Sure, I know Iowa’s nearly two weeks away. But I’m taking next week off, so I thought I’d go out on a limb and make some predictions now about the January 3rd caucuses.
Barack Obama will win the Democratic caucus in a landslide. OK, that was too easy. Let me take a stab at the Republican race.
For a while, it looked like Newt Gingrich was the one to beat, but the latest state and national polls show his popularity sliding. As I observed in a previous column, all the other ABM (Anybody but Mitt) candidates enjoyed five to six week upswings in their poll numbers followed by sharp drop-offs. Gingrich reached the six-week mark about 7 days ago – and right on cue, his poll numbers started to tumble. [I’m beginning to think that an ADD epidemic has struck one-fifth of the GOP primary electorate. The predictability of their fickleness is uncanny – and unnerving.]
Gingrich will continue to go down, and no matter what share of the vote he takes in Iowa, if he doesn’t win it, he will have performed below expectations. And that is what will be focused on – performance against somewhat arbitrary, and often erroneous, expectations.
This brings us to Ron Paul. He is surging in Iowa (and experiencing a slight but noticeable increase in his national support). Among the four Iowa polls out this week, he leads the pack by 3 to 6 points in three polls (Insider Advantage, Iowa State U, PPP) and places second to Mitt Romney by 5 points in one (Rasmussen).
Are the polls accurate – will Ron Paul win Iowa? It’s really too early to tell. Iowa caucus polls are notoriously unstable even in the final weeks. We saw this for both parties’ caucuses in 2008. It wasn’t until the final Des Moines Register poll conducted just days before the last caucuses that we got a fairly accurate read of what would happen.
Dave Peterson from Iowa State’s Harkin Institute, one of the academics behind that poll, remarked, “I think Paul probably under-polls.” I actually think it’s the opposite in this case – Paul is now over-polling, mainly because the demographic mix of voters in these Iowa caucus polls may unrepresentative of actual caucus-goers.
Keep in mind that approximately 610,000 registered Republicans and 705,000 registered independents are eligible to show up at the GOP caucuses. In 2008, just 119,000 did. In 2000, the turnout was about 88,000. In other words, much less than 10% of the eligible electorate will show up. Determining who they are for a poll is not easy.
For one, most of the public pollsters rely on self-reported intentions of caucus attendance. Unlike primaries where turnout records are publicly available, the Iowa Secretary of State does not maintain caucus attendance – the parties do. If you want past turnout reports you have to buy those lists directly from the parties at a steep cost. Furthermore, past caucus attendance is really not a good barometer of current intentions (again, unlike primaries).
Polling the Iowa caucuses is notoriously difficult. Which brings us back to Ron Paul. He does extraordinarily well among younger (i.e. under the age of 45), independent voters. The released cross-tabulations from two polls demonstrate this.
The Insider Advantage poll shows Paul with nearly half the vote among “likely caucus goers” under the age of 30 and leading with 30% among those age 30 to 44. He runs even with Romney (22% each) among those age 45-64, but Romney pulls ahead among those age 65 and over with Paul dropping to 5th place. Among self-identified independent voters, Paul leads with 28%, compared to 19% for Rick Perry and 17% for Romney. Among Republican partisans, Paul enjoys a narrow 22% to 20% edge over Romney, with Perry (15%), Gingrich (15%), and Michele Bachmann (11%) close behind.
The Iowa State Poll paints a similar picture. Paul garners a clear majority of the vote among those under the age of 45. However, Newt Gingrich has the lead among older voters. It’s worth noting that the Iowa State Poll did two things that are different from most other polls. They used a very long field period, starting their interviews on December 8, when Gingrich’s popularity was still at its height. They also used a panel sample – in other words, they re-interviewed a subset of the voters they spoke to in their November poll. It’s unclear what impact the use of a panel has on their results, but the extended field period certainly does not capture the exceptional fluidity of this race.
I think that these polls may be inflating Paul’s support because they over-represent a segment of the electorate that is less likely to turn out. Keep in mind that the caucuses are a long process. You have to listen to hours of speeches from representatives of each candidate before you get to cast your vote. And you have to do this on a cold Iowa workday.
Who’s more likely to show up under those conditions – younger, independent-minded voters or older, partisan stalwarts? Exactly. Now, this runs counter to what some other astute observers have opined.
I’m not saying there won’t be young voters at the caucuses. It just won’t be as many as the polls suggest. According to the 2008 Iowa exit polls (or technically “entrance” polls), 27% of GOP attendees were under the age of 45. The Insider Advantage poll has this group at 40% of the electorate and Iowa State has them at 37%. An October Des Moines Register poll noted that seniors are less likely to self-report intended caucus attendance than they have been in past years. While I accept that, I’m not sure if young independents will be as motivated to come out as the more recent polls suggest (young conservatives may be a different story).
The real issue here is partisan identity. In the 2008 exit polls, 13% of GOP caucus goers identified themselves as “independent.” However, voters of this affiliation make up 30% of the Insider Advantage sample and 38% of the Iowa State sample. Even without a contested Democratic caucus to draw away some independents (as can be argued happened in 2008), there is no way that independents will make up anywhere near that proportion on January 3rd.
In the end, I predict – and here’s where I go out on a limb – that Mitt Romney will win Iowa with about 27% of the vote. Ron Paul will come in a close second, but since many polls show him with a lead, he will be seen as having underperformed. I also think the third place finisher will be Michele Bachmann (perhaps due to an influx of young conservatives). She will not be far ahead of Gingrich and Perry in total votes, but her “surprise” showing will be the story of the caucus.
So with the media focused on Romney’s win (“Is he invincible?”) and Bachman’s better than expected performance, Ron Paul’s strong effort will likely get squeezed out of the media coverage. And that’s what really matters as attention turns to New Hampshire and the South.