Who are the undecideds?

Politics & Elections
1

Inspired by Sam Wang’s PEC post here, which was interesting but still left me unsatisfied.

In short what he did was to look at where the undecideds are and found that they are more concentrated in states that had gone to Romney.

Slightly longer version - at this point in 2012 about 10% were undecided and now about 15%. He graphed the states’ current margins against the Obama/Romney final margins and concluded that the a failure to “bring undecideds home” would hurt Trump more than it would hurt Clinton. That 5% increase in undecideds all going to Clinton would lead, he concludes, to a Clinton 381 EV, Trump 157 EV race, and all going to Trump would lead to a Trump 318 EV, Clinton 220 EV.

And yes ultimately it is individual state performances that matter. But …

It seems to me that his data in some way actually shows more upside for Trump than down: more of the undecideds’ “home” to return to is voting GOP. If both bring their portion of undecideds “home” Trump’s increase is more than is Clinton’s. Yes mostly that would means keeping Red states more Red , but it really does get granular depending on exactly who the undecideds are. The undecideds in PA, for example, are not necessarily voters whose “home” last time was Obama just because it off that margin.

Not much data out there on that that I can find.

The Reuters/Ipsos Polling Explorer has its limitations but an advantage of their technique is a large enough pooled n that they can offer us the ability to analyze by various filters.

Not much difference I can find on numbers of undecideds (two-way only) using different filters on the “likely voter” universe. Yes Voted Obama is 16% now undecided to Voted Romney now undecided being 22%. And “some college” or less about 22% compared to “college” or greater 19.5%. But by lean conservative to very vs lean liberal to very a bit more undecideds on the Liberal side, 20.6 to 18.4. Whites by age under 40 or 40 and over? About the same.

Nevertheless mostly consistent with the idea that current numbers (Clinton up about 5) include a fair number of likely voters who went Romney before now saying undecided and who might be able to be brought back.

So, demographically, who, do you think (data if you have it), are the various undecideds, and what do you think will make them decide how to vote (or if not to) if they don’t have enough to decide by now?

2

Oh.

Also.

Please, if you are an undecided, please contribute what your demographic is and what factors would go into your personal choice. Thanks.

3

As a number of observers have pointed out, there’s been some tendency for Trump to run better in key states, as far as available polls, than he has nationally. And the difference seems to be mainly narrower Trump leads than Romney’s victories in states Romney won handily, and which Trump is unlikely to lose if he remains competitive nationally and in the key states.

The tendency mentioned seems less right now where the RCP national average today closed to 4.7 against Trump not much different than the actual 3.9 national popular loss by Romney. And more recent state polls in NC, FL, OH and PA are more generally in line with that. Now Trump’s fractionally behind in NC (Romney by 2), down 3.7 in FL (Obama by 1), 2.5 in OH (Obama by 2), with PA being the major (apparently, this time) swing state where Trump is doing relatively better than Romney down 2.3 v Obama won by 5.

Anyway it would seem the focus on ‘undecided’ like voting in general should be on those states. Trump isn’t that likely IMO to win all of them which he’d pretty much have to (give or take a handful of other plausible substitutes like VA or NH). But if it’s close, those are the places which count, and some conservative Romney voters in AZ or KS staying home isn’t likely to matter.

Also though it may go without saying, ‘undecided’ can also correspond to respondents who have passed through a polls ‘likely voter’ screen (where that’s applied, a lot of polls this far out are ‘registered voter’) but aren’t going to vote.

4

I’d like to think they are just low-information likely voters, but I guess I’m, for the first time, in the ‘undecided’ category.

Hilary is the sane choice for me, but there is something appealing about having someone in office that is not a puppet of a party. I couldn’t stomach seeing a true conservative in office, but I could handle seeing Trump there. The buffoonery would at least be entertaining. But when I start thinking about the Supreme Court ramifications and how I’d like the court to view itself, it’s Hilary all-the-way. Double rainbow. Her “scandals” don’t bother me at all… she’s made bad decisions maybe, but I haven’t seen anything nefarious from her.

Then again, I don’t take my vote too seriously as I’m in the deepest of the deep-red state. If I lived in Ohio or Florida, it may be a different story.

5

The problem with isolating the undecided vote is that it is likely to be equaled or perhaps swamped by the number of committed voters who either make a special effort to vote on November 8 or decide it’s not worth their bother. I’d be surprised if their number were not also five percentage points of the total.

I understand the desire to look at parts of the electorate and try to gauge their meaning. It’s like astronomers looking at refining Sedna’s orbit rather than plotting all the planets and dwarf planets at once - a huge task in and of itself with monstrous uncertainties.

Wang’s article doesn’t in the end say much more than what everybody else has been saying. **Everything **has to break Trump’s way in order for him to stand a chance. Undecideds alone won’t do it.

6

I’m undecided, for the first time in 27 years I’m considering not voting. Similar to Nate, Clinton is my obvious choice, but I really , really, really don’t want to vote for her, and I’d take any reasonble alternative.

FTR, BLACK MALE, 48, 6 figure anual salary

7

Stuffy, how have you voted in the past and what could happen to impact your choice?

Exapno, yeah, that was why I ended up unsatisfied by Wang’s take.

Clearly what happens with undecideds is not the single issue of the election, but I would like to have a better grasp on how big of a potential factor they could be.

Again that Reuters/Ipsos tool has significant limits but taking it at face value looking at past Obama vs Romney voters, there are 9% of 2012 Obama voters who say they will vote Trump and only 3% of past Romney voters who will vote Clinton. IOW past Romney voters who dislike Trump relatively are more not going all the way over to Clinton and are holding as “undecided” whereas as past Obama voters who dislike Clinton are relatively more willing to go all the way over to Trump.

Will turnout among supporters be a bigger deal? Very likely so. I think the biggest issue for pollsters is going to be figuring out who actually is a likely voter this time.

Nevertheless the fate of current undecideds is a more significant issue this cycle than last time. Those who have decided are pretty solid in their decisions, more than in most years.

8

OTOH, what percentage of “undecided” responses can be assigned to a variation of “Take your annoying, intrusive, obnoxious poll and shove it up Gallup’s ass”?

More seriously, if a person \doesn’t want to stop and engage a pollster but still clings to civility, something like “I haven’t made up my mind, goodbye” would seem to be the default (if mendacious) answer.

9

The bottom line to me is this: Romney and McCain lost. And where can you look and find voters who didn’t vote for McCain or Romney but are going to turn out for Trump? Trump’s having a hard time holding on to traditional Republican voters much less winning over Democrats to his cause.

To me a lot of those undecideds are probably people thinking “I usually vote for the Republican candidate but I might not this year.”

10

Except that everything we know about voters, from every kind of interviewing situation and every type of study, shows that many people truly do not decide who to vote for until the last minute. They are “low information” voters, true, but more importantly they are low involvement voters. They do feel that voting - at least in presidential elections - is somehow important or necessary or proper but they don’t engage with the process. They don’t follow the horse race, they don’t discuss politics with others, they don’t have a commitment to a party line. Something will steer their vote to a particular name in November but they couldn’t tell you whose in July.

Why shouldn’t we believe they exist? It would far more remarkable if we couldn’t find 10% of the voting population who acted like them.

11

Well here’s some data on who were the undecideds in 2012 anyway.

:slight_smile:

12

I’ve voted democrat almost exclusively. I switched to independent almost 15 years ago because I feel the democrats have moved too far right. They could move back left, and I will acknowledge recent overtures by Clinton in that direction. I just find it less than believable due to the way they threw previous constituencies under the bus in the name of political expediency.

13

Please vote. Although EV’s decide the contest, the popular vote number sends a message, and the Republicans not only need to be beaten, but beaten badly, so badly that they change course substantively.

ETA: Plus the Senate’s in play.

14

If your state is safely blue or hopelessly red, and you plan to use your presidential vote to send a message, why not send the message you actually want to send? I’ve spoken here several times about the need for progressives to back Clinton; but if I lived in a state whose EVs were locked up, I’d probably vote for Jill Stein instead.

15

My message is I actually think Hillary Clinton is the best choice. If the choices were Clinton, Cruz, Johnson, Sanders, Stein, and Trump and the votes were tied and I got to choose who would be the next President, she’d be the one I’d pick.

16

Personally agreed, my "message"would be that I want Clinton as president, one, and two, that our country rejects wholeheartedly and overwhelmingly the message that Trump represents. But appreciate that some are not thrilled with sending the first message (even though given the actual choices they, perhaps with great reluctance, concur), and want to think that they can send a message they are not happy with the choices they have.

They know who they want to vote against but they still are not crazy about who that means voting for.

If you can get to the graphic in this WSJ article it is telling. It putatively is more about jumping lines than about undecideds but what they did was poll that compares relative performance of trump and Clinton compared to a generic Congress race in their party. Clinton overperforms a generic Democrat by 27, 20, 14, and 12 points in college-educated White Republican women, non-college-educated White Republican women, and college-educated White Republican men respectively. Trump overperforms a generic Republican by 14, 12 and 10 points in college-educated White Democratic men, non-college-educated White Democratic men, and non-college-educated White Independent men respectively. That’s a fair number of people who are not happy with their usual party’s standard-bearer. Some of those make up the undecideds.

The NYT is putting it this way:

2000? That’s when another Clinton ran for the first time. People do often forget that WJ Clinton was only found to be a satisfactory choice by 33% of the electorate at ths point, running less than GH Bush was and 10 points less than HR Clinton is. Not sure though that those not happy with either of them voting Perot sent any sort of message that was intended.

17

Erm… you’re off by 8 years.

18

2000? Don’t you mean 1992?

Generally, the undecided are those who don’t follow politics that closely or those that haven’t admitted even to themselves who they’re going to vote for. Most of us vote for the same party for president as we always do, a few switch back and forth, and a few genuinely make up their minds at the last minute. For the latter group, appeals to them based on logic are a waste of time as they simply don’t take it seriously enough that they even want to vote based on reason.

19

I think Hillary Clinton is a political opportunist, which makes her unappealing in the eyes of many. For years, being an opportunist in the democratic party meant that the center-right and center-left vote was the ‘baby’ that needed nurturing, leaving only scraps for the true left.

I get that people don’t like the fact that Hillary is shifty, but it might be her ability to shift and to adapt to political mood changes that makes her particularly appealing at a time when the opposition party is becoming intransigent to the point of insanity. I agree that Clinton is not a strong liberal – she probably needs a push to become one. But she is absolutely a strong anti-conservative, and that’s a good start. Few people have as much of a burning hatred for republican conservative radicals as Hillary Clinton – so I’d argue that’s a good quality. Her malleability makes it possible for her to be pushed by people like Elizabeth Warren, Barack Obama, and even Bernie Sanders. That’s another good quality.

Donald Trump would be a disaster from which this generation would not recover. We already elected one low-information president in George W Bush, and that led to the worst economic and national security decisions over the last 100 years. And yet, George W Bush was arguably more prepared to be a president in 2001 than Donald Trump would be next year. Americans seem to have this disdain for politicians like Hillary who seem too complex, but we’ve had the simple guy you’d invite to a backyard barbecue and it was an unimaginably horrible clusterfuck. Donald Trump is not the guy you’d invite to your barbecue; he’s more like the nature boy Ric Flair, screamin into a microphone and bragging about his Armani suits and Rolex watches while in reality he’s pretty much broke. Either way, not the right choice for the country. More than that, people have a duty to vote against him. Otherwise, they’re voting for him.

20

Apologies but in addition to myself I blame the NYT this time. They mention 2000 as being the past low point and the Pew graph shows that the low point was WJ Clinton v GH Bush (yes in '92) while 2000 actually was a high voter satisfaction year.

So year wrong, but people right, and point as intended.

Still my apologies.

Also per the NYT article … this year the undecideds are actually following closely. Maybe they do not want yet to admit to themselves.

The undecideds are a bigger number than in the past. I suspect that some of them are different than past undecideds.