Pole Vaulting Through the Polling Media

Christine Merser, We the People Dispatch

Here we go. Kamala Harris' polling numbers will start rolling in any second, and our legs aren’t long enough to vault through it. More importantly, it’s going to get major news coverage. Do not give it more importance than it deserves.

In my mind, not to be confused with fact, we still consider polls accurate, as a way to see what will happen in the future, and that’s our first mistake. We love being told who is going to win a month before a ballot is cast. It makes us feel better (or worse, but I’m manifesting, so going with better). Gallup is in our DNA (or those of us who remember Gallup polls), and we believe Gallup. He was Oz behind the curtain.

But times they have changed. This election has the potential to change in a second. It did after the debate. It did when Biden stepped down last Sunday. Note to self?

How about we don’t take them seriously… yet?

How about we do not talk about them with fear and loathing? (How can women possibly be voting for the men that will put a Handmaid’s Tale cloak on us; one that looks not surprisingly like the Scarlet Letter’s Hester Prynne?) Polls make us crazy people.

How about we totally ignore national polls around the Prez election when our reality is that the popular vote won’t count? Ask Hillary. She won it but never got to pick her desk for the Oval Office.

A Little History

Political polling has a storied history, dating back to the 1930s with George Gallup, who pioneered scientific polling methods that aimed to accurately reflect the public's views. Over the years, polling techniques became increasingly sophisticated, employing random sampling, quota sampling, and various statistical methods to achieve high levels of accuracy. During much of the 20th century and into the early 21st century, polling was considered a reliable tool for predicting election outcomes and understanding public opinion on key issues.

Polls Today

Then everything changed. But we still treat them with the respect they no longer deserve. Why, you ask? If you want to dive deep, the info behind the bullets below, is at the end of this post.

• Decline of Landline Use (see the end of this post for additional info)

• Rise of Online Polling (see the end of this post for additional info)

• Polling Aggregators and Data Models (see the end of this post for additional info)

• Increased Polarization (see the end of this post for additional info)

• Methodological Innovations (see the end of this post for additional info)

And, then there are some additional concerns I have, but these are not based on science. Just my thoughts.

• I think people lie. I think the woman sitting on the couch doesn’t want her husband to know how she is voting. I think some of my friends don’t want me to know how they are voting. Can you imagine? Sheesh.

• I think people are afraid to speak up. Repercussions abound. Anonymous no longer exists.

• I think we are a fast-moving society. I might intend to vote tomorrow but start scrolling at 8 am and do not get to the polls.

Given these changes and challenges to gathering accurate results, you might ask why we have them anymore. If they are not accurate, why have them? Because we like them? Again, this is where the media doesn’t want to lose that hook, and they do tell us things, just not for sure.

What To Do

Don’t Share Polling Data: Don’t pass on polls like they are a done deal. Do not give them the oxygen to make people complacent about showing up to vote, or feel like what’s the point.

Look at Polling Averages: Individual polls can vary widely, but averages of multiple polls tend to be more reliable. Aggregators like FiveThirtyEight, RealClearPolitics, and others provide useful averages that smooth out anomalies.

Consider Methodology: Pay attention to how a poll was conducted. Was it online, via phone, or a combination? What was the sample size, and how was it selected? Understanding the methodology helps gauge the poll's reliability.

Check the Margin of Error: It’s like looking under the hood to see if the car engine looks right. The margin of error indicates the range within which the true value likely falls. A small margin of error generally means more confidence in the results, while a larger margin suggests more uncertainty. It might be in the small print. If you don’t see it, ask about it in the comments section. Or ignore it.

Evaluate Pollster Track Records: My recommendation is Steve Kornacki at MSNBC. Great track record and explains all these things I’ve listed in a much more engaging way than I have.

Account for Trends and Movements: Rather than focusing on a single poll, observe trends over time. Consistent movement in one direction across multiple polls can be more indicative of public sentiment than a one-off result.

Be Wary of Outliers: Outlier polls that significantly deviate from others should be viewed with caution. These may result from methodological issues or sample anomalies.

Contextualize Poll Results: Polls are snapshots of a particular moment in time and can change rapidly, especially during a dynamic election cycle. Consider the broader context, including recent events, economic conditions, and major campaign developments.

It’s a lot. I get it, but this is so important to our peace of mind and approach to sharing information and managing it. If you are not willing to do the homework, maybe you should skip reading the polls this go-around. Another binge of The West Wing instead perhaps?

Christine Merser, We the People Dispatch

---

Back Up Information:

Decline of Landline Use: Ten years ago, a significant portion of polling was conducted via landline telephones. Enter cell phones, and this forced pollsters to adapt. Reaching a representative sample of the population has become more challenging as cell phone numbers are less accessible, and younger demographics are more likely to be cell-only users.

Rise of Online Polling: As if the decline of landlines wasn’t enough to contend with, many pollsters have turned to online surveys. While online polling can reach a broader audience, it also adds some serious challenges such as ensuring a representative sample and dealing with self-selection bias, where individuals who choose to participate in online polls may differ significantly from the general population.

Polling Aggregators and Data Models: The rise of polling aggregators like FiveThirtyEight and RealClearPolitics has transformed how polls are analyzed. These platforms compile and average multiple polls, using sophisticated models to predict outcomes. While this can provide a more comprehensive picture, it also introduces new complexities and potential sources of error. Big errors.

Increased Polarization: The past decade has seen heightened political polarization, making it more difficult to obtain accurate responses. People are more likely to distrust pollsters or give socially desirable answers rather than their true opinions, known as the "shy voter" phenomenon.

Methodological Innovations: Pollsters have increasingly adopted new methodologies, including weighting responses based on demographic factors, using advanced statistical techniques to correct for biases, and employing experimental designs to test different approaches. These innovations aim to improve accuracy but also add layers of complexity to interpreting results.