only things left aside from wild-cards that have the potential to change anyone’s votes. The VP debate will probably give Trump’s ticket a slight bump because Mike Pence is like every president in every B movie you can think of: bland, boring, but projecting strength; Kaine will almost certainly make better arguments, but his warm-fuzzy goofiness and sensitivity are not going to look presidential at all next to Pence or reassure undecideds. Though Kaine is brilliant and will no doubt perform very well, it is hard to see him win the optics battle next to a guy straight out of B-movie casting for the role of president. But then we can expect to see Trump continue to disappoint in the following debates, negating that bump. And we will still be stuck with trying to guess how very strange, indecisive people who seem to have to really mentally wrestle with whether to vote Clinton or Trump despite both candidate’s ubiquitous presence on any and all forms of media for month after month after month. With it almost impossible to guess how these undecideds will vote, the election is still up for grabs, though we can give Clinton an edge.

Between Brexit and the defeat of the Colombian/FARC peace deal at the hands of Colombian voters, we must make sure we who actually know what’s at stake get out and vote and to annoy our less mature friends to do the right thing. Democracy can be unpredictable, and can also be a scary thing, too; let’s make sure it isn’t this November.

If you appreciate Brian’s unique content, you can support him and his work by donating here. 

Feel free to share and repost this article on LinkedInFacebook, and Twitter (you can follow him there at @bfry1981), and here are many more articles by Brian E. Frydenborg. If you think your site or another would be a good place for this content, or would like to have Brian generate content for you, your site, or your organization, please do not hesitate to reach out to him!