Seven Questions About Iowa
Published in
5 min readJan 31, 2016
- Will the polls reflect the actual universe of people likely to turn out in a caucus — and what’s the downside for Donald Trump (if any)? I and others have been raising this question for months, and so it’s at the top of the list of questions I hope to have answered on Monday night. Even gold standard Iowa surveys like Ann Selzer’s have been drawing from a pool of potential Republican Iowa caucusgoers that numbers 300,000, give or take (based on Iowa voter registration statistics and the number of voters in a poll who screen in to be questioned about the Republican caucus). The actual turnout number seems certain to be no less than 120,000 (the 2008 and 2012 totals), and the 160,000 I predicted on Twitter the other day would be considered a blowout. Normally this doesn’t matter so much, for two reasons. First, turnout is much higher in primaries and general elections, meaning that most of the people polled will have voted and those pollsters inaccurately believe will vote can’t throw as big a wrench in the works. This is not so in a caucus. Second, we haven’t had a Republican candidate quite like Donald Trump before, who polarizes the electorate based on likelihood to participate. In most Republican primaries, the preferences of unlikely voters are roughly those of likely voters. If these suspicions about the polls including people who will never vote are well-founded, we will see Donald Trump underperform his final polling numbers on Monday night, and he is particularly vulnerable to this in a caucus state like Iowa. If he doesn’t, his numbers in New Hampshire and other states are also likely to hold up.
- Does late momentum matter more than the final polls? In 2012, Rick Santorum was 6th a week out and 3rd in the final polls. The final polls in the 2004 Democratic caucus had John Kerry in third place with 17 percent and John Edwards at 11 percent, with Howard Dean leading at 27 percent. Kerry and Edwards ended up dominating the field, with 38 and 32 percent respectively. As reliable a guide as the Selzer poll has been in the past, it reflects, on average, where the race stood on January 28th, four full days before the caucus. In those four days, between 30 and 40 percent of Iowans will be making their final decision, which means that, at best, a “gold standard” final poll, traditionally conducted, will only be 60 to 70 percent complete. If we see Marco Rubio doing much better than the final polling, a blowout in either Clinton or Sanders’ direction, or a big surprise of any kind, it won’t necessarily mean that the polls were wrong, but that they didn’t capture this late momentum.
- Will field matter? The tightening race between Hillary Clinton and Bernie Sanders are what strong ground games were created for. Even when Clinton led national polls by 40 points, she started the campaign with 50 staff on the ground in Iowa and this has grown to more than 100, a decision which now looks foresighted. Bernie Sanders has an operation on the ground that is near that of Clinton in its size and sophistication, but it got started later. If Clinton pulls out a win, notch one for the importance of field. Likewise if Sanders performs much better among young voters (reprising Obama in 2008). If Trump underperforms, that also likely means he could have benefited from a stronger turnout operation to bring out his coalition of the unlikely, who’ve often been quoted in the media professing basic ignorance of the caucus process.
- Will Trump alternatives finally get media oxygen? The theory that Donald Trump would fade as the voting got closer rested on the notion that he couldn’t sustain his dominance of media coverage forever. But, alas, Trump’s dominance of the news cycle has continued. Trump is gaining strength as the caucuses near — mostly because he’s been effective at swatting down serious challengers like Carson and Cruz. His debate stunt also showed that his strategy of sucking the oxygen out of the room could work within days of a vote, not just in the silly season of July and August. The question is now whether this will actually continue indefinitely, and likely result in his nomination? Will the media metrics in the final 48 hours show that Iowans shifting their attention to the non-Trump alternatives, a result we will also then expect in other states? And how will the dynamics change once we have this crucial first data point of 2016 in hand, and media coverage kicks into higher gear?
- Are voter registration numbers a good predictor of Republican caucus turnout? We know that Iowa Republican voter registration has gone up slightly in the past few months, while Democratic registration has been flat. This small increase for the GOP is nowhere near what the Democrats saw in 2008, when caucus turnout doubled from 2004, and Barack Obama scored a blowout opening to the primary season that would propel him to the nomination. But is voter registration a good proxy for turnout on the Republican side, given that voters on the edge of participating are not likely to be unregistered young people, but simply Republican general election voters who wouldn’t normally participate in a caucus? The Iowa Secretary of State lists 612,112 active Republican voters as of January 4. Even without a wave of new registrants, there may be plenty of room for record caucus turnout.
- Was Ted Cruz’s last week really as bad as it seemed? The media proclaimed Ted Cruz the loser of the debate, with a particularly nasty Des Moines Register front page on Friday morning. But while this wasn’t Cruz’s best debate, the narrative that this has worsened his slide in Iowa polls seems to be a bit of a media creation. Cruz has consolidated the conservative lane relatively early, and enjoys a harder base of support than other candidates. While stylistically Thursday was not his best performance, Cruz has the advantage of having a message that a large number of Republican base voters want to hear, which lowers the bar on his being Cicero in the debates. The Selzer poll — imperfect a measurement as it is, with one day of polling post-debate — does not yet show Cruz in any kind of death spiral. Will #marcomentum, the media meta-narrative, or the fallout from his social pressure mailer push him into one?
- Are social pressure mailers now ruined? The science is clear: Social pressure tactics increase turnout. But most of them aren’t prominently sponsored by candidates, deployed in a nomination contest, or use inflammatory language like “Voting Violation.” And while it’s well known that voters don’t like them, they also aren’t accompanied by bad news cycles and exposure to voters outside the intended universe. Will this have a meaningful impact on Cruz’s chances, or like most inside baseball stories, is the fallout simply a distraction? Perhaps more importantly moving forward, will this clumsy execution deter others from using this valuable technique in a general election, when the risk of alienating voters on your side is lower?
