How credible is the "Shy Trump voter" theory

[rafta_rafta]

The Trump campaign has been pushing a narrative of "the shy trump voter" who decides to cast his vote for Trump despite publicly not acknowledging so.

The same strategy was used by the Romney campaign to explain his deficit in the poll numbers and turned out to be totally bogus.

But since the 2016 campaign has been so unlike others in the last few presidential elections, could the polls really be underestimating Trump's vote share?

[Atlas Has Shrugged]

I actually think it is a lot more realistic with Trump considering what he says. As a neighbor of mine who I met last night was saying during our hour long conversation (he too is a poli-sci major at FSU), it can be very hard to be a Trump supporter some days.

Then again, I think most "shy Trump" voters are actually Republicans who are feigning to be "NeverTrump" but secretly are still partisans at the end of the day. I don't think most "shy Trump" voters who are registered as independent would be as shy. This is just anecdotal of course, but I have always found that the most outspoken people I know in politics, be they far-right, far-left, centrist, etc, are almost always independent.

[NOVA Green]

I think that this theory has been debunked somewhat on 538.Com based upon no significant variance based upon polling methodology that would frequently favor the "shy voter" theory.

Now one might argue about 538 and Nate's accuracy this year, but it does seem to make sense to me conceptually.

[Erich Maria Remarque]

I think that this theory has been debunked somewhat on 538.Com based upon no significant variance based upon polling methodology that would frequently favor the "shy voter" theory.

Now one might argue about 538 and Nate's accuracy this year, but it does seem to make sense to me conceptually.

Link? Don't remeber them debunk it.

[elcorazon]

not credible at all. Frankly, you could argue that there might be some shy Hillary voters too! Could balance out if there were such a thing.

[Erich Maria Remarque]

Shy Trump effect exists. Common, is it so difficult to imagine a high educated whites [men] who are not ready enough to admit. Ecpecially cellphones = Trump supporter being with his frinds/in bus etc etc, while answering poll.

The problem is that we can not estimate it. In other countries it usually 1-4 pps.
+ Trump is VERY controversiall. VERY.
+ Clinton Bradley classisk effect.
-  Shy Clinton effect exists as well, but probably no so large as Trump's
- Hispanics might be underestimated as well.

Indication:
+ latter part of primaries.
-  former part of primaries.
-  difference between online polls and regular ones have increased lately (I'm not sure, read it                               somewhere, but can't find it for now).
+ high number of undesiders/third party.
-  Clinton is also responible for this ↑

So that effect is real, but can be cancell out by many other things. For intance by underpolling hispanics in some states. So I would say 0-4% effect.

[Flake]

If anything, Trump supporters are much more vocal than most other voters.

[elcorazon]

It all depends. I feel like if you are in many less urban settings, your peers are mostly Trump supporters who HATE Hillary. Saying you're for Hillary is similar to saying you're a communist and you hate this country. Not to mention you're voting for the fatigued, weak woman rather than the strong man with endurance. How could you admit it?

[Erich Maria Remarque]

If anything, Trump supporters are much more vocal than most other voters.

We're not talking about rally-like voters. Those will probably come with placate and sing a song about him while voting.

[Xing]

I think it's been mentioned before, but Trump supporters are generally far from shy. I think there are probably more voters who might be shy to admit that they're voting for "Lyin' Crooked B**** $h*tlery."

[Erich Maria Remarque]

I think it's been mentioned before, but Trump supporters are generally far from shy. I think there are probably more voters who might be shy to admit that they're voting for "Lyin' Crooked B**** $h*tlery."

Again you are talking about just one [tiny?] part of his base. Is Whie educated not such group?

[EliteLX]

You can not compare this effect with Romney.

People are genuinely afraid to support him in fear of backlash or voice their opinion in the slightest. This has not happened in previous cycles, because the only serious backlash for GOP support was always from some left-wing nuts, blatant liberal hacks, or MSNBC lol. This cycle you have even 16 year olds on Twitter just ripping anything related to Trump to join the wave.

[Seriously?]

My guess is that it is more of a function that not many "live voter" polls have come out since early August, so the numbers that we are using for comparison purposes are a little bit dated.

My guess is that the "shy Trump voter" theory would count for a few percentage points, at most, and would be more of a phenomenon where a voter goes to the poll conflicted or slight lean Hillary and pulls the lever for Trump at the end of the day despite social pressure to pull the lever for Hillary.

As such, it's overstated.

But it will be interesting to see the variance between Internet/IVR/Live Voter polls over the next few months to see if I am wrong.

[An American Tail: Fubart Goes West]

I think there will be a point of 2 of Shy Trump voters, but there will also be a point or two of Shy non-native English speakers for Clinton.  I don't think polls are going to be far off at the end of the day, but Trump could beat his polls in the East while Clinton beats her polls in the West and in Florida.

That would be interesting to see.

[136or142]

I think this is just Trump setting the groundwork for the "Hillary Clinton stole the election" post election response.

[NOVA Green]

I think that this theory has been debunked somewhat on 538.Com based upon no significant variance based upon polling methodology that would frequently favor the "shy voter" theory.

Now one might argue about 538 and Nate's accuracy this year, but it does seem to make sense to me conceptually.

Link? Don't remeber them debunk it.

ok--- haven't found the particular mention that I recalled in one of the articles posted, but the theory goes something like this...

If there is a "shy Trump" voter effect than there would be significant variance between automated versus landline/cell person-to-person based polls.

Unfortunately, I haven't been able to find the exact reference (Which I believe was contained within the body of a larger article over the past two months), so maybe one of our Atlasians that has a better exact recollection than an aging man like myself, might be able to pull up the exact article which I do recall reading in the recent past.

That being said, Trump has not been doing particularly well in automated polls versus "live-caller" polls, which is where typically one sees the "shy voter" phenomenon that doesn't feel comfortable telling a real person that they support Trump, or any other political leader that might be extremely unpopular among their friends/family/neighbors/coworkers, etc...  (David Duke in the early '90s anyone???)

So, although I don't have the exact article that I recall reading at my fingertips, here are many other 538 posts regarding aspects of "automated" polling versus other forms.

http://fivethirtyeight.com/features/election-update-its-too-soon-for-clinton-to-run-out-the-clock/

http://fivethirtyeight.com/features/election-update-leave-the-la-times-poll-alone/

http://fivethirtyeight.com/features/election-update-national-polls-show-the-race-tightening-but-state-polls-dont/

http://fivethirtyeight.com/features/election-update-clintons-lead-is-clear-and-steady/

http://fivethirtyeight.com/features/the-state-of-the-polls-2016/

[OneJ]

Less credible than Hillary's actual "Silent Majority".

[ProudModerate2]

There could be a small amount of individuals in the populace, that are "shy trump" supporters to their family, friends, neighbors, co-workers, etc.
But not really in polls that are conducted.
Polls are almost entirely private, with no one else around when the poll is conducted and the individual is providing answers. Thus there is practically a zero adjustment necessary to poll numbers, when considering these shy voters.

[Wisconsin+17]

Interesting argument, that would argue that we are not actually going to sit out. Heck, I did it for Romney.

[Maxwell]

Less credible than Hillary's actual "Silent Majority".

[Mr. Morden]

The "shy X voter" theory always pops up as a reason why we should believe robo-polls and internet polls more than live interview polls.  The argument is that people might lie to a live person on the phone, but not to a voice recording.  However, I haven't seen any actual evidence that line interview phone polls are less accurate.  If anything, the evidence suggests that they're *more* accurate.  So why should we stop believing them now?

I think there will be a point of 2 of Shy Trump voters, but there will also be a point or two of Shy non-native English speakers for Clinton.

Do you really mean "shy" non-native English speakers in the sense of not wanting to admit who they vote for, a la the "shy Trump voter" theory?  Or are you talking about voters who simply aren't reached by English-only polls in the first place?

[King]

Will be cancelled out by Apathetic Trump Voters: Republicans who hate Hillary but might not muster up the energy to go vote.

[Averroës Nix]

I think yes, at least in terms of people lying or remaining silent toward their friends, coworkers, and families.

And I have to confess that I prefer things this way; I would need to reconsider my life choices if I were spending a significant amount of time around people who thought that I would be OK with then talking that way.

But I'm not convinced that it's making an important difference in polling.

[Classic Conservative]

It's probably mostly people who don't want to tell their family members, friends and acquientences and this goes for both sides.

[NOVA Green]

The "shy X voter" theory always pops up as a reason why we should believe robo-polls and internet polls more than live interview polls.  The argument is that people might lie to a live person on the phone, but not to a voice recording.  However, I haven't seen any actual evidence that line interview phone polls are less accurate.  If anything, the evidence suggests that they're *more* accurate.  So why should we stop believing them now?

I think there will be a point of 2 of Shy Trump voters, but there will also be a point or two of Shy non-native English speakers for Clinton.

Do you really mean "shy" non-native English speakers in the sense of not wanting to admit who they vote for, a la the "shy Trump voter" theory?  Or are you talking about voters who simply aren't reached by English-only polls in the first place?

Thanks Mr Morden---- glad that someone was able to present a more concise explanation, and I am 100% positive it got mentioned within the past month or so on 538.

I recall the subject of "shy voters" going back way into the late '80s/ early '90s in Poly-Sci literature usually involving latent support for candidates widely perceived as racists (And i think that is the case) not willing to discuss their voting preference with a live interview, but would happily express their voting patterns in an automated format.