As a former political pollster, or at least a reformed one, I have an abiding interest in how the research community addresses the last few years of near misses in public political polling. Many of those at the center of the debate are friends and former colleagues, bosses, and even a few competitors. It’s been impossible for me to avoid the conversation.
While I rarely poll for candidates anymore, political polling, especially during an election cycle, is the visible tip of the iceberg for anyone who pays attention to current events and news. If an alien were to beam onto Earth and follow U.S. news during an election year, their impression would be that the daily ebb and flow (mostly flow) of political horse race polls defines the entirety of primary research in our country. For practitioners on the outside looking in, we know this isn’t true, but are nonetheless compelled to watch, hope, and even help find a way forward.
Some of the brightest minds in the business have already weighed in on the apparent demise of political polling and many are trying to understand where the polling industry goes from here. I could ruin the movie by sharing what I believe is the ending right now, but I’ll save that until after I offer a few observations about our current situation:
1. We don’t know what we don’t know.
We’ve seen participation in phone surveys, especially political ones, steadily drop over the last decade. But study after study reassured us that the attitudes of those who do participate in phone surveys isn’t substantively different from those who don’t. With thoughtful sampling and minor weighting, pollsters have been able to account for demographic differences between our surveys and the overall electorate. But the impact of former President Trump in both 2016 and 2020 was clearly a factor, from people not participating in polls to outright lying to pollsters. What we don’t yet know is how much of an impact Trump had, and we really won’t until he isn’t on the ballot — either as a candidate, or as a gatekeeper.
2. Other measures for political viewpoints must be considered.
Research as a science has always evolved, and it might be more important now than ever in order to account for those who try to game the system by lying about who they plan to vote for. The “horse race” is but one of many different ways of understanding the shape and texture of a political race, even if they aren’t the most media friendly.
I was pleased to see in the most recent Democratic pollster-led memo a significant head-nod toward other forms of research. Online polls don’t yet have the rigor necessary to conduct CD-level research, but smart people are working hard to incorporate opt-in survey panels into political sample frames. Secondary research — understanding what people think by analyzing what they say or do — has been a tactic used in non-political research for more than two decades, and while it doesn’t help illuminate the horse race, it does help shed light on the political undercurrents that might get missed in a more traditional survey.
3. Democrats can’t be the only ones leading this effort.
Primary researchers writ large have a stake in this work. I’ve been doing this for more than 20 years, and I have seen at least three existential industry crises in my career. Each time, however, the primary research industry figured a way out, around, or straight through the problem. Political research exists today because we have consistently answered the bell. From academics to corporate researchers, we all have a vested interest in restoring faith in the basic premise that by asking a few questions of a representative audience, we can provide the insights necessary to solve any problem. If we allow the public to lose faith in political research, the impact will spread beyond political pollsters. Those of us who work in the connected spaces, like public affairs, crisis and reputation management, and brand and influencer messaging, will all find it increasingly hard to justify large budgets on critically important research.
4. Political polling isn’t going away.
This is an inexorable truth. Horse race measures in recent years have undermined faith in an industry that produces so much more good data than a simple, media-friendly question. The insights and guidance that political polling generates will continue to provide value to political campaigns. But we do have a continuing responsibility as an industry to ensure the data we provide is meaningful and reliable. Republicans and Democrats alike share this responsibility. GOP pollsters were also fooled in 2020, and there are just as many smart Republican pollsters out there as there are Democrats. At the end of the day, the solution must be a smart solution, not a partisan one.
I remember not too long ago when Democratic and Republican professionals happily beat each up during their day jobs, and then worked together on bi- or non-partisan efforts, or just sat down for a beer after work. Some of that, I think, has been lost in the last decade or so, and I for one hope that the search for a polling solution will bring a little of the bipartisan love back to politics. The success of our industry will depend on it.
Jason Boxt is the founder and CEO of 3W Insights, a strategic research consultancy. Prior to 3WI, Jason led the DC office of PSB Research, and before that headed up research at the public affairs firm Glover Park Group. He got his start in politics at the Democratic polling firm Global Strategy Group.