Pollsters by now are used to getting hit with the negative post-cycle headlines that point to publicly released misses. This year has been no different as attention has zeroed in on Iowa’s J. Ann Selzer, whose pre-election survey pointed to a surprise win for VP Harris in the Hawkeye state.
President Trump ended up winning the state by double digits – a miss that has resulted in fresh attention on the path forward for survey researchers.
Many pollsters are shrugging off the post-cycle hubbub. Mike Greenfield, CEO of Change Research, said he felt “incredulous” when he saw Selzer’s pre-election poll showing the Democrat leading. “It didn’t seem plausible, but I feel like most of the other polling held up pretty well,” he told C&E. Greenfield credits using vote recall as a way to keep his own surveys accurate.
“If there’s a systematic shift and people who took multiple surveys with us in 2022 [were saying] they voted for one candidate and then when we survey them again in ’23, if they tell us they voted for somebody else, we’re going to adjust for that. We didn’t see that. There’s like 99-percent consistency in terms of how people say they voted over time.”
Moreover, he isn’t concerned about the so-called “shy” Trump voter or any potential lasting difficulty relating to picking up support for a similarly styled Republican candidate.
“There are two things which are always potential issues. One is to what extent do people accurately forecast their own turnout?” he said. “And then the other is that there’s just selection bias and who answers surveys.”
He added: “You can’t know everything about the people who aren’t taking your surveys.”
Change Research is continuing to unpack its own polling post-Election Day. One data point it found noteworthy relates to post-cycle attention being paid to podcast interviews done by the presidential candidates.
The company found that Harris won podcast listeners by four points in its most recent survey (50 to 46 for Trump). But that was a decrease from Biden’s performance among podcast listeners. The president won them 53 to 41 for Trump in 2020. The difference was even more stark among podcast listeners under the age of 50. Harris won that group 54-41, but Biden won 59-32.
As for the pre-election dustup over a possible Harris appearance on Joe Rogan’s podcast: “It could be consistent with this [data], but I would be hesitant to assert that without seeing more,” he said. “It could be [part of] a shift that is broader, that has been in the works for over the past four years.”
