I think it’s pretty unlikely that a committed Republican is going to lie by saying they’re voting for Biden. Either they hang up without answering, or they say they’re undecided.
He’s also looked at the polling data, for '16, '18, and now, and has seen no data suggesting their significance.
"… To test this theory, CloudResearch recently sampled American voters in search of what they term “shy voters.” Their results show that Trump supporters were “significantly more reluctant to share their opinions on phone surveys compared to Biden supporters.” Almost 12 percent of Republicans and nearly 11 percent of Independents, were also almost twice as likely to be reticent than Democrats (about 5 percent).
These seemingly small percentages could have major November implications…"
Interesting … they seem to be the only ones who’ve found firm evidence so far. (The linked article is biased; here’s a link to the source.)
It is important to notice the limitations that the source reports:
Limitations
As with all studies, there are several possible limitations to the conclusions one can draw from this study as well. First, we used a nonprobability sampling frame, which means the results of this study may not generalize to the whole population. This remains true despite the fact that we replicated the main findings in this report across two samples that were recruited with quotas set to match the demographics of the U.S. population (Study 1) and quotas set to match the demographics of voters within each political party (Study 2). Additionally, it is important to remember that this is self-report data, which may not perfectly align with people’s real world behavior.
As a counter I do remember that another pollster friendly to conservatives (Rasmussen) came with a poll the day before the 2018 midterms that showed that the republicans were a point up over democrats and it could result with the republicans keeping the house. It was not to be.
These may be subsumed in the group of white non-college educated whites who are less likely to vote so are already given higher weighting by pollsters when they make their predictions. I am polled fairly regularly, as I am a registered voter in a swing state. They always ask about education because of data that shows that college educated whites are more likely to respond to polls. I’m sure there’s considerable overlap between these “Trump supporters significantly more reluctant st share their opinions” and the non-college educated white voters that are already given extra weight in the predicted outcomes because of known response rates.
This should have read: These may be subsumed in the group of white non-college educated whites who are less likely to reply
Right, the number that matters isn’t so much “reluctance”, since that can be adequately handled with demographic sampling (as long as ever relevant factor, like education level, is included). It’s “inaccuracy”. It’s one thing to say “I wouldn’t answer the question” and it’s another to say “I would give the wrong answer”.
I do think there was one sign of that in 2016, which was pretty high numbers of undecided and third-party voters that ultimately became Trump voters. Whether they were Trump voters all along and just shy about it or truly decided late is unknown.
There just aren’t very many undecideds or third-party voters this cycle. So it would have to be liars.
Even if there are shy Trump voters - and I have no doubt that there are - there’d have to be a colossal number of them to swing the outcome of November’s election. Biden has a massive lead by this point.
I’m not sure how much we can rely on 2018 to confirm they’ve fixed the polling. Trump wasn’t running then so if there were any shy voters it would have been minimal.
I’m still nervous about this and I mentally add 3% to all of Trump’s polling.
Over at 270towin.com they’ve got a listing of various pundit forecasts; one of them (Princeton Election Consortium) has a range of three maps - one based on current polling, one based on Biden outperforming polls by 3%, and one based on Trump outperforming polls by 3%. You can find them here. You might find it useful.
As of right now (ie, if the election were held now), if Trump outperforms his current polling by 3%, it would be 319 Biden - 219 Trump.
On the main page, there’s a map based on the consensus of ten organizations to play with. I’ll look through it, compare it to current polling from time to time. Anything above the margin for error (5%) for Biden I give to Biden, while anything less goes to Trump. I’ve been getting numbers in the 290+ for Biden that way.
From an recent Economist article:
The president’s advisers suggest the polls are missing many “shy Trumpers”. But the alacrity with which Trump supporters defend his chauvinism and sometimes display their own makes that seem unlikely. It perhaps also explains why, notwithstanding his lock on his base, Mr Trump is struggling to expand it.
I don’t know if this has been said, but if you are afraid to say who you are voting for maybe your choice, well, sucks.
More and more I think the constant “shy trump voters” we hear about is just to placate his base and is “code” among the rwnj’s that the fix is in, don’t worry about the polls.
There are legit reasons for someone to not want to say who they are voting for. A Biden supporter in rural red America may not want to. A black voter surrounded by white supremacists may not want to (although, in his example, his skin color would probably tell everyone his vote anyway)
trump voters do lie to pollsters. From shame, paranoia, or perversity. Shame is self-explanatory, paranoia has been covered, and as for perversity, well if they all told the truth the polls would be more accurate and then they wouldn’t be able to complain about how inaccurate the polls are.
also, a Trump voter who works at
Goodyear Tires.
But if they all told the truth, the polls would also show better results for Trump, in which case they wouldn’t actually have anything to complain about.
I’m pretty sure this is less of a “thing Trump supporters really do” than a “thing Trump supporters say that other unnamed Trump supporters are doing in order to convince people that he has a better chance of winning than he really does.” Generally, the perception that a candidate is doing well HELPS that candidate, since people like to vote for candidates they perceive as popular. Trump supporters who SAY he’s doing better than the polls indicate clearly know that this is the case, or they wouldn’t be saying it. And yet, they expect people to believe there’s a significant number of other Trump supporters who lie to pollsters in order to make Biden appear more popular than he is? Not buying it.
(Actually, if anybody were to do this, I would expect it to be Democrats, who actually do have a tendency to believe that it is dangerous for their candidate to be perceived as the frontrunner, because “complacency” or something. But as far as I can tell, this belief is confined to politically active Democrats, and it’s a reaction to the trauma of losing several very close presidential elections. Republicans, who WON those elections, do not share it.)
Eh, I’m not sure if they know it, vs want to vote for Trump, and feel (perhaps unconsciously) that they want to vote for someone who is popular. So, via motivated reasoning, they cling onto the idea that their chosen candidate really is popular, just secretly so. It makes sense in a way that it would be a popular piece of motivated reasoning, because it connects well with the general “fake news,” anti-establishment, and white grievance themes connected with Trump.
You’re applying logical thinking to folks who don’t think logically.