Category Archives: Polling

Two More Polls: Where Things Stand in Iowa

At this rate I’ll never get my grading done.

Two more polls just in and (fortunately!) they are consistent with my earlier analysis posted today.  The first poll is a Rasmussen automated survey conducted Dec. 28, and the second is a telephone survey by American Research Group (ARG) from Dec. 26-28 (both polls have a margin of error +/-4%).  For ease of interpretation I’ve put the five most recent Iowa polls in the following table. (Note, however, that the CNN poll took place with a day off for the Christmas holidays.)

Candidate Public Policy Polling (12. 26-27.11) CNN (12.21-24, 12.26-27.11) Insider Advantage(12.28.11) American Research Group (12.26-28.11) Rasmussen (12.28.11)
Romney 20% 25% 17.2% 22% 23%
Paul 24% 22% 17.3% 16% 22%
Gingrich 13% 14% 16.7% 17% 13%
Santorum 10% 16% 13.4% 11% 16%
Perry 10% 11% 10.5% 9% 13%
Bachmann 11% 9% 11.8% 8% 5%


Once you account for differences in sampling and the dates of these polls, they tell somewhat similar stories. Keeping in mind all the usual caveats about polling a caucus state, what do these more recent surveys suggest is happening in Iowa today, five days before voting?  I offer the following observations.

  1. Romney is going to get about 23% of the vote, give or take a few percentage points.  That is, he’ll finish close to what he did four years ago.  Whether that is enough to win depends on several other factors.  See below.
  2. Paul’s support, I think, may be softening just a bit, probably because of doubts, fueled by negative campaign ads, regarding his foreign policy views.  Note that his unfavorable ratings have risen in the last week and are now in negative numbers.  Of course, with him it all depends on the relevant turnout of independents and younger voters.   Will the college crowd be home for the winter break?  Will they turn out? We know that he has a core group of supporters who will show up no matter what.  But I don’t think that constitutes much more than 20% of likely voters.
  3. Gingrich seems to have weathered the avalanche of negative ads directed at him and is stabilized at about 16%.  In my view, he has a greater potential upside than either Romney or Gingrich, primarily because his support cuts across demographic and party lines.  His biggest problem is extremely high unfavorable ratings.  Much depends on whether his last-minute advertising pitch can swing voters his way.  He is now, finally, on the air with new television ads.  Unlike Romney and Paul, however, he lacks a campaign infrastructure for getting out the vote.
  4. Santorum seems to be benefitting from the “last Christian standing” dynamic; the evangelicals have danced with everyone else not named Mitt or Ron at least once (please, no touching on the dance floor), so it now appears to be Santorum’s turn.  He’s peaking at the right moment, has high favorable ratings (because no one has bothered to attack him) and could challenge for third place.
  5. Perry is a wildcard – he’s blanketed the state with advertising and has made a strong pitch for the conservative vote but it’s not clear that he’s attracting more than 12% support.
  6. Bachmann is in danger of getting winnowed.  The key question, for me, is whether her supporters peel off in the caucus room and switch to Santorum.  Note that unlike in Democratic caucusing in Iowa, where there’s plenty of jockeying for position, there’s not a lot persuasion that goes on in the Republican meetings, and the vote is by secret ballot.  So it’s a lot easier to dump the one you brought to the dance.

The results five days from now, it seems to me, hinge on two factors.  What is the relative percentage of independents (and even Democrats) that turn out?  And do the evangelicals coalesce behind a single candidate, or two, or three?  I think Romney’s and Paul’s support is what it is in the polls.   I’m less confident about the other four.

More as polling comes in.

Is Mitt It? The Pundits Say Yes, the Polls Say No

We are witnessing a fascinating struggle between the polling data in Iowa, and how Republican Party leaders and pundits are interpreting that data.  Yesterday, the National Journal ran this story under the headline “Romney: The New Frontrunner in Iowa”.   It linked to a number of print and online articles that collectively indicated Romney was poised to win the Jan. 3 caucus and position himself to close out the nomination contest shortly thereafter. Thus, Politico proclaimed that “Romney was in striking distance of Iowa Win.”    The Washington Post proclaimed  that Romney’s “stealth campaign” gave him a “real shot at winning Iowa.”   This Huffington Post piece  ran under the headline “Romney poised to do well in Iowa.”  The Hill opined that Romney “could lock up GOP nod with win at Iowa.”   The Los Angeles Times went so far as to claim that Romney could “win” Iowa even if he didn’t win there!  Josh Krashauer concludes his National Journal piece with this confident assertion:  “Make no mistake: If Romney wins the Iowa caucuses, he’s on a glide path to the Republican nomination.  And with newfound scrutiny over Paul’s racial record, Gingrich losing momentum and the evangelical base vote split between Bachmann, Perry and Santorum, it’s looking awfully likely that Romney will come away as the big winner on Jan. 3.”

Big Winner?  You might think this assessment and the accompanying flurry of news story touting Romney’s prospects are driven by recent polling data indicating that Romney was pulling ahead in Iowa, or at least that his poll numbers were on the rise there.  But you would be wrong.  The reality is that there are no signs of a Romney surge in Iowa.  At best one might argue that he’s climbed a couple of percentage points back to where he was a month ago, but even that’s uncertain, given the probabilistic nature of survey sampling.

Consider the latest Iowa poll released late yesterday by Insider Advantage.  It shows Romney at 17.2%, in a dead heat with Paul (17.3%) and Gingrich (16.7%).   (The bigger news in the poll is that Santorum has climbed to 4th, at 13.4%, which is consistent with yesterday’s CNN/ORC poll that had him third, with 16% of the polling support. More on that in a later post.) One week ago Insider Advantage had Romney at…. 18%.  Well, that’s one poll you say – surely he’s climbing in all the others?  Nope. PPP’s latest had Romney at 20% – exactly where he was a week before in the previous PPP poll.  Rasmussen?  The last two polls, again a week apart, have him at 23% and 25%.  Really the only poll that suggests any growth in Romney’s support is yesterday’s CNN poll, which has him at 25%, which is 5% more than its previous poll. That previous poll, however, dates back three weeks.

Indeed, right now Romney is at 21.8% in the RealClearPolitics composite polling tracker – which is almost exactly where he was a month ago and, in fact, about where he was last July!   Indeed, he’s on track to finish below the 25% he received in Iowa four years ago!   For all intents, despite several millions dollars of paid advertising, countless visits, endorsements by party leaders, and even a change in clothing (Mitt’s ditched the prep look and is now dressing down) Romney’s been treading water for four years in Iowa.  That’s the real story.

Look, Romney could very well finish first come Jan. 3 – but if current trends hold it won’t be because of a surge of support in the last two weeks of the race. Iowan voters show no signs of coming around to Mitt.  It will be because unlike in 2008, the Tea Party/evangelical voters who comprise more than 50% of the likely caucus voters didn’t coalesce behind a single alternative candidate.   Their failure to do so make it possible that Romney’s support will go down from 2008 – and yet he will finish first.

If that happens, it will be very interesting to see how the media spins the results.  For reasons that I think must be based on evaluations of candidate electability, the Republican party establishment  – by which I mean party leaders (see his endorsers) and their media counterparts (think Will, Brooks, and Krauthammer) – are completely in the tank for Romney, despite the fact that he has yet to demonstrate deep or broad support among hard-core Republican partisans.  There’s a dispute in the political science literature regarding whether the winning candidate gets a boost from Iowa.  (I’ll address this in a separate post.)  But it partly depends on media reactions and the expectations game.  If Mitt can’t increase his support in Iowa from four years ago, that suggests he’s not gaining strength as a candidate.  But that may not be how the media interprets the results.

By the way, the Insider Advantage poll released late yesterday showing a three-way tie in Iowa included 69% Republicans, 27% independents, and 3% Democrats – that’s a relatively low number of Republicans, which suggests the poll may be slightly understating Santorum’s support although it’s hard to be sure without an ideological breakdown as well.  The other interesting finding is what appears to be a gender gap in both Paul’s and Romney’s support;  Paul leads among men with 23% of the vote, but he gets only 12% among women.  Romney’s support is reversed; he receives the highest support among women, with 23%, but only gets 11% from men.  Gingrich, in contrast, polls equally well with both genders (17%).  Indeed, of the three front-runners, Gingrich shows the most consistent support across party, gender and age lines.   He comes in second among Republicans, second among Democrats and third among independents.  And while Paul draws disproportionally from those 44 years old and younger, and Romney’s core support is from the 45 and older crowd, Gingrich draws about evenly from all age groups.  Maybe Gingrich, and not Romney, is most electable?

Will It Be Santorum? In Iowa, Timing is Everything – the latest Poll

In politics, as in life, timing is often everything.  A second Iowa poll has just been released, and it suggests that conservatives in Iowa may be  – I stress may be, since this is one poll – coalescing behind Rick Santorum.  This poll, conducted by ORC on behalf of CNN, is based on telephone interviews with 452 likely Republican caucus participants, and was in the field from December 21-24 and December 26-27.  (The sampling error +/-4.5 percentage points.)  As the following table shows, it has Romney leading with 25%, followed by Paul at 22%.  The big surprise is that Santorum has climbed into third with 16%, which statistically ties him with Gingrich.

Romney 25%
Paul 22%

Santorum 16%
Gingrich 14%
Perry 11%
Bachmann 9%
Huntsman 1%
Someone else (vol.) *
None/ No one (vol.) *
No opinion 2%

Santorum’s rise in this poll seems to be fueled by the anti-Mitt and anti-Paul groups I discussed in my previous post – that is, social conservatives and Tea Party activists. Among those who self-identify as “born again” Santorum leads all candidates with 24% support. Among “conservatives” he is in a statistical tie with Romney (22% to Santorum’s 21%) at the top.  Note, however, that in contrast to the PPP poll, this survey does not appear to include self-identified moderates or Democrats, although it’s hard to tell for sure because explicit breakdowns by party or ideology are not provided.  Still, it is the first evidence we have that conservatives may have decided that Santorum is this year’s Huckabee.

If so, it would be only fitting; Santorum is the only Republican competing in Iowa who has not experienced a surge in support.  If, in fact, conservatives have now decided to coalesce behind him, his timing is impeccable, since a surge at this late date will not provide time for candidates or their shadow SuperPacs to reorient their negative message machine against him.

In terms of issues, perhaps the most interesting finding coming out of the poll is the number (60%) citing the deficit as the most important economic issue, trumping both jobs (20%) and taxes (13%).  These are not Romney conservatives, for the most part.

What is most startling about this poll, however, is that with just six days to go, fully 43% of those polled say they may still change their mind!  At the same time, however, there are significant blocs of voters who say they won’t support specific candidates under any circumstances; 35% say they won’t vote for Romney, 36% say they won’t vote for Bachmann, 39% won’t support Newt and 41% won’t support Paul.  Interestingly, however, only 25% say they won’t vote for Perry. (Unfortunately, for some reason Santorum wasn’t included in this question!)

So, where do things stand?  It appears that caucus goers may be sorting themselves into three voting blocs: “establishment” Republicans who back Romney, libertarians who support Paul, and social conservatives who are the biggest bloc, but who are still working on that coordination problem.  I think it’s pretty clear that despite the drumbeat of support by the party establishment, Romney has come close to maxing out in Iowa – indeed, it’s not clear he’ll even match his total from four years ago.  (I’ll have another post on how the pundits are missing the story with him in a bit.)  It also appears that Paul may have hit his ceiling. With six days to go,who wins Iowa may depend on just how quickly and thoroughly social conservatives move toward a single candidate.  Is it Santorum? If so, that may finally change his google, er, issue!

P.S. 6:14 p.m. I hope everyone appreciates that I’ve avoided stooping so low as to engage in the obvious Santorum word play regarding his polling move from behind….how he’s come from the rear of the pack …. ok, never mind.

P.S.S.  As I look through the internals of the CNN/ORC poll, it does seem they don’t include any non-Republicans in their survey, in contrast to the PPP poll.  That explains why Santorum can be third in one poll, but trailing the field in the other one. Just another reminder that who turns out next Tuesday makes all the difference in the world.



All Iowa, All The Time: Who Will Finish Second?

With less than a week to go before the Jan. 3 Iowa caucus, the latest PPP Iowa poll continues to show Ron Paul in a slight lead over Mitt Romney, with Newt Gingrich in third.  None of the numbers have changed very much since the last PPP poll from a week ago.   As I noted then, Paul’s lead is predicated on the assumption that independents and Democrats will show up in considerably greater numbers in the Republican caucus than they did in 2008 – a not unreasonable forecast given that there is no Democratic caucus this time around. In the PPP poll, fully 16% of those polled voted in the Democratic caucus in 2008 – not the Republican.  Whether Democrats will actually turn out in those numbers is the key to evaluating these polls.

As Mark Blumental notes at, most recent Iowa polling tells a similar story to the latest PPP results: Paul holding a slight lead over Romney, with Gingrich trailing both by about 4-5%. Here’s Pollster’s listing of the six previous Iowa polls, and the trends.

As I’ve repeatedly cautioned, however, these polls are based on assumptions regarding the level and composition of the Republican caucus come Jan 3 – assumptions that involve not a little guesswork. Remember, only about 5% of Iowa’s roughly 2 million registered voters will actually attend the caucuses.  In this vein, Blumenthal looks at past polling the week before the Republican Iowa caucus to see how reliable it was.  As this chart shows, in general the final week polls do a good job forecasting the winner.

When they have been off, however, it usually has been in underestimating the support for the second-place candidate.  (The exception was Romney in 2008 who slightly underperformed his final polling numbers.)  Why might this be?  Note that each of the three instances in which the Iowa polls underestimated the second place finisher’s vote involved a social conservative.  To me that suggests that conservative voters, who tend to turn out disproportionately in caucus events, essentially solved their “coordination” problem in the final week and decided to support a single candidate who came closest to embodying their values. Now look at the Republican candidates in 1996, 2000 and 2004.  In each of these races there were potentially two or more candidates who could appeal to the social conservatives. And in the end these voters tend to move in large numbers to the candidate who came closest to their social conservative views. So, in 1988 conservatives really had to select from among Pete DuPont, Jack Kemp and Pat Robertson.  Most of the undecideds went to Robertson. Eight years later conservatives could choose from among Phil Gramm, Alan Keyes, Steve Forbes and Pat Buchanan.  They broke late for Buchanan. Finally, in 2000 Gary Bauer, Keyes, Forbes and Orrin Hatch battled for the conservative vote, with Forbes getting most of the late deciders. What about 2008?  Here Huckabee really faced no opposition for the conservative vote, and he won their support early.

So what does this say for 2012? Here’s where things get tricky.  Paul’s libertarian values are not fully compatible with a true social conservative’s world view; most of his polling support comes from independents, Democrats, ideological moderates and young voters.  That’s not the profile of the voting bloc that pushed Huckabee to the top four years ago or that coalesced behind conservative candidates in the past. If I’m right, this suggests that Paul may have reached his ceiling at about 22%. Nor do I see these voters flocking to Romney – his support in the polls this time around hasn’t even consistently reached the 25% he won in 2008.

So where will the true conservatives go? Note that evangelical Christians constitute about 47% of the projected caucus turnout. Fully 52% of those surveyed get their news primarily from Fox.  Right now, the four other candidates – Gingrich, Perry, Santorum and Bachmann – are winning a combined 44% of the polling vote.  If any of the four could win over this vote, they would take the caucus easily over Paul or Romney.  But is anyone primed to do so?

At this point, the barrage of negative ads targeting Gingrich these last two weeks may have fatally wounded his support among conservatives. In the PPP poll, his unfavorability rating is the highest of all six candidates. He is counting on a last minute advertising campaign launched by a SuperPac to turn those numbers around, but it may be too late.

That leaves Perry, Santorum and Bachmann.  Of these Santorum has the highest favorability/unfavorability rating in the PPP poll, at 56/29 (with 15%) unsure, and he is the second choice of 14% of those polled, which is the strongest support in this category. But these figures are only slightly better than Bachmann’s, who’s the second choice of 12% of those polled, (tied with Gingrich for second in this category), and who has a 53/37/10 favorable/unfavorable/don’t know rating. Perry is the second choice of 10%, with 48/40 favorable/unfavorable ratings. But he has the strongest media presence in the state, vastly outspending Bachmann and Santorum on paid advertising. I tend to think that Santorum may have stronger upside  than Bachmann, primarily because of his higher stature as a former Senator which may give him a bit more credibility; Bachmann is still fighting the perception that she’s less seasoned.

However, keep in mind that social issues are cited as the most important problem by only 12% of those surveyed, and illegal immigration by only 3%. In contrast to past years Iowan caucus goers are focused primarily on jobs and the economy which could play more to Perry’s strengths than to Santorum’s or Bachmann’s.

Finally, 28% of those surveyed say they may yet switch to another candidate, and 5% say they are unsure of who to support.

In considering all these factors, it seems to me that there’s a high likelihood that social conservatives won’t coalesce behind any single candidate this time around.  If they don’t solve their “coordination” problem and split the vote, that benefits Paul and Romney.  And here’s where the media expectation game becomes important – even though neither of the two frontrunners will have commanded anywhere near a strong plurality of the Iowa vote (never mind a majority), the media will invariably suggest that one of these will have the “momentum” coming out of Iowa and heading into New Hampshire.  That won’t be true, strictly speaking, but that will hardly matter to the media coverage.

Before we get to that point, however, there’s still six days to go in Iowa and the race remains very fluid – much more so than in past years. Conservatives may yet decide to rally behind a single person – if they do, that candidate will be the perceived winner coming out of Iowa, and they may be the actual winner as well.

In the meantime, if anyone tells you they know what is going to happen, they are lying. Unless it’s me.

Stay tuned.

Last Minute Polling

Busy day today, but I wanted to send one quick Iowa polling update.

In Iowa, the most recent poll (by American Research Group) shows very little change, with Paul (21%), Romney (20%) and Gingrich (19%) essentially tied for the lead. (The poll’s margin of error is +/- 4 points).   No other candidate reaches double figures, but 12% remain undecided.  Once again, Paul does better among independents. who constitute 22% of the poll.  Although the race remains very fluid, with a week to go it appears to be a toss-up among these three.  Perhaps the most interesting takeaway is how steady Romney’s support is – he never gains, but doesn’t lose either.

If I can, I’ll post a more extended discussion later tonight.