One more time with feeling: Ignore the polls!

September 4, 2023

          Sigh. Fourteen months out from the 2024 presidential election and the public is being subjected to the “every-other-week” poll that shows the race between Biden and Trump is “too close to call.” We will be treated to this form of disinformation and mental abuse as long as major media outlets and respected commentators treat the polls as if they are meaningful and predictive. They are neither. Instead, they are clickbait wrapped in statistics that misleads by confusing precision and truth. If someone tells you that the universe will end in 3,198,642,971.25 years, that is a “precise” prediction. Whether the prediction is “true” is a different question entirely. So, too, with the polling.

          Over the weekend, Rupert Murdoch’s Wall Street Journal published a poll that allegedly shows that Biden and Trump are tied in the 2024 presidential race. While the poll itself has no predictive significance, it is being treated with credulity and alarm by political commentators who should know better. Although I think it is unproductive to talk about polls fourteen months before elections, the WSJ poll lit up my inbox on Monday morning with notes from worried readers.

          For those of you tired of reading my response to such polls, I apologize for the repetition. You may want to set aside this newsletter and start afresh with tomorrow’s newsletter. To those of you who need reassurance, read on! Because we will see many similar polls over the next fourteen months, I will use the WSJ poll as an example of how pollsters can distort the truth and why we should generally ignore the polls.

          Let’s start with the most pernicious aspect of polling: Many (most?) polls are conducted not to survey public opinion but to influence it. Republican pollsters “flood the zone” with underpowered, poorly designed surveys to skew the results of polling aggregators like FiveThirtyEight.com. Although polling aggregators are aware of such attempts to manipulate polling averages, aggregators nonetheless include sham polls while claiming that they “discount” those polls to account for their obvious bias.

          The countermeasure of discounting poorly designed polls didn’t work in 2022, 2020, or 2018—at least in the sense that major media reported the low-quality outlier polls with the same seriousness as higher-quality, well-designed polls. “Mission accomplished” by Republican operatives, who created a misleading media narrative by ginning up a bunch of low-quality polls.

          But, most importantly, “horse-race” polling in the Trump era is a joke. Sticking to the old paradigm of “which candidate will you vote for” ignores the fact that no one who votes in 2024 will be deciding between Biden and Trump. They will be deciding between democracy and autocracy, moderation and extremism, equality and white nationalism, reproductive liberty and religious intolerance, dignity and bigotry, climate protection and fossil fuels, the Constitution and insurrection, and order and chaos. Biden and Trump are surrogates for all of that—and more. Any pollster who fails to capture voter motivations relating to those issues in surveying voter preferences about presidential candidates is living in a pre-January 6th world that no longer exists.

          Any pollster who does not adjust their models for the post-Dobbs victories for reproductive liberty in Kansas, Ohio, Michigan, Minnesota, and Colorado and the losses of anti-choice candidates Herschel Walker, Blake Masters, Kari Lake, Dr. Oz, Doug Mastriano, and Daniel Kelly deserve the embarrassment they will experience when their polls get it wrong in 2024.

          Moreover, survey design can “prime” respondents to express negative views of one candidate by asking questions that precondition the respondent. And survey design always works against the incumbent, who is held responsible for whatever challenges the nation faces at the moment. For example, consider the following two survey designs:

Survey Design A: Do you think Joe Biden is too old to be re-elected? Do you believe Joe Biden is responsible for inflation? Do you believe Joe Biden’s emphasis on green energy is causing gasoline prices to increase? As between Joe Biden and Donald Trump, which candidate do you prefer in the 2024 presidential election?

Survey Design B: Do you believe the verdict finding that Donald Trump sexually abused E. Jean Carroll is relevant to character? Do you believe Donald Trump should have told the American people he paid his mistress (an adult film performer) to keep quiet before the 2016 election? Do you believe that Donald Trump acted quickly enough to protect Congress during the insurrection on January 6th? Do you believe that it is appropriate for Donald Trump to promise that he will “lock up” his opponents if reelected? As between Joe Biden and Donald Trump, which candidate do you prefer in the 2024 presidential election?

          In case you think I am exaggerating the use of “priming” in presidential surveys, here is the WSJ’s description of its own polling:

Voters overwhelmingly think President Biden is too old to run for re-election and give him low marks for handling the economy and other issues important to their vote, according to a new Wall Street Journal poll that offers a stark warning to the 80-year-old incumbent ahead of the 2024 contest. 

The negative views of Biden’s age and performance in office help explain why only 39% of voters hold a favorable view of the president. In a separate question, some 42% said they approve of how he is handling his job, well below the 57% who disapprove.

          So, before asking which candidate voters prefer, the WSJ (and other poll organizations) ask extensive questions about the economy, border security, the war in Ukraine, unemployment, etc. Because Trump is not the incumbent president, none of those “priming” questions were—or could have been—asked about him.

          Finally, while I have no training in polling, the WSJ’s description of its survey methods and pool of respondents seems suspicious. The WSJ says it conducted “telephone and text-to-web surveys of 1,500 registered voters,” but the results reflect responses by only 750 voters—exactly 50% of those surveyed. That round number is too tidy to be the result of chance. How is it that precisely half of the respondents ended up being included in the survey?

          If that 50% represents the percentage of those who self-selected to answer a web-based survey or telephone call from the WSJ, what does that self-selection say about those who volunteered to answer versus those who ignored a survey request from the WSJ?

          I will stop here. You get the idea.

          To be clear, we should not dismiss or reject all bad news. Neither should we be gullible marks for Rupert Murdoch, who does not want to see Joe Biden reelected. Still, there is information in the WSJ polling that Democrats should use to refine their message. For example, the survey respondents held the following beliefs:

But 58% of voters say the economy has gotten worse over the past two years, whereas only 28% say it has gotten better, and nearly three in four say inflation is headed in the wrong direction.

          Accepting as true that the above results reflect the responses in the survey, those responses reflect views that flatly contradict the facts. The economy has improved dramatically over the last two years and inflation has been decreasing for the last sixteen months. That belief gap is a problem for Democrats—one that is exacerbated by the irresistible urge of the media to create a negative narrative—regardless of the facts.

          The incessant negative narrative is itself exacerbated by false equivalencies that are casually accepted and promoted by the media. For example, the WSJ compares the relative “honesty” of Biden and Trump. As to Biden, the only “honesty” factors relate to President Biden’s alleged involvement in Hunter Biden’s business dealings, which the WSJ says are non-existent. As to Trump, “honesty” presumably covers his promotion of the Big Lie, his attempted coup, his refusal to return defense secrets, his obstruction of justice, and his lies about his sexual assault and affairs.

          There is simply no equivalency in the “honesty” of the two candidates, but the WSJ survey forces a false equivalency by compressing the above issues into a single comparison of the candidates’ relative “honesty.” Every major news outlet is guilty of the same false equivalency every time it counters Trump's manifest corruption with a mention of “Hunter Biden’s business dealings” without adding that NO ONE has found any evidence that Hunter’s business activities were illegal or that Joe Biden was involved. Simply repeating the phrase “Hunter Biden’s business dealings” without more to create the illusion of fairness and balance is journalistic malpractice.

          I have wandered far afield from my original objective—to explain why Democrats should not overreact to the WSJ poll—or to any of the other “made for TV” polls that will wash over us with regularity between now and November 2024. We are the intended targets of those polls, and we must fight the pollsters’ efforts to manipulate us.

          The most important lesson can be illustrated by a thought experiment. Assume three different scenarios: That polls consistently show Biden in a statistical tie with Trump. Or that Biden consistently trails Trump by 5%. Or that Biden consistently leads Trump by 5%. What would you do differently in each of those scenarios?

I hope the answer is “Nothing.” In the lead-up to 2024, we can take nothing for granted, leave nothing on the table, and spare no effort. We have learned from bitter experience that Trump will enlist our foreign adversaries and encourage violence to gain the presidency. We should assume he will do the same in 2024. A 5% lead by Biden guarantees nothing. We need to beat Trump by tens of millions of votes.

Concluding Thoughts.

          Getting Biden’s message out has proved difficult. Biden has been on the road for much of the summer, touting his economic achievements. He was campaigning over Labor Day weekend, hitting Trump hard on the economy. See The Hill, Biden bashes Trump’s jobs record to mark Labor Day. Although The Hill reported on Biden’s speech, none of the major newspapers or cable channels did so.

          The refusal of major media to cover Biden’s speeches is frustrating. It is not enough to complain about the unfair treatment of Biden compared to Trump (for whom every encounter with reporters resulted in wall-to-wall coverage). We must be part of the solution. By writing, posting, calling, texting, and canvassing (or donating to those who do), we can spread the word ourselves. That is hard work and expensive, but we have no choice. We must break through the information gap; we must ensure that Biden wins.

          The good news is that the facts are on our side. We need only communicate them—unlike our opponents, who must deny them in the face of incontestable evidence. As between the two positions, a rational person would want to be in our shoes every day of the week. Let’s take that happy thought and get back to the hard work of redeeming democracy for our generation!

Talk to you tomorrow!