During this national election season, hardly a day goes by without mention in the media of how candidates and political parties are faring in the public opinion polls. We live in a time when polls and pollsters dominate the political landscape. Many of us want to know who’s up, who’s down, who can move the country forward and who just doesn’t have the skills to solve national problems. We also want to know what policy issues are upper most in the minds of the voters and whether those issues blend in with our concerns.
Public opinion polls and the individuals to perform the complex task of determining the views and the preferences of Americans are in the business of prognostication, predicting the future of politics. But this process of prognostication is fraught with numerous pitfalls as trying to accurately find what Americans are thinking and who they support at any one time during a campaign is a difficult task. In recent years, pollsters have had numerous problems determining with a degree of certainty who the American voter supports and why. The problems with determining with a degree of certainty during the last two presidential elections has led to many Americans losing faith in polls and charges that the individual pollster is blinded by partisan concerns to show a particular candidate ahead in the polls when the reality is quite different.
In most public opinion polls, the pollster will develop a statistical program that seeks to categorize the potential voters according to a series of socio-economic-political characteristics that would provide an insight into the general voting public. Those characteristics are often age, sex, religious affiliation, income level, education, place of residency, party identification and employment. The next step, and the most difficult, is to contact by phone what the pollsters feel are a representative sampling of the American public in a specific time frame. But rather than contact thousands and thousands of Americans with these characteristics, the pollster will contact a sample grouping of 2,000–3,000 people and present them with a series of questions that not only determines their viewpoints but also identifies which candidate they prefer to vote for in the upcoming election. Usually, the poll will have a disclaimer that the accuracy is + or - a certain percentage, often 2–3%. The disclaimer is not an attempt to escape claims of error, but rather to alert the public to the difficulty of arriving at complete certainty.
The difficulty of arriving at certainty is compounded by factors such as what type of phone is being called — a landline or a cell phone, whether the respondent is being truthful in his/her responses, and whether the characteristics used to develop the poll are indeed reflective of the American public. In the 2016 and 2020 presidential elections, numerous pollsters were criticized for “undercounting” voter support for Donald Trump, especially his voter base in rural America. Although the polls were not off by a large margin, in an extremely close national race, the incorrect level of certainty can create a false sense of confidence. The campaign of Hillary Clinton in 2016 was influenced by a misreading of the polls that had her as the predicted winner, only to find out that Trump was the victor in key battleground states like Michigan. Since 2020, pollsters have refined their statistical models and developed what they believe is a more accurate count of the Trump support that will provide a clearer picture of the American voting pubic. Furthermore, one way that polling has changed in recent years is that organizations that conduct the surveys have moved to online questionnaires that reflect the reality of increased Internet use, especially among the millennial generation.
One of the more serious problems associated with polling and pollsters is that the field is crowded with an ever-growing number of organizations conducting polls. Besides the more recognizable polls such as Gallup, Harris, and 538, some new polling groups such as YouGov, Morning Consult and Ipsos have gained a level of notoriety. Also, colleges and universities are getting in the polling business as Emerson College, Marquette Law School, Quinnipiac College, Siena (partnered with the New York Times), Monmouth University (partnered with the Washington Post) and the University of Massachusetts-Amherst have developed a reputation for accuracy. The reason for the growth of polling organization is that candidates at the national, state and local level spend millions of dollars to hire pollsters who they believe will provide them with an accurate and up to date analysis of how their candidate is doing, what issues to accent and how to improve the image of the candidate to the voting public.
Between now and the national elections on Nov. 5, candidates, the media and the American people will be watching closely to see how the various races for public office will play out. Because all polls have a margin of error and polling has in the recent past been problematic, there will be understandable skepticism about who’s up, who’s down, who can move the country forward and who lacks the skills to solve critical national problems. The key question is how influential will the polling results be in how Americans cast their ballots? Should one candidate be ahead in the polls by a large margin it could convince voters not to participate, while a close polling result could bring out more voters to support their candidate. Then there is the question of how independent voters will cast their ballot. Since Independents are not wedded to either political party and now make up the largest voting bloc, their movement from one party candidate to another makes prognostication difficult. Whatever the impact of polling on the national elections, it is certain that will remain an integral part of the electoral process in our country. We Americans have this ingrained desire to find out how their candidate is doing, and political polling prognostication remains one of the best ways of determining the ups and downs of the individuals running for public office.