October 23, 2012 – 11:31 am
New Hampshire remains a sneaky pro-Romney state within the Battleground context. I’m your typical jaded conservative in the Northeast frustrated with the reflexive liberal bent to this region. New Hampshire is that last holdout and I have to say this election far more than any other in recent memory every time I look at the state I really think it is going to break for Romney. The voters know him well due to its proximity to Boston and Obama’s activist policies don’t mesh well with the “Live Free or Die” state making it ripe for the taking. The latest from American Research shows Mitt Romney with a slender 2-point lead, 49 to 47. Independents are dead even 48 to 48 and that’s an important voter group often accounting for over 40% of the New Hampshire electorate.
The party ID is R +6 (Dem 29, Rep 35, Ind 36). In 2008 it was also D +2 (Dem 29, Rep 27, Ind 45) and in 2004 it was R+7 (Dem 25, Rep 32, Ind 44). This poll is very low on Independents and too high on Republicans. It is within the 2004 and 2008 bounds but still a bit too many Republicans. I’d prefer something closer to R +2. Obama is stumping in the Granite state so his poll numbers must be showing not much different in this poll.
| For President |
Percent |
| Barack Obama |
47 |
| Mitt Romney |
49 |
| Other |
1 |
| Undecided |
3 |
October 23, 2012 – 11:14 am
American Research group weighs in with a tight poll of Nevada showing President Obama with a 2-point lead, 49 to 47. Nevada could be a tough state for Romney to flip. It’s disparate population make it a tough state to poll with Republican support often showing higher in polling than at the voting booth — a change from most states. My gut says Nevada may go to Romney but it will do so by 2-3 points below his national vote margin. That is, If Romney wins nationally by 5, he wins Nevada by 2-3. If Romney wins nationally by 1-2 points, he loses Nevada. But Romney is in Henderson, Nevada today and Reno, Nevada tomorrow so it’s not like he isn’t going all out for these 6 electoral votes. The early voting numbers are also a boon to Romney at this juncture, but he must begin outperforming with Independents if he wants to secure the state. In this ARG poll, Obama is leading with Independents by 6-points, 50 to 44.
The party ID is D +6 (Dem 43, Rep 37, Ind 20). In 2008 it was D +8 (Dem 38 Rep 30, Ind 32) and in 2004 it was R +4 (Dem 35, Rep 39, Ind 26). This is a shade closer to 2008 but Democrats did close strong on voter registrations to mostly reverse what had been great GOP gains relative to 2008. The enthusiasm gap may play a part in this favoring the Democrats a bit too much, though.
| For President |
Percent |
| Barack Obama |
49 |
| Mitt Romney |
47 |
| Other |
1 |
| Undecided |
3 |
October 15, 2012 – 12:31 pm
The latest from ARG in Virginia has Mitt Romney nominally ahead by 1-point 48 to 47. The party ID is D +5 (Dem 40, Rep 35, Ind 25) which compares to 2008 of D +6 (Dem 39, Rep 33, Ind 27) and 2004 of R +4 (Dem 35, Rep 39, Ind 26). Pretty aggressive in favor of Team Obama here on turnout. The male/female breakdown is a hair more towards men than 2008 but only 1%. The racial breakdown is close but a shade higher on Whites (2%) over the 2008 turnout and 1% lower on African-Americans. The deciding factor in Virginia will be the knife fight over turnout and Mitt Romney has met the challenge this year unlike Obama’s incredible advantage in this area last time so I’m comfortable with the direction Virginia is trending.
| For President |
Percent |
| Barack Obama |
47 |
| Mitt Romney |
48 |
| Other |
1 |
| Undecided |
4 |
October 15, 2012 – 12:20 pm
Haven’t seen a lot in Iowa recently but ARG comes to the rescue with a dead heat poll of the Hawkeye State, 48 to 48. The Party ID is D +1 (Dem 35, Rep 34, Ind 31) versus D +1 (Dem 34, Rep 33, Ind 33) in 2008 and R +2 (Dem 34, Rep 36, Ind 30) in 2004. Although it matches the 2008 turnout we see in Iowa that wide swings in party affiliation don’t occur like in other states so it’s hard to take issue when they are off by 1 or 2 % at most. The male/female split is marginally more aggressive towards women but close enough to the 2008 split. Otherwise the numbers are pretty clean. Good poll for Romney and another argument for the media talking heads to switch this state back to toss-up where it always was:
| For President |
Percent |
| Barack Obama |
48 |
| Mitt Romney |
48 |
| Other |
1 |
| Undecided |
3 |
October 12, 2012 – 11:26 am
The latest from American Research Group shows Romney with a 4-point lead in New Hampshire, 50 to 46. The party ID in the survey was R +6 (Dem 29, Rep 35, Independent 36). In 2008 it was also D +2 (Dem 29, Rep 27, Ind 45) and in 2004 it was R+7 (Dem 25, Rep 32, Ind 44). A fairly aggressive turnout in favor of Romney but at least not outside the 2004 to 2008 outer bounds.
| Candidate |
Percent |
| Barack Obama |
46 |
| Mitt Romney |
50 |
| Other |
1 |
| Undecided |
3 |
Interview dates: October 9-11, 2012
Sample size: 600 likely voters
Margin of error: ± 4 percentage points, 95% of the time
October 12, 2012 – 11:14 am
The latest from American Research Group shows another lead for Romney in Florida, this time 3-points 49 to 46 over President Obama. The party ID in the survey was D +3 (Dem 40, Rep 37, Independent 23). In 2008 it was also D +3 (Dem 37, Rep 34, Ind 29) and in 2004 it was R+4 (Dem 37, Rep 41, Ind 23). Pretty aggressive turnout in favor of Obama for a poll where he’s losing by 3 … not good.
| Candidate |
Percent |
| Barack Obama |
46 |
| Mitt Romney |
49 |
| Other |
1 |
| Undecided |
4 |
Interview dates: October 8-11, 2012
Sample size: 600 likely voters
Margin of error: ± 4 percentage points, 95% of the time
October 9, 2012 – 11:52 am
The post-debate surge carries over into Ohio where American research group find Mitt Romney narrowly leading President Obama by 1-point, 48 to 47. The party ID in the survey was D +9 (Dem 42, Rep 33, Ind 25) which compares to 2008 of D +8 (Dem 39, Rep 31, Ind 30) and in 2004 it was R +5 (Dem 35, Rep 40, Ind 25). So a better turnout than 2008 at Obama’s peak of enthusiasm versus an unarmed opponent and yet Obama still trails by 1-point. Holy-moly that debate was bad. Why on earth is the Obama campaign still talking about it?
| For President |
Percent |
| Barack Obama |
47 |
| Mitt Romney |
48 |
| Other |
1 |
| Undecided |
4 |
October 9, 2012 – 11:46 am
American Research Group (ARG) out with a new slate of polls confirming the expected surge from Romney after last week’s debate. In the Colorado survey Romney opens a 4-point lead, 50 to 46. The party ID was R +1 (Dem 33, Rep 34, Ind 33). This compares to 2008 of R +1 (Dem 30, Rep 31, Ind 39) and 2004 R +9 (Dem 29, Rep 38, Ind 33). So interestingly it is identical to the 2008 turnout but Romney leads by 4. This is huge for Team Romney. While I crush the pollsters for turnout models like this, always arguing for a split between 2008 and 2004, Colorado is a state trending Democrat with registrations and political affiliation in every election since 1996 so a split between 08 and 04 may not be perfect in this state. That said, the Obama 2008 turnout also isn’t happening, it’s a likely voter survey, Romney hits 50% threshold and the 3rd Party threat is minimal so by any measure this is a great poll for Mitt Romney:
| For President |
Percent |
| Barack Obama |
46 |
| Mitt Romney |
50 |
| Other |
1 |
| Undecided |
3 |
September 28, 2012 – 4:39 pm
As the survey groups get more reasonable, the polls tighten … funny that. The latest from American Research Group in Virginia shows President Obama with a 2-point lead 49 to 47. The party ID is below Obama’s 2008 turnout but still well off the 2004 result.
Party ID:
ARG: D +3, Dem 39, Rep 36, Ind 25
2008: D +6, Dem 39, Rep 33, Ind 27
2004: R +4, Dem 35, Rep 39, Ind 26
| For President |
Percent |
| Barack Obama |
49 |
| Mitt Romney |
47 |
| Other |
1 |
| Undecided |
3 |
September 28, 2012 – 4:22 pm
I’m always jealous of the “Live Free or Die” motto of New Hampshire. It’s probably one of the reasons the state is so competitive in most elections — plenty of Independent thinkers. Romney is going to have to pick up his game with Independents as he trails the President 52 to 41. Overall President Obama maintains a 5-point lead 50 to 45 in a survey more reasonable than many previous ARG polls lately. The Party ID was dead even at Dem 30, Rep 30, Ind 40 versus D +2 in 2008 (Dem 29, Rep 27, Ind 45) and R +7 in 2004 (Dem 25, Rep 32, Ind 44).
| For President |
Percent |
| Barack Obama |
50 |
| Mitt Romney |
45 |
| Other |
1 |
| Undecided |
4 |
September 24, 2012 – 2:32 pm
We’re going to chalk this up to just being a bad week for American Research Group because their Battleground State polls last week were at least reasonable assumptions on party breakdown (except for Colorado), but today’s are straight out of David Axelrod’s fantasy turnout on election day. Florida is only the last in an ugly week of unrealistic polls from ARG. Their voter turnout model is D +6 (Dem 42, Rep 36, Ind 22) versus D +3 in 2008 (Dem 37, Rep 34, Ind 29) and and R +4 in 2004 (Dem 37, Rep 41, Ind 23). Too many Democrats and not enough Independents. Romney leads by 1-point among Independents and both candidates have plenty of work to do securing their own base. Obama garners 83% of Democrats while Romney holds sway with 84% of Republicans.
| For President |
Percent |
| Barack Obama |
50 |
| Mitt Romney |
45 |
| Other |
1 |
| Undecided |
4 |
Interview dates: September 20-22, 2012
Sample size: 600 likely voters
Margin of error: ± 4 percentage points
September 24, 2012 – 2:09 pm
Just kidding, but that’s probably what you’d have to believe to put any weight in this latest American Research Group poll. Iowa is one of the more constant states in America when it comes to party identification, especially for Presidential contests. From the three elections we have data for the change has been the following: 2004 R +2, 2008 D +1, 2010 R +4. Somehow after that sequence and the litany of reasons for people to sour on Obama, ARG comes out with a poll over-sampling Democrats by 4 points or D +4. This is the state that voted out every state supreme court justice it could in 2010 because the body approved a measure in favor of gay marriage. I’m just sure President Obama’s embrace of this issue warmed him right up to this state. That said, here’s the party ID breakdown of D +4 in the ARG poll (Dem 36, Rep 32, Ind 32) versus D+1 in 2008 (Dem 34, Rep 33, Ind 33), and R +2 in 2004 (Dem 34, Rep 36, Ind 30). Romney is dead-even with Obama among Independents (47 to 47), GOP voter registration advantage today is at 30k and growing versus a -108k deficit in 2008 yet somehow they construct a poll with a growing Democrat advantage? Yea, I don’t think so either. Note also, many of the numbers are quite similar to the Nevada poll. Not just the top-line results that are identical. Obama gets 92% of the democrat vote in both poll and Romney gets 87% of the Republicans in Iowa and 88% in Nevada. The age breakdown preferences are also identical (assuming the Iowa typo should read 51). I’m wondering if there wasn’t something in the question wording, but no matter. The survey is unrealistic as it is:
| For President |
Percent |
| Barack Obama |
51 |
| Mitt Romney |
44 |
| Other |
1 |
| Undecided |
4 |
Interview dates: September 20-23, 2012
Sample size: 600 likely voters
Margin of error: ± 4 percentage points
September 24, 2012 – 1:43 pm
We beat up on these guys below for their national poll, but this Nevada one looks even worse. President Obama has a 7-point lead over Mitt Romney 51 to 44 in the Silver State. Obama gets 92% of Democrats and Romney gets 88% of Republicans and also 53 % of Independents yet he trails by 7%. Is this Michigan or Massachusetts? The party ID breakdown is D +5 (Dem 42, Rep 37, Ind 21) versus 2008 of D +8 (Dem 38, Rep 30, Ind 32) and 2004 R +4 (Dem 35, Rep 39, Ind 26). The only two takeaways are Romney needs to work harder for the GOP vote and this poll woefully under-samples Independents who favor Romney 53 to 41. Romney has work to do in the state, but this poll isn’t reflective on the state of affairs in Nevada:
| For President |
Percent |
| Barack Obama |
51 |
| Mitt Romney |
44 |
| Other |
1 |
| Undecided |
4 |
Interview dates: September 20-23, 2012
Sample size: 600 likely voters
Margin of error: ± 4 percentage points
September 24, 2012 – 8:48 am
This is a Battleground State blog but every now and again I’ll blog a national poll because t either provides a good read on the electorate or it’s so bad I feel the need to point that out. This poll, like far too many over the last couple of weeks, falls into the latter category. American Research Group constructed a survey where Obama ‘s locks down 90% of the Democrat, Mitt Romney locks down 92% of the Republican vote and also Romney leads among Independents by 7% (50 to 43) , yet he trails overall. That simply won’t happen in this election. If each candidate locks down his base and one candidate is disproportionately carrying Independents, that is the candidate who will win. In some of the recent state polls that person was Barack Obama, but in this poll it is clearly Mitt Romney. The party ID weighting was D +6 (Dem 40, Rep 34, Ind 26) just below the 2008 best in a generation for Democrats D +7 (Dem 39, Rep 32, Ind 29) and well-off the 2004 election where party ID was tied (Dem 37, Rep: 37, Ind 26).
Demos: The White Vote
Additionally there are some quirks in the demographic make-up. The white vote (a slight Romney demo) comprises 77% of the electorate which was true in 2004 but it dropped to 74% in 2008. This is a major point in polling and for the election. The percentage of white voters comprising the electorate has been steadily and consistently dropping ~3% in each Presidential election since 1992. That would lead poll weightings to drop the white percentage to 71 or 72% (most I’ve seen say 72% which is defensible for various reasons). This issue is a large bone of contention with Democrats and in this instance they have an argument.
African-Americans
Interestingly Obama leads 92 to 5 among African-Americans. This is worse for Obama than it might seem. That disparity is closer to the historic norm (not 2008′s hopey-changey apex) for African-Americans voting Democrat. This is part of the enthusiasm drop I often complain about which I believe will dampen Democrat turnout on election day — making most of the polls over-sampling Democrats null and void.
Obama’s 47% problem
“A total of 47% of likely voters say they would never vote for Obama in the general election and 44% of likely voters say they would never vote for Romney in the general election.”
Off-topic: As a reminder I don’t blog PPP polls because they purposely skew polls to further an agenda. They’re not a polling firm who happens to be Democrats. They are a polling firm who skews polls to further the Democrat’s agenda.
September 18, 2012 – 3:25 pm
Mitt Romney’s home away from home in New Hampshire remains as tight as ever. The latest from American Research Group shows Obama leading by 2 -points:
The following results are based on 558 completed telephone interviews among a statewide random sample of registered voters in New Hampshire. Of the 558 interviews, 167 were among registered Republicans, 156 registered Democrats, and 235 undeclared voters. The interviews were conducted September 15 through 17, 2012. The theoretical margin of error for the sample of 558 registered voters is plus or minus 4.2 percentage points, 95% of the time, on questions where opinion is evenly split.
| For President |
Percent |
| Barack Obama |
48 |
| Mitt Romney |
46 |
| Gary Johnson |
2 |
| Undecided |
4 |
September 14, 2012 – 12:19 am
American Research Group has a new poll in Ohio and unlike the Colorado poll, the demographics underneath the top line results conform to type. Romney leads Obama + 8 among men while Obama leads +10 among women. Romney leads among whites by +10 and while Obama’s lead with African-Americans is unsurprisingly high, it does conform to historic norms of 90% support, below the incredible 96% in 2008. Interview dates: September 10-12, 2012. Sample size: 600 likely voters. Margin of error: ± 4 percentage points, 95% of the time.
| For President |
Percent |
| Barack Obama |
48 |
| Mitt Romney |
47 |
| Other |
1 |
| Undecided |
1 |
September 14, 2012 – 12:02 am
American Research Group does a variety of polling but the cross-tabs aren’t provided so we’re left with just the results. There are more than a few interesting things in the results as the gender gap both was is very muted (Romney +1 with men, Obama +4 with women). Also Obama was only +15 with Hispanics which was curious. Interview dates: September 10-12, 2012. Sample size: 600 likely voters. Margin of error: ± 4 percentage points, 95% of the time. Note the party ID favors Democrats by 2-points (Dem 34, Rep 32, Ind 34) which is meaningfully greater than Obama’s 2008 performance which was actually R +1 (Dem 30, Rep 31, Ind 39).
| For President |
Percent |
| Barack Obama |
49 |
| Mitt Romney |
47 |
| Other |
1 |
| Undecided |
3 |
A new poll out American Research Group shows Obama leading 51 to 43 among registered voters. This is important for Obama for a couple of reasons:
- First, an 8-point lead in a Battleground state is a sizable lead worthy of some confidence boosting in the Obama campaign
- Second, this state is in Mitt Romney’s back yard and voters are likely very familiar with him, yet Obama has a sizable lead
- Finally, President Obama is above 50% in this poll — something he almost never is in any other Battleground state polls
Despite these positives, there was a sneaky silver lining for Mitt Romney in the same poll. Buried at the very bottom was this nugget:
Among 417 registered voters in New Hampshire saying they will definitely vote in November, Obama leads Romney 49% to 46%, with 6% undecided.
This is the far more reliable “likely voter” screen pollsters use when we get closer to election day. At this stage of the election, a 3-point gap with 6% undecided is fertile territory for the challenger.
Interestingly, New Hampshire is one of the few Battleground states, where by registration, Republicans usually outnumber Democrats. In 2004 the D/R/I split was: 25/32/44 or R+7. In 2008 the party ID split was 29/27/45 or D+2. The American Research Group party ID was 30/34/36 or R+4:
The following results are based on 535 completed telephone interviews among a statewide random sample of registered voters in New Hampshire. Of the 535 interviews, 180 were among registered Republicans, 161 registered Democrats, and 194 undeclared voters. The interviews were conducted June 21 through 24, 2012. The theoretical margin of error for the sample of 535 registered voters is plus or minus 4.2 percentage points, 95% of the time, on questions where opinion is evenly split.
