Ricochet is the best place on the internet to discuss the issues of the day, either through commenting on posts or writing your own for our active and dynamic community in a fully moderated environment. In addition, the Ricochet Audio Network offers over 50 original podcasts with new episodes released every day.
The Scarlet Letter
In 1982, Los Angeles Mayor Tom Bradley secured the Democratic nomination for Governor of California, a first for an African-American politician since Reconstruction. The polls showed him easily defeating the Republican George Deukmejian and the early post-election editions of The San Francisco Chronicle carried the headline “Bradley Win Projected.” But in the end, Deukmejian won by less than 94,000 votes (in an election that saw almost 210,000 votes go to third-party candidates.)
When the question arose as to how the pollsters screwed it up, the theory of “The Bradley Effect” was born. In a nutshell, white voters were supposedly “ashamed” to admit that they were supporting a white candidate over a black candidate and lied to the polling companies, especially if the individuals asking the questions were black. It is a theory that has been both pushed and dismissed without proof one way or the other for almost 40 years now.
In the last three years we’ve seen three massive failures of public opinion polling worldwide: the 2016 US presidential election, the UK referendum on EU membership and, most recently, the Australian Parliamentary elections. Of those three, Australia is the most interesting. Down Under, the Aussies have mandatory voting. If you don’t vote they will seek you out afterward for a reason. If the authorities aren’t satisfied with your answer there’s a modest fine. ($20AUS for the 1st offense, $50 for repeat “offenders.”) Since its institution in 1924 turnout has never been lower than 91%. Modeling the electorate or predicting turnout is not their problem. The only logical conclusion is that respondents are deliberately misleading them.
The idea that someone in 1982 either felt shame or feared retribution is laughable. That can no longer be said today. Give the wrong opinion now and within minutes your picture could be on social media with pleas for crowdsourcing your identity, with the ultimate purpose to make your life a living hell with everything from threats to your life, your property and your livelihood.
Of course, none of the legitimate pollsters would put up with any of their employees doxxing survey participants. But the thought only has to exist in the minds of the public to put a chill upon the whole process.
Make no mistake about it: de-platforming, de-monetizing, doxxing and shaming on social media is the new Puritanism. Like the heroine of Nathanial Hawthorne’s story, we’re all to be branded with a new scarlet letter, be it “C” for Conservative or “T” for Trumpist. That’s going to push the gap between public pronouncements and the private actions of the voting booth farther and farther apart.
Published in General
One way to get the actual vote totals to match the polls is to eliminate the secret ballot. The Democrats have been trending that way for years. Remember “card check”? Union certification ballots would be public, so the union could intimidate the workers into voting the right way.
Much worse, people usually fail to recognize that an absentee ballot is not a secret ballot. (“What? You want to vote in person instead of letting us fill out the ballot and send it in? What are you trying to hide?”) The same applies to internet voting.
Not surprisingly, Democrats and RINOs are promoting both. In olden days, the Dems would pay a wino five bucks to vote, but they could never be sure what he did, alone in the voting booth.
I thought I had seen the last of pollsters when we got rid of our landline. And for a long time things were good. Then just the other night I got a call on my cell and I’ll be darned if it wasn’t a pollster. Unfortunately, with my summer job I need to answer the phone and I can’t screen calls like I do the rest of the year (not on my contact list, I don’t answer. No voice mail left, it wasn’t important.)
This has not been my experience.
Another problem with polls of “likely” voters. If you found 1000 Democrats who told a pollster they were “very likely” to vote, and 1000 Republicans who claimed they were “very likely” to vote, do you think the actual result would be similar or not? I think Republicans would be somewhat more reliable to carry out their claims ( as long as they weren’t Republican politicians).
Pollster: Sir, what is your race?
Stad: The Kentucky Derby. It’s a classic!
There is also the matter of the model. I lost faith in Fox News polls in 2016 when I saw an interview with a guy from the ‘brain room’ condescendingly talking about polling, that if you didn’t have a model like 2012 your numbers would be off. When I could think of several ways it was likely to diverge seriously from 2012, and I don’t do polling.
Absolutely. It’s amazing how little emphasis is placed on the model which is crucial to accuracy. And that’s probably because it’s largely a guess, or in defrence to ‘science’, based on former models that may well have lost basis or relevance. It starts out as either opinion or 4 year-old data. And they never talk about it. There’s fake news, but let’s not forget, fake weather, and fake polls!
In fairness to all political parties, it was commonplace in the 18th century to “bribe” voters with whiskey.
There actually were Republican big city political machines, but not for the last 100 years or so.
The questions were always, to me, what caused polls to go south. I’ve done the calls and even written a few. It’s a tricky thing to write a polling question that isn’t leading in some way.
Several here say they lie to pollsters and otherwise mess with them. The pollsters are aware of this. Others do it unintentionally: whatever answers one perceives as “right” are chosen, as opposed to what the person called actually thinks. Something with basic psychology, I guess. They try to please the person on the phone. In other words, if the question leads in any way, you’re looking at a skewed result.
Taking the two phenomena together, you’re looking at a growing margin of error. But…it’s big business!
I heard the basic problem with both Euro and Australian voter polling is the standard practice of omitting those who say they are undecided or decline to answer. In both cases, I understand there was a large enough such group to change the election outcome.
Also they pressure undecided voters into picking a candidate.
The canvassers have a quota to meet. Human nature being what it is, I’m sure a significant part of the responses they record are made up out of whole cloth. Whatever sounds plausible to them.
Here was my last interaction with a pollster, from September 2018:
After the usual questions about my age and where I lived, she asked:
“Do you consider yourself White, African-American, Native American, Asian or Pacific Islander, or some other race?”
“No.”
“I’m sorry, sir, you have to chose one.“
“I don’t consider myself any of those things.”
“I need to put a race down.“
“Okay, Human.”
“I can’t put that down.”
“I guarantee that I’m not a fish in disguise.”
“It needs to be one of the choices: White, African-American, Native American, Asian or Pacific Islander, or some other race.”
“Race is an artificial construct historically used to separate and divide people. By its very nature, race is a racist concept. I am not a racist, therefore I don’t use racist labels for myself or others.”
“Okay sir, thank you. That is all the questions I have for you. Have a blessed day.” [Hangs up]
Conclusion: Only racists are allowed to answer polls.
Very good!