Who is Gen Z voting for? That’s the big question for election pollsters

Something incredible happened in the 2018 midterm election. Young people showed up in record numbers. According to Census data, 36% of 18- to 29-year-olds—encompassing both the youngest millennials and oldest Zoomers—reported voting in that election, nearly doubling the 20% turnout that had previously held steady for decades. Whether it was due to a general Trump-era rise in political awareness, or the March for Our Lives movement that started earlier in 2018, the message was clear: Gen Z was shaping up to be an electorally formidable voting bloc.

Now that roughly half of Gen Z is old enough to vote, and capturing the youth vote is a major imperative for both parties, polling young voters is arguably more important than ever. But there’s just one problem in uncovering the raw data behind all those coconut-tree TikToks: getting accurate polling for Gen Z is extremely challenging.

“Polling in all generations has gotten harder, but it is especially hard with the younger generation,” says Paul Johnson, vice president of advanced analytics at the Harris Poll.

Reaching Gen Z

Some youth-based polling challenges are not unique to the current crop of 18- to 29-year-olds. Whichever generation happens to be the youngest at any given moment has typically always been the most difficult group to poll. They’re either too busy hard-launching a career and partying, or they’re as distrustful of institutions as they’re ever going to be. (Or both.) Gen Z does present some distinct difficulties, however.

“Older generations still answer the phone sometimes,” says Chris Jackson, public polling lead at Ipsos. “Gen Z, they really don’t.”

Once upon a time, more or less every American could be reached on a landline telephone. Those days are obviously long gone. Thanks to the rise of robocalls and sketchy spam texts, though, fewer Americans are even reachable by cell phone now, if the call is from an unknown number.

Considering that Zoomers were born into this technological environment, while older generations have adapted to it, they are the least likely to take a survey over the phone. And forget about trying to reach them by text either.

“One of the most common ways of doing text-based polling is known as ‘text-to-web,’ where the text includes a link to take a survey online,” says Alex Podkul, senior director of research science at Morning Consult. “But because Gen Z are digital natives, they’re better about not clicking on links from strangers.”

Making matters more complicated, as hard as it may be to get any members of Gen Z to take surveys, pollsters should ideally include a demographically representative swath of Gen Z in their data. They shouldn’t just reach out to the politically hyperactive young super-donors most likely to respond, but rather a wide spectrum that collectively resembles what this age group actually looks like, according to race, ethnicity, gender, geography, educational level, and other variables.

Creative election polling

So, how do pollsters recruit members of Gen Z to take surveys? There’s a range of promising alternatives beyond the brute force tactic of making hundreds and hundreds of phone calls in search of the youthful unicorn willing to pick up.

“We really try to reach people where they are, and for Gen Z respondents, that means online,” says Podkul.

Morning Consult’s recruitment strategies include traditional web-based outreach on social media, over email, and through online communities. The organization has also lately used more innovative techniques, however, such as tying the survey to an incentive in mobile games. Fresh out of extra lives in a particularly engaging game? Take this quick survey and get some more.

But even just optimizing a survey for display on mobile is a way to make the process smoother for Gen Z respondents.

“One thing that we see in our research is that Gen Z is much more likely to take our surveys that are fielded online through their mobile device,” says Podkul. “When you compare to older generations, they’re more likely to take it on a desktop, laptop, or tablet.”

Another way to reach young people is to first recruit their parents.

Ipsos researchers sometimes do this with “probability panels”—scientifically designed pools of participants who agree to take a number of surveys over time, thus becoming a well that pollsters can reliably keep returning to.

In certain cases, these probability panels end up hosting multiple branches of family trees.

“Sometimes we ask our panelists if they have adult children, and if they do, to connect us with them so that we can recruit them into the panel as well,” says Jackson.

It’s the polling equivalent of nepo babies, and it’s helped Ipsos surface lots of Gen Z respondents.

Shrinking attention spans

Once a pollster has convinced a young person to take a survey, though, what helps ensure they actually complete it is making the survey relatively painless.

“It really does help to keep the survey short and engaging in every way possible, to keep these people in the survey environment,” says Johnson.

Traditional social science surveys used to be 25 to 30 minutes long, and filled with in-depth questions. Most young people simply don’t have that kind of patience anymore. (To be fair, few people in any age demo do anymore.) Each organization is different, but at Ipsos, surveys on average now run seven to 10 minutes. Sometimes, they’re even shorter. Whatever it takes to prevent people from bailing out midway through.

Of course, because of sweeping technological shifts and shrinking attention spans, the entire field of polling is in a state of flux right now.

“I think there will definitely be an industry conversation on the other side of the election,” says Podkul. “Once we can sort of judge the accuracy of how well a poll did among this particular subgroup.”

Perhaps whatever pollsters end up learning from this general election will translate into some new strategies for the next one—the first election Gen Alpha will be old enough to vote in.