As pollsters, we want your opinion.
We want your opinion if you’re locked in a Twitter war with a “liberal journalist.” We want your opinion if you’re locked arm-in-arm at the neighborhood anti-Trump protest. And we even want your opinion if you can name more Game of Thrones characters than members of Congress.
Although pollsters value all opinions equally, obtaining the honest views of each demographic isn’t always so easy. It’s understandable why a well-respected opinion research industry group recently concluded that there has been a precipitous drop in polling participation rates.
If your pollster tells you it’s getting harder to get people – especially in certain demographic groups – to participate in survey research, they are probably ignoring the latest tools available to the modern-day researcher.
Until November 2016, the most famous polling debacle was the erroneous 1948 Chicago Daily Tribune headline “Dewey Defeats Truman.” That mistake didn’t end the industry because the industry learned what went wrong and adapted.
The industry is currently in another one of those transitional times. Not just because some polls failed to predict a win for President Trump, but because with all the tools we have available to us today, response rates should be on the rise, not on the decline. We’re able to reach a broader cross-section of people today than at any other time in history. Telephone (landlines and mobile), online and in-person research are all more affordable and more accessible than ever before.
It’s also never been easier to conduct bad research. How can we discern the good from the bad?
It starts by keeping up with demographic trends, news consumption habits and use of technology. As the public’s communication mediums change, so too must our polling methods. The best pollsters identify respondents’ preferred mode of contact and understand the evolving attitudes toward political expression. Pollsters must understand the way that the American people relate to politics and digest national political news.
If you’re reading an online magazine dedicated to practicing politics, you probably need to reset your perspective. That starts by understanding the subtle differences in how most people engage with national politics.
In a recent national Probolsky Research poll, we asked U.S. adults how often they consume national news and then, how often they talk about it. The results weren’t shocking, but they’re far different from the glory days of the Big 3 national network news broadcasts.
News Consumption: “How often do you listen to, read or watch news about national politics?”
- 46 percent of U.S. adults say at least once per day.
- 19 percent of U.S. adults say never or almost never.
Clearly, America is engaged. But, consumption isn’t the same as engagement. When we asked a follow-up question designed to answer how that translates into shaping the opinions of friends and family members, the results were starkly different.
News Engagement: “How often do you talk about national politics with friends, family, co-workers or others?”
- 46 percent of U.S. adults say never or almost never.
- 23 percent of U.S. adults say the topic comes up at least once per day.
Political junkies connected to the non-stop news IV need to take a step back. Most of America is not like us. They may not unplug – it seems no one does anymore — but their Facebook feed is not all politics, all the time. Their conversation at Sunday supper at Cracker Barrel might not include a single mention of how their member of Congress voted on the ACA repeal.
Pollsters, too, must change the way we communicate with this group of aware-but-disengaged voters. When a pollster calls the average American household to try to complete a survey, we may be asking them to do something that feels unnatural to them – to talk about politics.
This doesn’t guarantee poor results. In fact, calling voters (one landlines and mobile phones) is still the most accurate way to poll. But as the realities of communication change and conducting research in an online environment or through some other mode becomes the preferred method for large swaths of respondents, then we must meet the public where they are.
Creativity, technology and a large helping of good old statistics is the new recipe for how opinion research will yield accurate results into the future.
Adam Probolsky is CEO of Probolsky Research, a market and opinion research firm.