Polls Aren’t Just Wrong, They Propagate Racism and Sexism

COMMENTARY

Based on “objective” exit poll data, President-elect Donald Trump won the 2024 presidential election because of “White men,” “Catholics” and those with “no college degree.” Because it’s “data,” that can’t be racist or sexist, can it? Yes, it can.

Data can be manipulated to tell any story you want. And poll data can unwittingly deepen the divisions in our nation. 

As a previous data professional at the U.S. Department of Transportation, Humana, Fannie Mae and Thermo Fisher Scientific, and now CEO of The Hidden Speaker, I see demographic data all the time.

In theory, it helps you better understand your audience’s needs.

Demographics describe “who you are.” They don’t describe “why you do things.” It would be much more helpful to have data on “what issues concern you most” or “what topics you enjoy reading” — psychographic, emotional data about your intent, motivation and beliefs.

Because being Black or White or Asian does not cause you to read this article. Interest in politics does. 

When the BBC shows a graph that 56% with “no college degree” voted for Trump, it implies being “uneducated” causes you to be Republican. Research shows our brains are wired to see those causal relationships, especially to confirm pre-existing beliefs. When you see headlines like “Trump overwhelmingly leads rivals in support from less educated Americans” by PBS or “Why education level has become the best predictor for how someone will vote” by CNN, it confirms that bias over and over again.

Republicans become a mass of “uneducated White men” or a “basket of deplorables” à la Hillary Clinton, or “garbage” à la President Joe Biden. 

So how do we do better?

Well, we can’t ignore demographic data completely. Since women gained the right to vote in 1920 and the Voting Rights Act of 1965 outlawed racially discriminatory voting practices, we’ve been tracking how Americans vote by race and sex. That information is undoubtedly historically valuable, and those demographics are inextricable from the political conversation. 

The solution is more careful data communication. 

We need to stop ogling at simple charts that compare “Whites” vs “Blacks” vs “educated” vs “uneducated” as monolithic red and blue bars.

Yes, that data is true.

Yes, it’s very straightforward to understand.

But those associations without context can lead to false causal relationships. We instead need to focus on charts that show intent, motivation and beliefs: the actual causes for voter decisions. 

Take for example ABC’s exit poll. If you scroll past all the demographic data, you begin to find more of the emotional kind we need. Most Republicans feel strongly that abortion should be illegal, and that undocumented immigrants should be deported. You also see things you might not expect: 81% of Republican voters feel their family’s financial situation is worse today; 63% of Republican voters feel both Trump and Harris’ views are too extreme. 

Data like that turns Republicans into actual human beings. 

As we approach a tense Thanksgiving after another historical election, we need to focus less on the identities that divide us and more on the intent, motivation and beliefs that connect us.

Christopher Chin is founder and CEO of The Hidden Speaker, teaching organizations around the world how to tell stories with data. Originally a graduate of Columbia University with a degree in music, he went on to train at the U.S. Department of Transportation in data journalism and drive analytical storytelling at Fortune 500 companies like Thermo Fisher Scientific, Humana and Fannie Mae. Now he’s an international speaker and trainer with his signature TEDx talk “How Music Can Make Your Data Sing” capturing the unique and interdisciplinary nature of his expertise. He can be reached on LinkedIn.