Opinion: Navigating the reliability of political polls

Political onlookers are inundated with headlines touting the latest polls: national polls, swing state polls, internal polls and everything but a ski pole. Which polls should voters trust, if any of them?
Cowley: America has seen some high-profile polling misses. Pollsters and papers both predicted Harry Truman would lose. He won with 303 electoral votes. Polls said former President Donald Trump would lose in 2016 (he won), President Joe Biden would win by a landslide in 2020 (he squeaked out a win) and predicted a red wave in 2022 (it never came). Survey research is a science with a healthy dash of subjectivity.
An accurate survey requires talking to the right mix of people who proportionally make up the exact population of those who will vote on or before Election Day. A registered voter differs from an active voter, and neither necessarily means someone is a likely voter. Some research firms define “likely voter” differently; they vary on how much to count newly registered voters among other discrepancies.
I remain skeptical of surveys, especially since we know Trump supporters are reluctant to respond, which could also impact down-ballot results.
Pignanelli: “Polls? Nah … they’re for strippers and cross-country skiers.” — Sarah Palin
Polls are like fruit. You have to consume them when ripe, or they get stale. Also, the more of them you eat, the greater the benefits.
Many Americans and Utahns grumble about polls. But they are a popular product on television, newspapers and social media. This is because we all want to know what our neighbors, friends and others think.
Dictators do not utilize public opinion surveys for apparent reasons. This underscores the fact that polling and research are a fundamental aspect of democracy. We want officials, candidates, government agencies and other public affairs entities to inquire about the mindset of citizens.
The question is not whether polls should be trusted but how they will be viewed. The responses are a snapshot of the cross-section of the population. Questions may be tilted in attempts to tip the balance in one direction, but even that information can be worthwhile.
More news agencies are utilizing organizations that aggregate different polls to compile a broader perspective.
Fruit is an elemental part of enhancing a body’s nutrition. Polls serve a similar function of increasing the quality of responses from our leaders and the development of policies.
A recent survey from Noble Predictive Insights states Gov. Spencer Cox commands 54% of the vote compared to Democratic gubernatorial candidate Rep. Brian King’s 26%, with 20% undecided. King released an internal poll from Public Policy Polling that paints a far different picture, with Cox at 35%, King at 24%, Republican write-in candidate Phil Lyman at 19% and Libertarian nominee Robert Latham at 3%. What accounts for the discrepancy, and which is more accurate?
Cowley: Internal polls usually have an optimism bias. Most polling firms are partisan and, for a myriad of reasons, often skew in favor of their party. Therefore, I don’t trust internal polls to predict who is winning, but they are helpful for message testing and ascertaining what demographics can be persuaded.
Campaigns rarely share bad news polls unless it shows them gaining ground or the race is tighter than previously reported. Sharing internal polls with the press is a way to get earned media and boost optimism with supporters, which is what I assume King’s campaign is doing.
Pignanelli: Internal polls are an important part of a campaign strategy, especially for message testing, determination of trends, etc. The questions for such surveys can be intentionally leading to assist the development of a communications plan. When results are publicly released, they are often dismissed as biased.
Some local political analysts determined that the questions used in King’s poll were not misleading. However, a 19% result for Lyman suggests respondents were weighted with those opposed to Cox. However, the survey suggests that Lyman’s potential traction will reduce the margin between Cox and King. Conversely, Cox’s favorability ratings remain high, which will benefit him as the ballots are returned.
Another Noble survey indicates that Trump holds a 16-point advantage (54% to 38%), with 8% undecided. Also, Utah voters have a stronger preference for a generic Republican candidate than Trump. What does this tell us about Utah voters?
Cowley: A generic party-coded candidate almost always outperforms a real one. It’s why I think Vice President Kamala Harris avoided the media for so long and remains light on policy and substance. She benefits from voters not knowing too much about her, making assumptions about what she may or may not do if she wins, and dodging the baggage of the Biden/Harris administration.
Several surveys, as well as previous election results, show Utah Republican voters have reservations about supporting Trump. That doesn’t seem to be changing before Nov. 5. Trump will still win Utah, just by a smaller margin than previous Republican presidential candidates.
Pignanelli: Trump will win Utah. The polls reaffirm the continuing “complicated relationship” between local voters and the former president. This explains the disciplined answers that major Republican candidates provide when responding to questions regarding their support of the GOP nominee.
Republican Renae Cowley is a political consultant, lobbyist, social media influencer and professional rodeo athlete. Email: [email protected]. Frank Pignanelli is a Salt Lake attorney, lobbyist and political adviser who served as a Democrat in the Utah state Legislature. Email: [email protected].