Listen to me! I'm Trying to Save You Some Unnecessary Stress!
Here's why you must stop looking at election polls right now!
Between now and election day, we are going to see a lot of poll numbers. I am yelling at you right now, DO NOT LOOK AT THEM! I’ve given you plenty of reasons why you should mute polls, please reread them if you’ve forgotten or are freaking out. And inform yourself about not just on how polls work, but how they are used politically.
The best election polls are those commissioned by candidates for internal use. These polls tend to be far more detailed and extensive than the ones we see from even the most respected polling and news organizations and for good reason. Polls commissioned by candidates are used to make campaign decisions. They have to be as exact as possible so that campaign strategists have the best data to use to make the best decisions.
A poll done by the New York Times or another media organization might be very good, but, because the poll is conducted for journalists to use and to inform the public, there isn’t the pressure for the data to conform as close to reality as possible. The Times might get embarrassed over a faulty poll, but no one is going to lose their job or ruin their rep and thus their career. I’m not saying that the Times or other media operations are sloppy, just that they have far less at stake than a campaign does.
One huge caveat: While internal campaign polls tend to have deeper and more exact information, no one outside the campaign is likely to see that data unless a campaign has a specific reason to leak or release their numbers. As Mission Local reports, when campaigns do share internal polling data, they are going to tell us about the positive stuff - how well their candidate is doing, how popular they are - because good numbers get positive news coverage, which pumps up supporters, draws undecided voters, and increases campaign contributions. Makes sense. No one gets jazzed about a failing campaign. No one wants to vote for a loser. No one is going to cut a check to someone who is going down hard.
Campaigns will also release poll data to change a media narrative. In San Francisco, we have five solid candidates for mayor. One of the candidates never had a chance, one who shouldn’t have a chance has been pimped as a major contender, and one who has a chance has been dismissed as an irritant. Ignoring the also-ran, the no chance contender is Mark Farrell, a Republican-backed right-wing Democrat. The irritant who is a contender is Aaron Peskin, the only progressive running in the race and the only contender not backed by billionaires.
For months, media polls have had Peskin running a distant fourth, behind incumbent mayor London Breed, Mark Farrell, and Levi-Straus heir, billionaire Daniel Lurie. All summer the contest has been framed as Breed vs Farrell, with Lurie being a dark horse and Peskin being out. This week the Peskin campaign released poll numbers that show Lurie and Peskin nearly tied, with the winner determined by ranked choice voting.
The Peskin campaign released the poll to change the narrative, which they have done, and they did it with good timing, just as we got our ballots. And though the poll was floated for political reasons, the numbers make sense. San Francisco is a city split between moderate-to-conservative Democrats and left/progressives. Every mayoral candidate except Peskin is running from the right. Farrell is probably too right-wing for the city. London Breed’s administration has been a failure and is deep in multiple corruption scandals. Breed’s public image is also off-putting, only Farrell has worse negative numbers than she does.
Lurie has never held political office and, although Breed and Farrell have tried to use that against him, not having the stink of City Hall on him is a draw. He also is the least harsh of the right-leaning candidates and is more liked by the public than disliked. He is the nice guy “outsider” who appeals to those who are sick of “professional politicians.” Plus, he is a billionaire so theoretically he can’t be “bought off” or corrupted.
Peskin is not well-liked but he has a reputation as an honest, hard-working, smart, no-bullshit elected official. No one doubts that he can do the job (which scares the billionaires). He has an established base among the city’s progressives and left. He is the anti-Breed, running on empowering the people and against the billionaire class. He is pro-rent control and is backing major measures against corruption. Regardless of who right-leaning San Francisco supports, Peskin has the progressive vote locked up.
Because the Peskin poll rings true, people are taking the data seriously. Again, that was the intention for releasing it. Know that there is nothing unusual, dirty or controversial about the Peskin poll release. What Peskin’s campaign did has been a standard campaign practice since modern polling entered politics. Campaigns also release negative internal polling data when it pertains to their opponents, hoping to ward off voters from a candidate and demoralize their supporters. All of this information is fed to reporters, as well as an army of social media activists and bots. Some of the reporters are straight-down-the-middle types, grinders who are looking for a new angle to cover the horse race. Others are very much partisans.
No one would be surprised to know that the Trump campaign probably shares positive Trump numbers and negative Harris data with Fox News and the small fries in right-wing media. I am certain that the Harris campaign has some folks at MSNBC and some left-leaning journalist on speed dial. The people receiving the information might report on it or they might not, but their perceptions will be altered by the information and so will their reporting. And all of this happens behind the scenes, with you not knowing whether a pundit has an inside source or is relying on the polling data we see in the paper.
Listen, everyone is trying to get an edge, especially when races tighten and the election is wrapping up. The use of disinformation and dirty tricks intensifies the attempts to gain advantage. Campaigns like Trump’s will do anything to win, voter manipulation being its tamest tactic. Cleaner campaigns such as Harris’ will try to control narratives and spin information. None of this should be taken cynically but as a matter of fact and a general part of marketing and advertising, not much different than what goes down in soda or burger “wars.”
Finally, another key factor in polling is what is being asked and how it is being asked. Last week, Peskin campaign consultant Jim Stearns explained a type of internal poll to Mission Local. This from the Local:
A political poll typically starts by asking a horse-race question at the beginning: Which candidates a voter is aware of, and how they would rank them.
Next is a series of questions, looking for positive and negative responses, as a way of testing which of a campaign’s messages are appealing or persuasive. “There’s a big difference between saying, ‘Do you think the city should provide needles to illegal drug dealers that are lying on our street addicted to fentanyl?’’” says Stearns. “Or, ‘Do you think we should have medically supervised, safe consumption sites so that people don’t overdose?’”
Chisti3
At the end, the poll asks the horse-race question again — and by then, the results have usually changed. It’s important to know which set of candidate rankings are being released. The ones at the beginning? Or the ones at the end? And what questions are being asked along the way?
So, unless a campaign shares their polling data and methodology, or the reader digs deep into a poll (if they can), we have no idea what questions are being asked, what order they are being asked, and what answer to what question is being reported. When presented these numbers, we are being asked to asked to trust the people feeding us this information.
Sorry, but if something comes from a campaign, I’m a skeptic. If it comes from the Trump campaign, I am an uber-skeptic. If it comes from a pro polling firm and/or news organization, I am doing some digging. If the information is circulating on social media, I’m ignoring it. And unless I know the exact details of a poll – where it came from, what was asked, how it was asked, who was polled, how they were polled, what was released, what wasn’t released, etc. – poll numbers mean nothing to me.
I advise you follow my example. Stop looking at poll numbers. Start getting people to the polls. Substitute obsession and worry with action. Not only is it much healthier to avoid the worry, elections are won by activism, not through concern, even deep concern. Votes are the only numbers that matter.