January 2008 Archives
A few weeks back I posted a datapoint from Iowa showing that Ron Paul was getting a disproportionate share of those voters who are "Angry" at President Bush (as opposed to merely dissatisfied, or positive). One of John McCain's remarkable feats in Florida was to pull from both those who were positive about the Bush administration AND those who were negative:
In Florida, 7% of voters in the Republican Primary indicated that they were "Angry" with the Bush Adminstration. Who are these angry voters in the sunshine state? Here are some interesting highlights:
- 76% Believe abortion should be "legal," with 30% saying "legal in all cases."
- Angry voters tended to be older (50% were age 60 and over) males (62%) who identified themselves as "Independent." (49%)
- They favor candidates who "say what they believe" (39%) over candidates who "share their values." (14%)
- Unlike most Republicans in Florida, they would vastly prefer paying down the deficit (73%) to cutting taxes (26%).
- They think the economy is poor (59%) and by far the most important issue (51%). They aren't too worried about Terrorism, however (3%).
- Though they are concerned about the economy, they are a bit more well off than their more "satisfied" brethren--44% have household incomes greater than $100,000.
A great breakdown of some of our exit polling data from the recent Florida Primaries can be found in Marc Caputo's article in the Miami Herald from Wednesday. Caputo notes the impact of Governor Crist's endorsement of McCain, and also the three-way split of the evangelical vote in the Republican Primary (more on that in a future post).
Most of the post-exit poll analysis focuses on the big issues, and in South Carolina the primary issue was still the economy. Hundreds of thousands of South Carolinians are feeling the pinch on their pocketbooks, and over half of the voters we surveyed in last Saturday's Democratic Primary indicated that the state of our economy was the most important issue on the minds of these voters.
Still, change occurs at the margin--and even when an issue such as the war in Iraq gets pushed down the page, there are still plenty of people passionate about our continued presence in the Middle East. In South Carolina, 19% of voters still felt that Iraq was the most important issue. These voters closely resembled the rest of the sample in most areas--they weren't any richer, certainly--save one: their proposed solution to the Iraq problem. In one of the only areas of inquiry that this 19% markedly differed from the total sample, these voters' preferred solution continues to be immediate withdrawal:
"Only 19%," but a loud minority with a clear difference from their fellow Democrats.
With President Clinton becoming more of a presence in Senator Clinton's campaign, I though it worth revisiting this datapoint from the New Hampshire Primaries. We asked New Hampshire voters if they would still have chosen their candidate if Bill Clinton were eligible for a third term and on the ballot, and here is what we found:
Clearly, Obama's supporters are strongly behind their candidate, with Edwards' voters only marginally less so. Hillary Clinton, however, would lose handily to her husband, which might lead one to conclude that many of her supporters believe a vote for Senator Clinton is indeed a vote for a third Clinton Term. Yet, amongst all New Hampshire Democratic Primary voters, Bill was able to muster only 37% of this hypothetical vote. "Bill's Third Term" has certainly proven to be a polarizing issue not only for Republicans, but many Democrats, as well.
An interesting study in momentum: one of the questions we ask in the exit polls is "When did you finally decide whom to vote for in the presidential primary?" As this graph shows, there were plenty of minds made up in the last three days prior to the contest. Reports of the Fred Thompson "surge" appear to have been justified, as much of his support did in fact come within a few days of the primary. Huckabee clearly showed an Iowa "bump," followed by a New Hampshire "lull" and ending with some hard-fought campaigning in the last days. Showing the biggest "day of" bump was Mitt Romney, who had earlier that same day won Nevada handily, which may have given him a slight lift on primary day in South Carolina. Finally, Most of Ron Paul's supporters decided to vote for Paul some time ago; he suffered the most in terms of recent converts.
Last week, I noted that the Edison exit polls indicate evidence of a generational split among Michigan's African-American primary voters in their support for the Democratic candidates; younger African-American voters went decidedly for 'uncommitted' as a probable proxy for Barack Obama, while older African-American voters were somewhat more inclined to support Hillary Clinton. It will be interesting to see if this pattern also holds in the upcoming South Carolina Democratic primary on Saturday.
The exit poll data from this past Saturday indicates that a similar pattern of generational differences appears to have characterized the voting patterns of the self-identified born-again Christians in the South Carolina Republican primary. Older born-again Christian voters (i.e., those over 60 years of age) accounted for John McCain's margin of victory over Mike Huckabee in the South Carolina Republican primary election.
As everyone expected, born-again Christians played a leading role in South Carolina's Republican primary; 60% of the voters interviewed in the exit polls identified themselves as born-again Christians. This is a far larger proportion of South Carolina's Republican primary voters than in the states with the earlier Republican primaries (e.g., 23% in New Hampshire and 39% in Michigan), though it equals that of Iowa caucus goers (60%). Of the South Carolina Republican primary voters, Huckabee won the largest share (43%) of the born-again Christians, though McCain also managed to attract more than a quarter (27%) of these voters.
Interestingly, however, an analysis of the age distribution of the born-again Christian voters in Saturday's South Carolina Republican primary indicates that nearly half of those voters age 18-29 years (49%) and those whose age is 30-59 years (44%) voted for Gov. Huckabee, while only about 1 in 5 (19% and 21%, respectively) in these age groups supported Sen. McCain. Among those age 60 years and older, however, each of the candidates received an identical level of support (39%). This is important, given that nearly a third (32%) of the born-again Christian voters in the South Carolina Republican primary were age 60 or over.
So what are we to make of this? Well, 60% were born-again, 32% of these are older, and McCain did much better among this age group (39%) than he did among the younger age groups (about 20%). Had he done only as well among the older born-again Christians as he did among the younger, the older born-again Christians would have provided him (.6 * .32 * .20 =) 3.8% of the South Carolina Republican primary vote. Instead, the older born-again Christians gave McCain a margin similar to what they gave Huckabee, resulting in far larger portion of the South Carolina Republican primary vote: (.6 * .32 *.39 =) 7.5%. The difference between these (7.5% - 3.8% = 3.7%) is just slightly more than John McCain's margin of victory over Mike Huckabee in the South Carolina Republican primary. So it appears that John McCain can thank older born-again Christians for his margin of victory in the South Carolina Republican primary election.
Mitt Romney's win over John McCain and Mike Huckabee in Tuesday's Michigan Republican primary election received the most attention in the press. The Democratic primary election was largely overlooked, since it was uncontested--Barack Obama and John Edwards had requested that their names be removed from the Democratic party's ballot. Hillary Clinton's name was the only one of the three leading Democratic candidates to remain on the Michigan primary ballot.
Even though Obama's and Edwards' names did not appear on the Democratic ballot, Michigan's Democratic voters still had an opportunity to express, if only indirectly, their support for their candidates: by voting "uncommitted." Prior to the Michigan primaries, Representative John Conyers (the senior African-American congress member from Michigan's fourteenth district) sponsored radio advertisements asking Michigan supporters of Barack Obama to vote uncommitted in the Democratic primary. The strategy appears to have met with some success. According to the Edison exit polls, 23% of those who voted in Michigan's Democratic primary were of African-American heritage, and 68% of them reported voting "uncommitted." The exit polls indicate that Clinton received only 30% of the Black vote in Michigan's Democratic primary.
If the uncommitted vote among the African-Americans in Michigan reflects a preference for Obama, then Michigan may be a harbinger of the upcoming South Carolina Democratic primary on January 26. The Michigan exit poll numbers indicate that Clinton and "uncommitted" drew disproportionately from across the generations of African-American voters. The exit polls indicate that nearly all of the youngest (ages 18-29) African American voters cast their vote for "uncommitted"; nearly one in eight (13%) of the uncommitted African-American vote came from this age group, while this age group accounted for less than 1% of Clinton's African-American support. Clinton, on the other hand, drew disproportionately from middle-aged African-Americans (ages 30-59) with 87% (vs. 69% for uncommitted) of her African-American support from this age group. Finally, the exit polls indicate that both the Clinton and uncommitted vote among African-Americans came about equally (13% vs. 18%, respectively) from the older age group (age 60 and older).
Although it is not crystal clear, since Mr. Obama's name was not on the Michigan Democratic primary ballot, the Michigan exit poll data indicate that it is at least conceivable that there is a generational split in the African-American community. Younger African-American voters appear to have strongly favored Obama, while Clinton drew her support disproportionately from among middle aged African-American voters.
In the run-up to New Hampshire one of the big questions was whether independents would vote in the Republican or Democratic primary. In South Carolina, we have the same question, but with a twist: The two primaries there are one week apart. The Republicans go first (January 19th) and then the Democrats go one week later (January 26th).
The key thing to note is that one can vote in one election, but not both. They don't have party registration in South Carolina, so one can still choose which election to vote in, but the state board of elections is running both primaries so it will be impossible to vote twice.
Thus, it will be interesting to see how many voters the Republican tilt essentially 'takes off the table' for the Democratic primary. Will independents take themselves out of the Democratic race by voting in the earlier race, since it is getting so much attention now?
One of the key factors to watch will be the total turnout for the Republicans. If it is significantly higher than the previous record of 573,000 in 2000, then indeed there might simply be fewer bodies available to vote one week later.
And so who would a large independent turnout in the Republican primary help on the Democratic side? With Obama now expected to decisively carry the African-American vote, does this make it harder on Clinton to take the state with fewer white voters to come her way? Or does it take white independents, of the type who went for Obama in New Hampshire and Iowa, away from his count and allow Clinton or Edwards to squeak by him?
It will be interesting to watch and of course we will be helping to answer all these questions with our Exit Polling data.
Much has already been said about the "Uncommitted" vote in the recent Michigan Democratic Primary. Almost a quarter of a million Obama and Edwards supporters braved the lousy weather on the day of the Michigan Primary just to vote "Uncommitted." The Detroit Free Press, for instance, was one of many sources pointing out that when asked whom they would choose if all the candidates were on the ballot, 35% of Primary voters would have chosen Obama, compared to 46% for Clinton. Edwards would have finished third, at 12%. This is the "Hidden Election" of the last week - a glimpse revealed by the Michigan exit poll into what might have happened in Michigan if the Democratic Party had not taken away all of Michigan's delegates for scheduling their primary before February 5th.
Given the significant number of Obama and Edwards supporters who did brave the weather to cast their vote for...well...no one, it is worth a deeper look at this "Hidden Election." If we recast the data using the "who would you choose if all the candidates were on the ballot" question, we do see some parallels between Michigan and the previous contests in New Hampshire and Iowa:
- Obama would have beaten Clinton with 18-29 year-olds and 30-44 year olds. Clinton continues to win 45+.
- Continuing another pattern from prior contests, Obama would have won with college graduates, while Clinton would have won with lower income voters. Obama also would have tied Clinton with voters living in households earning more than $50,000 per year.
- Independents would have voted for Obama, while, again, Clinton wins with Democrats. Women would still have voted overwhelmingly for Clinton; however she would have split the male vote with Obama.
- 55% of Rural/Small City voters would have voted for Clinton, compared to 22% for Obama. But Obama would have taken 55% of the vote in cities with populations greater than 50K.
- Finally, in the most "hidden" data point of this hidden election, 20% of the voters who actually did vote for Clinton would have chosen another candidate if all were available (with 11% of the total actual Clinton Vote going to Obama.)
All of these points from the exit poll should be taken with a grain of salt since the Democratic Primary turnout was less than would have been normally seen in a competitive Democratic Primary. Just under 600,000 people voted in the Michigan Democratic Primary on Tuesday, compared with more than one million who voted in the last competitive Democratic Primary in Michigan - the 2002 Democratic Primary for Governor.
But even with that caveat the results for the "hidden election" question are comparable to the horserace numbers that we have seen in recent post-New Hampshire national polling. A Gallup poll released the day before the Michigan Primary showed a 12-point Clinton national lead over Obama - 45 to 33 with 13 percent for Edwards. The "hidden election" in Michigan was very similar --46% for Clinton, 35% for Obama and 12% for Edwards -- and may be an indicator of how the Democratic primaries may look when we get to the many big states voting on February 5th.
Farhad Manjoo offers an in-depth analysis of the facts surrounding the voting disparities between precincts in New Hampshire. Note especially the section entitled "Can exit polls tell us whether fraud occurred?" We have said in this space and elsewhere that it should not be surprising to see voting disparities between precincts--that is why we use the precinct model, after all--and that correlation does not imply causality.
Kudos to Salon for a balanced and insightful take on this issue.
So far we have asked Republican voters in three states what they felt was the most important issue of this year's election. The results over time present a different take on "consumer confidence:"
Admittedly, Michigan may be a bit of an outlier due to the condition of their state's economy, but if this trend continues (pushing Iraq and Immigration further down the list in terms of priority) this may favor the fortunes of executives (such as Romney, Giuliani or perhaps even Bloomberg) over legislators. Certainly, the issues the candidates will be talking about over the coming months will increasingly turn to matters domestic if this trend line continues in Nevada and South Carolina.
From Frank Rich's column in last Saturday's New York Times:
The exit poll of those who voted on Tuesday -- not to be confused with the pre-primary polls that misfired -- showed that Democrats are still looking for change (54 percent) over experience (19 percent) and that they overwhelmingly associate Mr. Obama with the former and Mrs. Clinton with the latter. By change, they don't mean merely a tuneup.Interestingly, though the overall percentage of voters wanting "Change" increased from Iowa to New Hampshire, this small increase was driven entirely by Independent voters. In fact, the percentage of persons identifying themselves as Democrats that indicated "Change" was the most important issue to them in selecting a candidate stayed statistically flat, with 52% of Iowa Democrats favoring "Change," compared to 50% of New Hampshire Democrats. Independents, however, rose from 53% favoring change in Iowa to 62% in New Hampshire. As has been noted in this space and elsewhere, the story of the primary season continues to be turnout--will the Independents (and Democrats) who are most desirous of change come out and vote with their feet? We'll have another piece of this puzzle for you this weekend, in Nevada.
In his take on what happened to the New Hampshire pre-election polls, Marc Ambinder gives us a theory of a different sort, and also conjures up our first Bea Arthur reference of the 2008 election.
Jennifer Agiesta and Jon Cohen ran a blog post in yesterday's Washington Post about the current conspiracy theory that ballots counted by optical scanning machines gave Senator Clinton the win in the New Hampshire Democratic Primary, and that hand-counted ballots actually favored Obama. Frankly, given the socio-economic and psychographic differences between the various precincts that use one method or the other, the only "suspicious" finding would be if there were no differential at all. In fact, there are loads of differences one could imagine between, say, a large urban precinct that requires optical scanning just to get through the night, and a small rural precinct whose handful of votes can be hand-counted in just a few minutes.
Our own Joe Lenski looked into the history of these precincts, and found that the differences between the two types of precincts is not new, and not unique to 2008--they are simply different sorts of precincts. As Joe himself noted in the Agiesta/Cohen piece, "unless there has been hidden election fraud in New Hampshire for the last three presidential primaries the 'evidence' being used by these fraudsters probably does not hold up to any rigorous statistical analysis." Kudos to Jennifer and Jon for digging into the numbers.
The nearly final turnout numbers from New Hampshire show 287,849 total voters in the Democratic primary and 238,909 total voters in the Republican primary. The Democratic Primary turnout beats the 2004 record by almost 70,000 and the Republican Primary turnout just edged the Republican turnout in 2000 by a few hundred votes. Combined, it beats the previous presidential primary (2000) record by more than 130,000. In fact, the 526,758 total voters beat what you typically see in an off-year general election for governor in New Hampshire (417,000 in 2006 and 453,000 in 2002). It even came close to the 578,638 total voters in the 2000 Bush-Gore general election, so it is kind of phenomenal.
So how did this extraordinary turnout break down by the party registration of the voters? As most followers of the New Hampshire Primary know, those who are registered Undeclared in New Hampshire have the choice to vote in either primary. New Hampshire also is one of a handful of states which allow same day registration on election day. The Edison exit polls show that the Democratic Primary consisted of 52% Registered Democrats, 42% Registered Undeclareds and 6% Election Day Registrants; the Republican Primary consisted of 61% Registered Republicans, 34% Registered Undeclareds and 5% Election Day Registrants.
Applying those numbers to the final turnout you get approximately the following picture of the combined electorate:
So in the end, roughly the same number of Registered Democrats and Registered Republicans participated in the primary even though there are slightly more Registered Republicans (256,353) than Registered Democrats (221,549) in the entire state.
But the key statistic is that by a 60-40 margin Undeclareds and Election Day Registrants chose to vote in the Democratic primary over the Republican primary. This margin resembles the 57D 39R split in the vote for U.S. House among Independents that we measured in our national exit poll in 2006. A continuation of this 60-40 split among Independents is a good sign for the Democrats as the Presidential election year unfolds.
One of the pre-election pollsters' most daunting tasks is to divine a model for the "likely voter." Since most respondents in a pre-election telephone survey will report that they intend to vote, pollsters use a series of questions that help them estimate which members of their telephone samples are actually likely to show up at the polls, and which respondents are likely to demur. Like grandma's favorite recipes, survey organizations typically keep their "likely voter model" a closely held secret.
In the pre-election polls for the recent New Hampshire primary, the likely voter models for the Republican primary performed well, though the likely voter model for the Democratic primary performed poorly. The pre-election estimates for the Republican candidates were derived from models that were (presumably) based on the 2000 Republican presidential primary. For both Republican primaries, the voter turnout was remarkably similar (238,606 in 2000 and 238,909 in 2008), so the pre-election polls' likely voter model worked well, and their predictions were reasonably accurate.
The voter turnout for the Democratic primary, on the
other hand, was unprecedented in at least two ways: the number of participants
and the proportion of registered Democrats who showed up to vote. In 2004, a
(then) record of 219,787 voters turned out to vote--the previous record for the
Democratic primary was in 1992, when 167, 819 voters participated. This year, a
record shattering 287,849 voters participated in the New Hampshire Democratic
primary--including nearly two thirds (66.3%) of the state's registered Democrats
(up from 43.3% in 2004). Simply stated, the 2008
I will leave it to the pundits to speculate as to whether or not Senator Clinton's emotional moment prior to the primary contributed to her "surprising" victory, but there are at least two hard differences between the 2008 data and the 2004 data that truly tell the tale. The pre-election opinion polls (which we do not do) all had Obama winning, and Edwards in third. While these polls were actually generally in agreement with the final results in terms of the Obama/Edwards numbers, they were dramatically off with respect to Clinton's final tally. The story in NH, as in Iowa, was one of turnout.
We will have a lot more detailed analysis on the turnout issue very soon in this space. For now, here are two big differences between 2004 and 2008: In 2004, the NH Democratic primary turnout was 54% Female, while this year it was 57%--and Clinton was very strong with women. Also, and perhaps even more significant, while the 2004 Democratic sample was 45% Democrat (and 45% Independent), in 2008 those numbers were 52% Democrat and 42% Independent. With the 2004 Republican Primary not as competitive as the 2008 contest, a higher proportion of Independent voters actually participated in the Democratic primary four years ago. Thus, while all the talk of these first two state contests has been the number of Independents mobilized to vote, the NH story is that registered Democrats supported Clinton by an 11 percentage point margin over Obama, and lots more registered Democrats (as a proportion) turned out this year than they did four years ago. The turnout issue is a deep one, with many implications for the pre-election polls--watch this space for more on this issue in the days to come.
Much has been made of the number of Democratic caucus-goers who made their decision in Iowa based upon which candidate "can bring about needed change" (52%), as opposed to "has the right experience" (20%) or "cares about people like me (19%). As has been reported here and elsewhere, Obama got the "change" vote, Clinton the "experience" vote and Edwards the "cares" vote. In fact, amongst the 19% who valued "cares about people like me," Edwards led with 44%.
What was fascinating about this data was the distribution by income. Amongst Iowa caucus-goers with greater than $100,000 in annual household income, the percentage of voters seeing a candidate who "cares about people like me" was 15%, and a whopping 65% of those affluent, care-seeking voters went for Edwards (with 18% for Obama, and 10% for Clinton). Interestingly, though fighting for the Middle Class has been the dominant theme of Edwards' populist campaign, it was Iowa's wealthy who claimed they chose Edwards for caring for people like them!
In her blog called "The Trail" Ann Kornblut of the Washington Post reported the following:
- After the public rally, the Clintons held a private pep talk with their staff members on the floor of their hotel. Former president Bill Clinton told downcast aides that there was some good news in the night's results: that the campaign had gotten far more than 70,000 voters -- far more than the campaign thought it needed, he said -- and that he is confident a post-caucus analysis will show that Clinton and Obama were tied among first-time voters.
Well, in fact, this is not the case.
Among voters in the Iowa Democratic Caucus who were participating in their first caucus, Obama crushed Senator Hillary Clinton 41% to 29%, according to our entrance polls.
In addition, former Senator John Edwards won among those who had caucused before. Clinton came in third among this group.
The answer is yes.
We looked at the two major-party (or most major) candidates in every contested election. You will recall that the 1789, 1792, and 1820 elections were effectively unopposed -- and there is the wild 1836 election where the Whigs ran regional candidates against Martin Van Buren -- we went with William Henry Harrison as the comparison here.
The biggest previous differential is the 1996 Dole-Clinton race -- Senator Dole is 23 years senior to Bill Clinton. Two other races had nearly as large gaps, the 1992 election between Clinton and George H. W. Bush had a 22 year difference, and the 1856 election between James Buchanan and John C. Fremont (at left) also had a 22 year spread.
Barack Obama will be 47 on election day 2008, and John McCain will be 72. Their 25 year gap would set a record.
As a side note, there appears to be no pattern to whether the younger or older candidate wins.
As a researcher, I love elegant information displays--nothing can truly bring to life--or obscure--important data like a compelling graphic. Edward Tufte has more on this to say than I ever could, but the New York Times posted a graphic of our Iowa data that I just had to share with you:
This graph, (taken from this story from today's New York Times) clearly distills the essence of Huckabee's surge in Iowa. Again, as a researcher, I love the way our data has been presented here--but as a marketer, I notice something else. Politicos will focus on the size of the "Shares My Values" bubble for Huckabee (and rightly so) as the story behind his Iowa success. What strikes me, however, is not so much the "values" aspect specifically, but the fact that he has any disproportionately larger bubble at all--look at Fred Thompson, for instance--his "bubbles" are all fairly even, which means that he didn't win the battle for any one position, but was a little bit of everything. Huckabee found a ladder that he could sit atop and successfully claimed it. Of course, step one in this process (as in any strategic marketing exercise) is determining which ladders are worth owning, and which ones are not as important in the eyes of your consumers--or voters. Still, being known for something is better than owning nothing in particular.
On Jan 3, Edison polled almost 4,000 caucus goers in Iowa to determine their candidate choice and the reasons for their preference. Our results showed wins for Barak Obama and Mike Huckabee along with a record turnout for Iowa. Edison entrance polls were used to determine key information about caucus goers, including demographic data, important issues and (in the case of the Democratic caucus) the second choice candidates. Edison's election team captured, processed and analyzed thousands of data points within the short duration of the caucuses and enabled our member clients and subscribers real-time access to in-depth analysis of the Iowa results.
Iowa Caucus Results
| Iowa Republican Caucus Straw Poll Results | Percentage |
| Huckabee | 34.29% |
| Romney | 25.32% |
| Thompson | 13.37% |
| McCain | 13.13% |
| Paul | 9.99% |
| Guiliani | 3.46% |
| Hunter | 0.44% |
| Tancredo |
| Iowa Democratic Caucus State Delegate Equivalents Results | Percentage |
| Obama | 37.58% |
| Edwards | 29.75% |
| Clinton | 29.47% |
| Richardson | 2.11% |
| Biden | 0.93% |
| Uncommitted | 0.14% |
| Dodd | 0.02% |
| Gravel | |
| Kucinich |
Caucus Entrance Poll Results
For more analysis and results from the Iowa caucuses, please visit our member sites:
http://www.cnn.com/ELECTION/2008/primaries/results/epolls/index.html#IADEM
http://www.cnn.com/ELECTION/2008/primaries/results/epolls/index.html#IAREP
http://www.msnbc.msn.com/id/21225980
http://www.msnbc.msn.com/id/21228177
http://www.abcnews.go.com/PollingUnit/Vote2008/story?id=4084499&page=1
http://www.cbsnews.com/stories/2008/01/03/politics/main3673124.shtml
http://www.cbsnews.com/stories/2008/01/03/politics/main3673260.shtml
Republican Party actual vote results can be found at http://www.iowagop.net, while Iowa Democratic Party actual vote results are located at http://www.iowacaucusresults.com/.
No, it isn't a 60's revival band. One of the more interesting stats from the Iowa Republican entrance polls was the metric of voters' attitudes about the Bush administration. On balance, most Iowa Republicans in the sample were favorable towards Bush, with more than two-thirds indicating that they were at least "satisfied," or "enthusiastic" about the current administration.
However, amongst the dissatisfied, and in particular, the "angry," there was an interesting datapoint. While support from those who were favorable towards Bush was distributed amongst all the candidates, 54% of the "Angry" voters supported Ron Paul in Iowa. This may not be the Republican equivalent of the "Change" vote we observed in the Democratic caucus, but it is a pretty good barometer of the motivations of the voters who helped Paul finish with 10% of the Iowa vote. If 5% of Iowa's Republican caucus-goers are "angry" with the Bush administration, knowing how "angry" the participants in New Hampshire's Republican Primary are may be a key to anticipating Paul's strength next Tuesday night. Based upon the Edison Media Research Iowa entrance poll data, for Ron Paul--the angrier, the better.
Now that the primary/caucus season is fully underway, we here at Edison Media Research have lots to talk about! With Iowa under our belts, and more to come in this compressed season, we have lots of interesting factoids, datapoints and anecdotes about the exit (and entrance) polling process that we will be sharing in this space. We hope you enjoy Politics By The Numbers not only as a "journal of election data and analysis" (as advertised!) but also as a rare glimpse into the nuts and bolts of the arcane science of exit polling itself, and all of the complications, considerations and great care that we put into our work.
We are fortunate to be the sole source of exit polling data for some of the smartest folks in network news, the AP and print journalism, and we look forward to seeing our data used in all sorts of ways over the coming months. We won't try to duplicate their efforts, but where we find the interesting "spicy meatball" of data, the forgotten factoid, or the quirky 'coincidence,' we'll spotlight it here. Questions or comments? Let us know.
From Minnesota.PublicRadio.Org on January 3, 2008.
Listen Here