top of page
Search
Jeffrey A. Rendall

The Right Resistance: What do political pollsters and a stopped clock have in common?

It happens every election (not just every four years for the presidential contest): Americans of every ideological predilection diligently watch for fresh poll results from the media, with their

level of enthusiasm rising or falling according to the figures.


Good poll numbers are akin to a sugar high, a temporary fix that raises eyebrows and causes hearts to race, only to subside as quickly as the phenomenon arose after reality sets in. Everyone wants to know how their “horse” is doing in the proverbial political “horserace” and sometimes their daily lives are haunted by the mystery.


The strange and peculiar -- and often erroneous -- nature of polling was especially exposed in the past couple election cycles, largely due to the presence of outsider first-timer Donald J. Trump in politics. Trump broke the paradigm on how Americans viewed the quadrennial extravaganza, bringing his own brand of in-your-face campaigning to a system that hadn’t seen much modification in decades. Understandably, Trump’s poll numbers have always been all over the place, first concerning his standing in the Republican Party’s 2016 primary race and then his head-to-head matchup with Hillary Clinton.


Of course, the polling industry’s 2020 record did not improve from its abysmal 2016 effort, consistently showing then-candidate Joe Biden with a sizeable and steady lead despite the Democrat doing little or no campaigning for the office. It’s safe to say few were surprised when Election Day came around and the actual vote counts revealed Trump’s ranking was significantly greater than the surveys had forecasted.


Simply put, the pollsters blew it again. Now, there’s proof. Byron York touched on the subject at The Washington Examiner:


“[T]he polls were not only wrong but more wrong than they had been in decades, including their dismal performance in 2016. A new report from the American Association for Public Opinion Research found that national election polls overstated Biden's lead over Trump by 3.9 percentage points. And state-level polls were even worse, overstating Biden's lead by 4.3 points.


“The polls’ performance in statewide races -- contests for senators and governors -- was even worse. There, they overstated the Democratic candidates’ lead by an average of six percentage points.


“Notice something? The errors went in one direction: Overstating support for Democratic candidates. ‘Whether the candidates were running for president, senator, or governor, poll margins overall suggested that Democratic candidates would do better and Republican candidates would do worse relative to the final certified vote,’ the report says. So all those Republicans who said the polls were biased against Trump and the GOP -- they were right.”


Bingo! Pfft. Conservatives and Republicans realized this long ago, and it didn’t take a fancy-schmancy white paper from some stuffy research outfit to confirm our suspicions. York’s piece only proved what most of us were sensing in our souls, that report after report of Trump and the Republican party being electoral dead ducks ahead of Election Day last year were exaggerated at best and dead wrong (with a pound of gusto) at worst.


Just talking with neighbors and friends was enough to infer that something was happening at the human level.


Remember? Whenever the subject of the election would come up in casual conversation with strangers, ardent Trump backers would lower their voices, look around to see who was within earshot and then let their opinions flow once the coast was clear. I recall being heartened when a female employee in a supermarket checkout line admired my son’s “Trump 2020” mask, only to explain she had to watch what she said because the thought police (or her bosses) might detect her underlying Trump support.


Her closing thought was, “If he loses this election, it’s all over for us. I pray every night.”


Unfortunately the woman’s prayers weren’t answered, or just not in the way she’d sent her requests heavenward to beg for the almighty’s assistance. With the emerging evidence of fraud and/or downright incompetence in heavy Democrat districts in Arizona and Georgia (among others), conservatives were proven prescient on their post-election hunches as well. Yet it also seems clear that liberals love polls but won’t stoop to accept an audit’s findings.


Why would anyone believe a poll but not trust an audit? It doesn’t make sense. A poll is conducted by someone paid to survey public opinion, an inexact science to say the least. I’m not a credentialed expert on the practice, but my understanding is that a polling entity takes a sample of registered voters or, from a more detailed calculation, likely voters, based on a list compiled through certain sources. If you want a valid representation of conservatives’ opinions, for example, a good place to start would be the Heritage Foundation’s membership roster. (Note: a Republican Party list would not be representative of the grassroots’ feelings, in other words.)


I used to work for a firm that took surveys based on various phone lists, but the results weren’t necessarily scientific or political. Even back in those days, most of the battle involved trying to coax or compel someone to participate. It felt kind of like hooking a fish and then bringing it to the boat, only to see it spit out the barb and swim away. It didn’t always work and the sense of loss from “losing one” was tangible.


In contrast, an audit deals with actual source materials. In a lawsuit, they would be called “evidence”. Tax audits compare W2s, receipts, etc. vs. the numbers on the tax return. It’s an arduous process, both for the subject of the audit and the auditor him or herself. No one’s happy. But one way or another, the audit produces a result. A poll doesn’t, since no poll samples everyone in a voting precinct. The actual survey responses are filtered through a formula of projected strength for a candidate.


Polls are educated guesses based on historic trends, demographic data, migration patterns and your basic dartboard throw predictions. Different firms use divergent methods. Kind of like following a ball on the roulette wheel, occasionally they hit the number spot on. But ask them to do it two times in a row and the challenge gets tricky. When bias enters the picture, you’ve got an even foggier outlook.


2020 was a joke. Here’s my take on what went wrong. I don’t pretend to be Karl Rove holding a white clipboard with figures scribbled in blue and trying to look smart and ahead of the curve by addressing each bullet point with a treatise on why this or that happened in location X. The answer is much easier to come by:


One, the polling firms don’t ask the right questions. Instead of asking what candidate or party the subject preferred and his or her likelihood to vote, why not sample why voters intend to vote? You won’t get tallies for certain candidates, but by identifying what’s motivating an action, there’s a better predictive value. “I’m voting because I’m ticked off at the government’s response to COVID-19”. Ding ding ding! Biden and Democrat voter!


“I’m voting because I’m disgusted at all the rioting that happened last summer.” Trump and Republican voter!


Pollsters already do this, of course, but it seems like they try to make connections that aren’t there. Perhaps it’s because media entities are attempting to bolster their left-leaning point of view. The COVID scaremongering from last year was way over the top and the talkers connected it to Donald Trump’s attitude and policies. Trump’s voters weren’t as vocal, and thus harder to detect.


Two, in 2016’s and 2020’s cases, pollsters didn’t account for the number of new voters Trump brought into the game. Again, it was informal intelligence gathering, but I talked with a lot of people who indicated they or members of their family had never voted before but were doing so in 2020 because of Trump. These people don’t show up on the political radar and probably wouldn’t be counted as either likely or even registered voters. There’s no way to calculate the actual number, but they were heard loud and clear on Election Day.


Lastly, distrust of the media is at an all-time high. The silent Trump voters and uncounted Republican voters wouldn’t accept the liberal experts’ interpretation of the polling data. Where was the sample taken from? Did it account for changing attitudes in minority communities? How do you poll the student population? What role would mail-in ballots play? Fraud? Mark Zuckerberg?


There were just too many intangibles in 2020 to believe the polling from liberal media entities. Like in 2016, the “professionals” ended up with egg on their faces again. Conservatives aren’t any more willing to trust the current marks on Joe Biden’s popularity or the very early predictions of 2022 support. Even a stopped clock is correct twice a day.


  • 2022 elections

  • 2022 polls

  • 2020 polls wrong

  • Donald Trump

  • Arizona audit

  • Georgia audit

  • Electoral college

  • January 6

  • Silent Trump voters

  • Joe Biden

67 views2 comments

2 Comments


kenmarx
Jul 23, 2021

Consider who the primary sponsors of polls are: NYT, MSNBC, WP, CNN, Fox, etc. All of them, including Fox, unfortunately, go at polling from a leftist viewpoint. I've been watching political poll results since the 1990's and have come to the conclusion that if the Republican is within five points of the Democrat, he will win. I think by skewing the poll results to favor the Democrat, the media are attempting to cash in on the "everybody's doing it concept."

Like

Van Snyder
Van Snyder
Jul 22, 2021

A clock that loses one minute per day is right every 1440 days.

Like
bottom of page