How would the Electoral College vote look if Mitt Romney and Barack Obama tied in the popular vote? Naturally there are myriad possible situations where candidates tie in the popular vote, and thus the possible distribution of Electoral College votes are legion. We sought to figure out which of these may be most likely.
I have to put a big (huge, leviathan, ungodly large) disclaimer here: this is not a projection for 2012. The numbers are predicated on what would likely happen if the popular vote split evenly between the candidates. This is a what if scenario, not an expectation.
The above figure shows the distributions of expected Electoral College outcomes under a 50/50 vote (that is, under a scenario agnostic to national swings). According to 1,000 simulations of state-level models, Romney finds himself winning the Electoral College (i.e., gaining more than 270 votes) about 53 percent of the time. Obama wins the Electoral College in roughly 47 percent of the simulations.
The reason for this major EC swing toward Romney is two-fold. First, in order to estimate a 50/50 popular vote, we had to scale up Romney’s support amongst white voters, under the assumption that support from black and Latino voters would stay roughly the same as 2008. That helps to push some narrow-win states for Obama, like North Carolina and Indiana, pretty firmly into Romney’s camp.
The second reason for some of the swing is reapportionment of electoral votes, which does Obama a slight disservice. Georgia, Texas and South Carolina, for instance, gain electoral votes, while Illinois, Iowa, New Jersey and New York lose them.
The general take away: in 2008, a 50/50 election would likely have still favored Obama electorally. In 2012, this does not seem to be the case. Rough parity in popular vote projections gives Romney a slight edge in the Electoral College.
(more…)
by Brice D. L. Acree Filed under:
Elections Forecasting
A couple days ago, Harry discussed the shady results posted by pollster Insider Advantage during the 2012 primaries. Insider Advantage previously worked for Gingrich, and as Harry pointed out, their results seemed to favor Gingrich until the closing hours of the race, when their results become suddenly more in-line with other polls in the field. Per Harry:
Now, I’m not saying that I know for sure that Insider Advantage polls are purposely biased towards Newt Gingrich, but doesn’t it look awfully strange that their founder’s former boss has been the beneficiary of surveys that are constantly different than the average poll?
The results Harry reported are good, but I process data better visually, and so I was curious what the graphical story behind Insider Advantage might be. Consider the following figures from Iowa and South Carolina:
The figures above show the trend in Romney’s lead over Gingrich as presented by Insider Advantage versus all other pollsters. It is startling how similar the results from Iowa and South Carolina are. In both cases, Insider Advantage strongly “favors” Gingrich vis-à-vis Romney in the couple weeks before the vote, but as the vote approaches quickly converges to the other pollsters’ results.
As Harry points out, the charge of bias is difficult to make. But the evidence, if not incontrovertible, should give the press pause before they publish Insider Advantage’s results as though they were accurate.
by Brice D. L. Acree Filed under:
2012 Primary Elections
Pollster bias, the idea that a survey house’s polls constantly favor one candidate or party vs. another, remains one of the more controversial subjects in the polling industry.
Bias in polling is an important subject because polls not only tell us who is winning, but they influence news coverage. Voters, especially in primaries, like to vote for viable candidates.
An arguably flawed, though not purposely biased, CNN/ORC Iowa poll a few weeks ago illustrates this point. The poll gave Rick Santorum his most favorable Iowa numbers to date and favorable news coverage followed. The news coverage assisted his Iowa surge and fundraising prowess to make him a viable candidate.
Missed in this CNN/ORC drama was the potential bias of another pollster: Insider Advantage (IA). IA is a “nonpartisan” polling firm headquartered in Georgia and founded by Matt Towery.
Towery’s firm has constantly help to shape the Republican primary narrative by frequently polling the early Republican primary contests.
What you probably don’t know is that Matt Towery “ran [Newt] Gingrich’s political operation in the 1990s”. This potential conflict of interest is known by some, but is certainly not echoed enough by those who cover and recite Insider Advantage polling data.
Of course, Towery’s past relationship with Newt Gingrich would not be a big a problem if IA polls showed no bias in favor of the former Speaker of the House.
Just in the past 24 hours, Insider Advantage released a poll showing Mitt Romney turning a 2% South Carolina edge into a 11% lead over Newt Gingrich in an amazing 4 days. This would suggest the opposite of a bias.
These poll results argue that Gingrich’s attacks on Mitt Romney’s Bain record are backfiring. A new We Ask America poll also indicates that Mitt Romney is gaining steam in the Palmetto State over the past few days.
But let’s remember that IA’s poll a few days ago was the most pro-Newt Gingrich survey among the numerous polls produced in South Carolina.
IA’s first poll was 4% more favorable to Newt Gingrich than the average of all the other SC polls taken during the same period.
It’s certainly not unusual for any one poll to be slightly out of the mainstream. A poll with 500 voters has a theoretical margin of error of 4.4% for each candidate’s percentage.
Yet, this is not the first time that IA has been the most pro-Newt pollster. Iowa and New Hampshire also saw its share of pro-Newt Insider Advantage polls, which does suggest bias.
During the December 11th to December 13th period, four polls were released in Iowa.
Three of the polls demonstrated a weakening in Newt Gingrich’s position, while Insider Advantage illustrated no such decline. IA was, in fact, 12.3% more favorable to Gingrich than the other pollsters.
The same pro-Newt Insider Advantage lean again popped up just after Christmas in Iowa.
Romney was ahead of Gingrich in every other poll, but IA landed Romney in a tie with Gingrich. This tie made Insider Advantage 9.2% more favorable to Gingrich than the average of all the other polls during the same period.
And just like Romney’s 9% turn around in SC, IA found Romney gaining just four days later to take a 7% New Year’s Day Iowa lead over Gingrich. This change in their final poll allowed IA to be among the top in the final pollster accuracy rankings.
How about New Hampshire?
Insider Advantage was mostly silent in the Granite State, but its last poll exhibited the same pattern apparent in Iowa and South Carolina.
All other polls with end dates between December 12th and 19th staked Romney to a double-digit lead, but not IA.
Insider Advantage was the only pollster during this stretch to peg Romney’s support under 30% and was 12.6% more favorable to Gingrich than the average of all other polls ending between December 12th and 19th.
The consistency of these pro-Newt numbers means that it’s not just random statistical fluctuations.
Could it be some constant methodological problem? I doubt it.
Newt Gingrich voters are not more likely than Mitt Romney’s to be subject to wild swings by pollsters with different methodologies. Both Gingrich and Romney voters tend to be older Republicans who all pollsters tend to capture pretty well, and they have been two of the most accurately polled candidates in both Iowa and New Hampshire.
These facts and figures instead lead me back to the bias accusation.
Now, I’m not saying that I know for sure that Insider Advantage polls are purposely biased towards Newt Gingrich, but doesn’t it look awfully strange that their founder’s former boss has been the beneficiary of surveys that are constantly different than the average poll?
Insider Advantage somewhat shady history also lends credence to my suspicions. They have, for example, previously been hesitant in releasing important technical details on how their polls have been conducted even to the newspapers that sponsor their polls.
Insider Advantage has additionally been among the least accurate pollsters over the past ten years. And as in Iowa and South Carolina, they have previously had rapid and probably unrealistic changes in survey data in the week leading up to elections to become more amazingly more accurate in their final surveys.
For me, any one piece of this evidence would not be enough to say Insider Advantage is not a great pollster, but together the mountain of evidence is too high. I just don’t think Insider Advantage polls are worth the press they receive.
Update: See Brice’s figures with this data here.
The emerging narrative from this primary cycle is the battle between “Romney” and “Not Romney.” Just a couple days ago, for example, Nate Silver suggested that Santorum’s path to the nomination required two phenomena:
First, Saturday’s statement would have to get voters to take another look at him at the expense of other conservatives like Mr. Gingrich or Mr. Perry. And second, the negative attacks by Mr. Gingrich would have to succeed in softening up Mr. Romney’s support, but not play to Mr. Gingrich’s direct benefit either, allowing Mr. Santorum to leapfrog both and win with perhaps 25 percent of the vote.
In a way, the narrative argues that the conservative wing of the Republican party suffers from a kind of coordination problem: none of the conservatives want Romney to win, but they cannot quite coordinate their behavior to beat him. Certain candidates have risen but been found unsatisfactory.
Thus, the narrative goes, Romney skates to victory because the conservatives spread their vote so thinly among too many candidates. For anyone (à la Santorum) to win, as Silver describes, a single candidate needs to consolidate the anti-Romney vote.
To what extent is this correct? Nate is correct in his assessment of Santorum’s path forward, but we should not interpret this as affirming the coordination narrative in South Carolina.
The narrative is only correct insofar as the “Not Romney” vote edges out Romney by quite a few points. According to our trend estimates, Romney holds about 30 percent of the vote in South Carolina. By comparison, the sum of support, according to our trend estimates, for Gingrich, Santorum and Perry is almost 42 percent. (I omit Paul because he has a very different base of support, like young and nontraditional Republicans, who would likely not vote for the more conservative candidates, if they voted at all.) From this angle, it looks like the narrative has legs.
There are two major problems, though. The most apparent is that no single candidate could perfectly consolidate all of those “Not Romney” voters. Santorum supporters, for example, split pretty evenly between Perry, Gingrich and Romney if pushed on their second-choice candidates (PPP crosstabs [pdf]).
A second, and less immediately apparent, reason to be skeptical is that even Santorum, Gingrich and Perry combined are losing ground.
True, most of the loss of ground comes from Gingrich tanking. But for Santorum to consolidate the “Not Romney” voters, he would need to be trending upward strongly enough to offset Gingrich’s plummet. He’s not. Even with his slight momentum, the “Not Romney” vote share is still falling (and fast).
After his impending South Carolina win, Romney will effectively end the nomination fight. The narrative of Romney versus “Not Romney” will fade too. For the next few days, however, it would behoove observers to note that even though Santorum is rising in the polls, it does not seem to have much of an effect on Romney’s support, nor does it represent a true consolidation of voters behind a single anything-but-Romney candidate.
On Tuesday Morning, I wrote “Mitt Romney seems destined for a big New Hampshire 15-20% win, and Jon Huntsman looks to come up just short of a second place finish.” The final returns verify that Mitt Romney won by 16.4% and Jon Huntsman did finish in third place.
Did the rest of the results lineup how we expected them to?
Dick Bennett of the American Research Group has done a fine job calculating individual pollster accuracy. Most survey houses did a great job and their results fell within the expected confidence intervals.
Few of Bennett’s findings are surprising and generally fit with the conventional wisdom. He determines that the greatest pollster error occurred with Ron Paul’s projected vote percentage.
If you look at Bennett’s chart, you’ll note the smallest Paul errors were those that polled earlier in the period. Both NBC/Marist and Washington Times/JZ Analytics ended their polls either on the Wednesday or Thursday before the primary. The pollsters with the largest Paul error were all in the field through at least Sunday (January 8th).
How is it that the pollsters who polled later were the least accurate in terms of Paul? Were later polls missing something that earlier ones were not?
The answer is not as clear as we would like it to be, but Stefan Hankin has found that the Public Policy Polling and Suffolk did undercount both young and independent voters. Paul performed best among these groups.
This discovery seems to confirm the pre-primary thoughts of Jeff Winchell before the primary, although Romney’s percentage of the vote was not greatly affected by this factor.
Oddly, NBC/Marist actually had about the same percentage of Independents (38%) and young voters (10%) as later polls. Thus, the only conceivable conclusion that I can reach is that both early and late surveys were missing Paul voters, and that Paul did decline in the lead-up to the election. It’s quite possible then that Paul’s true percentage of the vote was ~26-28% early in the week.
I am not too worried about the polls undershooting Paul’s percentage going forward as the Iowa polls did a fine job gauging his support.
Bennett also finds that those pollsters late in the period were more likely to correctly gauge Huntsman’s share of the vote.
The timeline of the Huntsman error is consistent with the overall feeling that Huntsman was rising in the lead-up to the primary.
But I have to ask whether it is fair to “grade” pollsters who end their polls earlier on the same scale as those who ended their polls closer to the primary? Unlike what we know about general elections, voter opinion in primaries can change quickly. We saw that in Iowa with Rick Santorum and in New Hampshire with Jon Huntsman.
Finally, before I turn my attention to the great performance of the aggregators, I must point out that Bennett finds the least accurate overall pollster was the University of New Hampshire.
It was not that UNH polls were “bad”, but that they were the least great of the great. Indeed, UNH is a fine pollster, and I rely on their surveys for NH Congressional district polling.
I bring attention to UNH only because before the primary all I heard from pundits was how this was the “it” pollster in New Hampshire. The truth is that, on the whole, no pollster is that much better than any other one. Most pollsters (with the notable exception of Zogby Interactive) tend to be equally good, and Tuesday’s results confirm that belief.
As for the aggregate of the aggregates, and the aggregators who make up that aggregate, it was a fine evening.
The average absolute candidate error (1.6%) was down vs. the “difficult” to poll Iowa Caucuses (2.1%). The largest individual candidate error (3.9%) was also down from Iowa (5.2%). The aggregate of the aggregates as well as each aggregator correctly placed Mitt Romney’s first through Rick Perry’s sixth place performance.
This impressive predictive power is not surprising. While most of the public concentrated on the Democratic polling disaster of New Hampshire 2008, most failed to recognize that the polling on the Republican side was fantastic.
Among the individual aggregates, the smallest (statistically insignificant) error overall belonged to my old friends at HuffPollster.
Polls & Votes and Real Clear Politics tied for the smallest (again statistically insignificant) largest error. All but Polls & Votes with Rick Santorum had their largest one candidate error with Ron Paul.
Overall, the pollster’s graveyard did not receive a new body this past Tuesday. Along with Mitt Romney, polling improved upon its Iowa performance. The only question left to answer is whether the polls can maintain their momentum and make it three-for-three in South Carolina.
I have a piece over at The Guardian on the relationship between the growth of the economy and President Obama’s chance of re-election. Check it out.
by Harry J. Enten Filed under:
Elections Forecasting
Friends at my alma mater will be casting their vote in the first in the nation primary later today. New Hampshire, polls would have you believe, is Romney country. The question it would seem is by how much. The media will play the expectation game and anything less than a sizable 15%+ victory for Romney is likely to be viewed as a disappointment.
Many in the press worry that the polls will repeat their 2008 New Hampshire Democratic Primary performance, but many of these same people forget that the 2008 New Hampshire Republican Primary was very well predicted. Indeed, most of the aggregates all point in a similar direction. They all have Romney winning with near 40% of the vote and by about 20%.
In fact, exactly like Iowa, all the aggregates have the order of the finishers identical. Mitt Romney is first, Ron Paul is second, and Jon Huntsman is third.
This is not to say that there isn’t some movement in the polls. All the aggregates have shown a decay in Mitt Romney’s support. If the Public Policy Polling poll is right, it’s possible that Romney could end up with less than 35%. This will almost definitely be spun by the media, even though Romney would win by a sizable margin.
Some analysts have argued that recent polls may be undercounting independents, men, and younger voters. These are all groups among Mitt Romney performs worse, and Ron Paul performs best. One, Jeff Winchell, believes Romney might only win by ~10% with Paul ending up in the mid 20s.
Count me as somewhat skeptical as pollsters did a very good job of gauging both of these candidate votes in Iowa, but it something to keep in mind.
The candidate most likely to surprise is Jon Huntsman, as polls often do not fully catch the rise of the “momentum” candidate. That said,only one poll has him overtaking Ron Paul for third. Huntsman will not be the Rick Santorum of New Hampshire, but he has been gaining almost every day. If he were ever to get 20% of the vote tonight, Huntsman probably will deem himself the “Comeback Kid”.
The rest of the candidates seem pretty stable in support. Newt Gingrich and Rick Santorum have seen that New Hampshirites were not open to their campaigns, although Santorum seems in a slightly better position than Gingrich. Their lack of a support is of no surprise to this analyst, but does not bode well for them heading into South Carolina.
Rick Perry is stable at 1%, and the only question seems to be whether Buddy Roemer will get a higher percentage of the vote than Perry.
The fact remains, like Iowa, New Hampshire has been polled until the very last minute. All the aggregates are pretty much identical, which gives me confidence that there will not be too great of a surprise. Polls in primaries are not perfect, and any one candidate may exceed or fail to meet their polling numbers.
Still, Mitt Romney seems destined for a big New Hampshire 15-20% win, and Jon Huntsman looks to come up just short of a second place finish.
On Monday Evening, I wrote “It’s still tight, and polls have gotten primary/caucus results wrong before. But when Mitt Romney says ‘We’re going to win this thing’, I gotta say the numbers agree with him.”
Though Romney only won by 8 votes, the fact is the polls correctly predicted the winner of this difficult to poll contest. But how “right” were the polls?
American Research Group’s (ARG) Dick Bennett has already done a solid review of individual polls. As he notes, most polls did quite well, and the results of them fell within their expected 95% confidence intervals.
If I take one issue with Bennett’s analysis, it’s that he examines the four-day sample of the Des Moines Register’s Selzer & Co. poll instead of the final two-day sample.
As I have previously noted, the publication of a CNN poll during the Selzer & Co. time in the field changed the dynamic of the race by aiding a Rick Santorum surge.
It was for this reason that the Des Moines Register published two sets of results: one for the entire 12/27-30 period and one for the 12/29-30 period. I, believing it was the more accurate representation of the race, used only the 12/29-30 poll for my analysis.
The 12/29-30 Selzer & Co. poll found Mitt Romney leading with 24%, Rick Santorum in second with 21%, and Ron Paul in third with 18%. This poll was the only one to correctly forecast first, second, and third place. It was the most accurate in predicting the spread between Santorum and Paul, and second most accurate in estimating the spread between Romney and Santorum.
Overall, the 12/29-30 Selzer & Co. poll was the “most accurate” Iowa poll employing ARG’s Martin, Traugott, and Kennedy measure of pollster accuracy. This is not to say that Selzer & Co.’s full four-day (12/27-30) sample should not also be scored. The fact is that it was published as the “main” poll, but I think it’s necessary to point out that the two-day sample was quite accurate.
What about the polling aggregates?
Back on December 27th, I wrote “the polls are more likely to get a lot more right than wrong”, and that 2012 would compete with 2008 for being the most accurately polled Iowa Caucus. In 2008, “the average difference between the final polling average and candidates’ actual percentage was only 2.7%. The biggest difference in percentage [for any one candidate was] 5%.”
As you can see in this table, the aggregate polling average for the 2012 Iowa Caucus was among the most accurate ever.
The largest error 6.2%, Rick Santorum’s, was slightly worse than 2008′s 5% error largest error. This largest error was half that of every year besides 2008. Furthermore, the average candidate error of 2.3% in 2012 was the smallest ever. Needless to say, the poll aggregates on average did very well.
What about the different averages themselves? As the table below illustrates, all had near identical average candidate errors*.
Though not statistically significant, 538′s largest error* (Rick Santorum as it was for all the aggregates) was the smallest at 5.2%. This makes sense as Santorum was surging, and 538′s model had a built-in momentum catcher.
But let’s bring it back one minute here.
Nate Silver came in with a very last minute altered projection in which Rick Santorum placed first, Ron Paul second, and Mitt Romney in third. If 538′s forecast had been this call, the order would have been incorrect, but the average error per candidate would have actually been only 2.2%. 538′s largest error would have been Mitt Romney’s 5.5%.
More interestingly, if our aggregate of the aggregates had included Silver’s late projection, it would have been slightly more accurate.
The average error per candidate would have only been 2.1%. The largest error for any one candidate would have been only 5.2% (for Rick Santorum). Thus, even if Silver second call was wrong, it still added some valuable information to our thinking that his original call did not.
But enough with Iowa, now it’s onto New Hampshire. While New Hampshire was the site of arguably the polling catastrophe of the 21st century, Iowa’s strong polling performance gives me confidence that history will not be repeated.
*Note, Selzer & Co.’s two-day sample had only a 1.5% average error and 2.6% largest error (Michele Bachmann), which beats all the aggregates.
Just a few moments ago, Jon Huntsman announced that he had secured the Boston Globe’s endorsement. Given that Huntsman has tried to turn New Hampshire into his firewall for the nomination, he must have a new spring in his step tonight. How have his chances changed, though? Probably not much.
Huntsman is trending in the wrong direction. In the past few days, Huntsman’s poll numbers have been dropping, not rising as one would expect as the race tightens. Consider the following graph from Charles Franklin at Polls and Votes:
Could the Boston Globe endorsement change the Huntsman momentum in the positive direction? Sure. The Globe endorsed John McCain in 2008 at a time when Mitt Romney dominated the polls. After the endorsement, however, McCain rises precipitously in surveys and eventually wins the primary.
We must be cautious, however, in drawing too close a comparison. Consider, first, that the Globe endorsement came much earlier in the race, giving McCain three weeks to overtake Romney. Second, McCain had a history with New Hampshire, having won the primary in 2000. Third, McCain was already trending upward at the time of the endorsement. Fourth, Romney had spent a lot of time and money in Iowa, and was greatly weakened when he lost to Mike Huckabee.
Huntsman faces a much more daunting challenge. The former Utah governor needs to turn his polls around quickly and with little cash on hand. He also needs to break through the media narrative of Santorum’s surge and Romney’s intractable strength. While the Globe’s endorsement doesn’t hurt Huntsman’s chances, I can’t see a scenario where it helps him enough to make much of a difference.
What would a bump look like, if it comes? Given similarities in the profiles of Romney and Huntsman, a bump in Huntsman’s numbers would probably need to come at Romney’s expense, which seems unlikely. My colleague Harry Enten pointed out earlier today that Romney voters are not fickle. As Harry argued, Romney leads the field “with 59% of the vote among those in the latest Suffolk Poll who said they were unlikely to change their mind or were sure about their voting choice.”
Perhaps Huntsman can build a coalition amongst undecided voters. The latest Suffolk Poll shows that 17 percent of respondents are undecided, and 57 percent of those are liberal-to-moderate. If Huntsman wins over all of those voters, with the help of the Globe’s endorsement, Huntsman could add almost 10 percent to his current showing of about 7 percent. In that rosy scenario, Huntsman would still only pull about 17 percent of the vote, still possibly insufficient to top Ron Paul (FiveThirtyEight projects Paul to take 24 percent), much less Romney (43 percent). And that ignores the fact that Huntsman is really in a second-tier tussle with Santorum and Gingrich; if one of these guys manages to come close to Huntsman, it dampens the upside of Huntsman beating expectations.
Even if Huntsman bounces up to second place, it’s hard to see how he translates that into momentum beyond the Granite State. Huntsman will not play well in South Carolina and probably not in Florida, meaning that a still-weak Huntsman would need to hold out until early February and hope for a big showing in Nevada and Maine.
It’s not impossible that Huntsman could still make a sizable impact on the nominating process. But if anything is going to stop Romney, it seems like it will be a conservative alternative, not a moderate Romney look-a-like who, in the best scenario, will take second in New Hampshire.
[Crossposted at my personal blog.]
The question dominating political analysis right now is whether Mitt Romney can hold onto his large New Hampshire lead or will he falter to a charging Rick Santorum. Santorum, after all, is coming off a very strong second place showing in the Iowa Caucus, and his northern (somewhat Pat Buchananesque) populism might take hold with New Hampshire voters.
Count me as a skeptic for the following five reasons.
1. Mitt Romney voters are solid Mitt Romney voters. Mitt Romney lead the field with 59% of the vote among those in the latest Suffolk Poll who said they were unlikely to change their mind or were sure about their voting choice.
Put another way, 38% of the entire primary electorate says their choice is Romney, and they are solid on that choice. No other candidate comes close to that percentage, and 38% is almost certainly enough to win the New Hampshire primary.
2. The truly undecided voters in New Hampshire are overwhelmingly liberal or moderate. There can be little doubt that Rick Santorum runs best with conservative voters. He won them in Iowa, but he garnered only 8% of the vote among liberals and moderates.
Problem is that among the 17% of undecided voters in New Hampshire in the latest Suffolk Poll, 57% are moderate or liberal. This number does not leave Santorum a lot of room to catch up.
3. It’s not like Ron Paul is going away. Lots of pundits like to discredit Ron Paul, but the truth is that he had a fairly strong performance in Iowa. Paul is polling at ~20% in New Hampshire and that is a large chunk of the electorate.
Early (and I stress early) signs from the Suffolk tracking poll is that he has also received a bounce out of Iowa. If Rick Santorum is hoping to capitalize on newly found viability, it seems he might have to compete with Paul for it in New Hampshire.
4. There just isn’t that much anti-Romney, pro-conservative vote in New Hampshire. Mark Blumenthal put it best when he said “even if every Bachmann, Perry and Gingrich supporter switched to Santorum, his support in the Granite State would only total roughly 20%”.
5. Historically, surges in New Hampshire require two parts: the underdog overperforming, and the favorite underperforming in Iowa. Whether it was Gary Hart and Fritz Mondale in 1984, Bob Dole and Pat Buchanan in 1996, or Howard Dean and John Kerry in 2004, the front-runner was perceived as doing worse than expected.
There is no doubt that Rick Santorum did better than the polls indicated in Iowa, but here’s a little secret… so did Mitt Romney by a little. Compared to the final polling averages, Romney did ~1.5% better than the final projections. Santorum certainly exceeded that doing ~6% better, but it’s not like his finish was shocking.
If I were to estimate (and keep in mind, it’s only an estimate) where Rick Santorum will end up in New Hampshire, I would peg it at ~15%. That’s actually slightly better than Nate Silver’s historical Iowa momentum into New Hampshire model would indicate based on the Iowa results vs. pre-Iowa polling in Iowa and New Hampshire.
Mitt Romney and Ron Paul are solid, and Newt Gingrich will hold onto some of his vote. This leaves very little room for a Santorum surge.
